Behavior and Speciation Mechanisms

Barry Sinervo©1997

Index

Modes of Speciation Classified by Geography

Clinal Variation

Parapatric Speciation and Allo-Parapatric Speciation

Ring Species in Gulls

Paint Experiment in Gull Species Recognition

Ring Species in Salamanders and the Evolution of Mimicry

Sympatric Speciation

Polyploidy

Reproductive Character Displacement and Mate Calls in Frogs

Production of Asexual Parthenogenetic Species

Multiple-Niche Polymorphism in Insects and Host-Plant Preference

Allopatric versus Sympatric Speciation: Case Studies

Conditions for Sympatric Speciation -- Assortative mating

Cultural Inheritance of Song and Speciation in Darwin's Finches

Deme Recognition and Female Mate Choice

Founder Effect Speciation

Runaway Process, Sexual Selection, and Speciation

Species are groups of actually or potentially interbreeding populations, which are reproductively isolated from other such groups.

Mayr 1942

Biological Isolating Mechanisms

Ernst Mayr's formulation of the biological species concept clarified a vexing question for evolutionary biologists: What is a species? Mayr's view of the species concept has widespread acceptance owing to its simplicity. However, how species arise by the process of natural selection remains a source of great controversy. The problem is that the process of speciation takes a very long time to complete, and it occurs on a very large geographic scale. In addition to Mayr's definition additional concepts are quite useful in delimiting species.

Hybrid sterility definitely separates two species, although it is not the sole criterion. Hybrid sterility is a special instance of reproductive isolation. In the cases of Reproductive isolation between species the barriers to mating do not just have to be due to incompatibility between genomes. The blocks could be due to behavior or morphology. These isolating mechanisms which serve to limit gene flow between species are categorized in terms of pre- or post-mating mechanisms:

A) Premating isolating mechanisms -- prevent union of gametes -> zygote

mates do not meet (seasonal or habitat isolation)
mates meet but do not mate (ethological or behavioral isolation)
mates meet but no sperm transfer (mechanical isolation)

B) Post-mating isolating mechanisms -- varying degrees of hybrid sterility or fitness

sperm transfered but dies before fertilization
zygote dies
zygote produces an F1 adult that has reduced viability (survival)
hybrid is viable, but partially or completely sterile (fecundity) or the F2 is deficient.

The answer to the adaptive value of pre- versus post-mating isolating mechanisms is revealed by asking the following questions. Which of the two mating isolation mechanisms is the most effective in reducing the risk of producing low fitness offspring? Which mechanisms is most efficient in terms of time and lost reproductive opportunity?

An organism that mates with a different "semispecies" may produce low viability offspring and this individual would have lower fitness compared to an organism that discriminates against such semispecies and mates with members of its own semispecies. The evolution of species discrimination mechanisms are clearly an advantage over one that only possess post-mating isolating mechanisms. Such behavioral descrimination ability should evolve rapidly in the area of incipient speciation, under the right conditions.

Biological barriers to interbreeding undoubtedly arise during the process of speciation because such blocks have selective value. Behavioral isolation can be the strongest form of reproductive isolation. The clearest examples of biological species concept in action are provided when two species which have overlapping geographic ranges do not interbreed owing to differences in pre-mating behaviors. One species or both species refuse to engage in the mating rituals because the stereotyped mating rituals are to different. Species might simultaneously possess both premating and post-mating isolating mechanisms in which case the probability of interbreeding is even more remote.

What are the right conditions? This is a very difficult question to answer. It requires an understanding of the field of population genetics and the crucial parameter is how much migration occurs from areas where each semispecies exists in its pure form. A second parameter, has to do with the degree of genetic similarity between the two pure semispecies.

Migration is the movement of animals between what are referred to as populations of interbreeding animals. Rates of migration are usually expressed as the proportion of individuals in a population that disperse from one population to the next each generation. If migration between the region of contact between the two semispecies and the rest of the semispecies ranges is too high, then the two semispecies will become homogenized into one species even if such interbreeding results in lower fitness because of hybrid unfitness. If migration is low enough, then the two pure semispecies will eventually become differentiated into two species owing to the evolution of isolating mechanisms. Evolution of such isolating mechanisms is contigent upon the degree of genetic similarity between the two semispecies.

Large differences in genetic similarity between semispecies are more likely to produce unfit hybrids than small differences in similarity. If the two semispecies have many genetic loci that differ in the kind of allele that they possess, then those alleles might not work very well together, and more importantly the way two different genetic loci work together may begin to breakdown. This is referred to as genome-wide genetic interactions. The correct operation of one genetic locus (lets say an enzyme) may require a particular form of the enzyme that is produced by a different locus (e.g., epistasis). These enzymes are most likely to be compatible in their function if they come from the same species. If they are from different species, they may not integrate well with other loci in the organism and the hybrid individual may die. Imagine such genetic interactions occuring at all possible loci (e.g., 10,000 loci in Drosophila or perhaps 100,000 loci in humans). The probability of a hybrid being more unfit has to do with how many loci in each semispecies are homozygous for different alleles. Hybrid breakdown occurs when many of these loci are fixed for different alleles in the two subspecies and the production of offspring by the hybrids is not possible because and F1 X F1 produces too many incompatible gene by gene combinations.

Genetic similarity drops as we change from the different levels of differentiation, going up the heirarchy of differentiation: populations -> races -> subspecies -> semispecies -> sibling species -> species. By the time species are compared there is generally little or no gene flow between species. As genetic similarity drops, the differences increase and the possibility of hybrid breakdown increases. A species is thought to have a coadapted genome in which many genes are finely tuned by the proper interactions with other genes.

Clearly genetic differentiation is occuring during speciation but it is not enough to define a species. It can vary from organism to organism and from mode of speciation to the next. In practice, even the criterion of no gene flow between species is not the defining feature of species. As we will see below, the biological species concept begins to break down in areas or at times of speciation or incipient speciation (semispecies). In such cases, alleles can leak across the hybrid zone from one species into another.

Index

Modes of Speciation Classified by Geography

To understand the origin of species differences it is necessary to consider how genetic differences arise between species. Geographic subdivision decreases gene flow. Depending on the population size, genetic drift can play a major role in promoting genetic differentiation. If population size remains low for a long period of time it is possible for genetic differences to accumulate between geographic areas by the random process of genetic drift. Selection can promote speciation because genetic changes can rapidly accumulate in one area where selection favors certain behaviors and morphology relative to another area where alternative alleles are favored by natural selection. Finally, the amount of gene flow which is usually described in terms of the migration rate is a key parameter that influences the rapidity of speciation. By lowering gene flow, geographic subdivision promotes the process of speciation.

Models of speciation relate to the degree of geographical subdivision for Allopatric -->Allo-parapatric --> Parapatric --> Sympatric which ranges extreme to none. Another parameter that can be used to descriminate the types of speciation is related to rates of gene flow and how alleles at many genetic loci are distributed across the species range.

Allopatric Speciation -- speciation occurs in geographic isolation,
Founder Effect Speciation -- a special kind of allopatric speciation in a small isolated population on the edge of a species range
Parapatric Speciation -- speciation in adjacent populations with gene flow,
Allo-Parapatric Speciation -- populations initially separated (allo-) but then secondarily come into contact with subsequent parapatric speciation,
Sympatric Speciation -- speciation within a panmictic population.

Figure 2. Models of speciation based on geographical subdivision.

Index

Diagnosing Models of Speciation from Clinal Variation

Patterns of variation across the species may provide some basic diagnostic features that distinguish the four modes of speciation. A cline is any systematic change in a trait across the range of a species. Clines can be abrupt or gradual. Presence of clines in morphological or behavioral traits are used as evidence for one mode of speciation over another (see Figure 3).

Presence of a cline would suggest a particular mode of speciation has operated in the past. Presence of a geographical barrier at the cline would reinforce the notion that the speciation event was due to allopatric speciation.The change in behavior traits, morphological traits, allozymes, or some other form of genetically-based variation resembles a step function. A very abrupt change in traits where the two species come into contact would suggest little or no gene flow in the present day. Lack of gene flow in the present day could be could be due to the evolution of reproductive isolation mechanisms during allopatric speciation in the remote past and the original barrier has been removed.
The Shape of the Cline might also be indicative of gene flow. Clines can be gradual across a hybrid zone and span 1000 of km or clines can be very abrupt and occur over a few 100 m. Is the cline maintained by persistent selection that favors one form of the species on one side of the cline and another form on the other side of the cline? If yes, then the two species along with their hybrid zone might be a candidate for the parapatric model of speciation. In the case of a cline that is being maintained by persistent selection, the cline may have arose as a primary response to different selection regimes in the species range.
Origin of the cline or hybrid zone is another attribute that might distinguish models of speciation. Is the cline due to secondary contact? In this case, the two species were in allopatry in the remote past and brought into secondary contact and weak gene flow began once the sibling species were rejoined at the hybrid zone. However, complete reproductive isolation did not occur during allopatry, but rather the the species differences were reinforced once the two species came into secondary contact. The cline is said to have evolved by allo-parapatric speciation.
If there is no cline and the two species can be found co-occuring across a large geographic range, sympatric speciation may have produced behavioral isolation. A transect across the range of the two species would show isolated pockets of one pure genotype, or the other pure genotype, with no present day gene flow between the two species.

Figure 3. Hypothetical change in frequency of allele frequency across the range of two species under allopatric, parapatric, and sympatric models of speciation.

Index

Allopatric and Parapatric Speciation

Any geographic factor that severely restricts gene flow will allow populations of a species on one side of the barrier to begin to differentiate from species on the other side of the range. Given enough time, the genetic differences became large enough to warrant calling the "races" on one side or the other a different species.Allopatric speciation occurs under the following conditions:

isolation of a colony (e.g., island and mainland)
division of range by extrinsic barrier (e.g., mountain range) or extinction of populations in the intermediate parts of the range.
isolation by geographical distance.

Parapatric versus Allo-Parapatric Speciation

In order for a species to arise by parapatric speciation, there must usually be a very strong change in the environment. Natural selection on one side of the environmental transition favors one set of traits compared to the traits favored on the other side. The existence of clines as evidence for parapatric speciation is generally not considered very good evidence.

However, if selection is strong enough it leads to disruptive selection at the abrupt transition that parapatric speciation occurs. The fitness of an individual is determined by the fraction of genes from each parental type. Individuals that are pure for each parental type will tend to have higher fitness than individuals with a genome that is formed from a mixture of the two parental combinations. These hybrids come from crosses between the two parental forms. However, animals that interbreed (e.g., mate with a different type) at such hybrid zones tend to show underdominance in fitness, that is their progeny the heterozygotes (hybrids) have lower fitness. Another way to think of this pattern is to view it as disruptive selection against individuals heterozygous for a large number of loci. The genes in the heterozygous condition have lower fitness than homozygous configurations because of the strong environmental differences on either side of the contact zone.

If there is any viability of the hybrids, alleles may leak across and introgress across the zone. Leakage of alleles across the hybrid zone is referred to as Introgressive hybridization. Some alleles may spread quite far across the zone whereas others will be quite abrupt and coincide with the hybrid zone.

The big question, can we distinguish between primary and secondary contact? In principle, we cannot ,which is the case and that is why the models of speciation are so fiercely debated. The process of speciation is by its very nature a historical process. The origin of the cline is a key component of all all models of speciation, and this event can only be inferred from the patterns of clinal variation that we see across the present-day species ranges. We see many fully formed species, and while we also see incipient species in the process of formation, the actual cases are difficult to resolve as allopatric versus parapatric speciation, because events have occured in the remote past.

Nevertheless, studies of the geographic pattern of races in a species and their likelihood of interbreeding have provided us with splendid examples of apparent allopatric speciation. Nothing is more challenging to the imagination that a special process of allopatric speciation which produces a ring species.

Index

Ring Species in Gulls

Ring Species arise from a peculiar form of allopatric speciation that takes place when the center of a species range is unoccupied because the habitat is unsuitable. Adjacent races all around the ring will interbreed, however the races at the termini are so divergent, that there is no interbreeding, and they can exist in sympatry without interbreeding.

The species range of Sea Gulls in the Arctic Ocean represents a classic example of a ring species. The polar ice cap limits the species range of Sea Gulls to a circumpolar ring.

There are a number of races of Sea Gulls as one traverses around the globe. Races from America freely interbreed with races from Europe. Races from Siberia freely interbreed with races from America. Races from Siberia freely interbreed with races from the Caucauses. However Central European races do not interbreed with those from Western Europe. All along the ring there is gene flow, but where the two ends of the ring meet in Europe there is no gene flow. The ring species of gulls forms a cline in that there are small changes in coloration and behavior of the gulls along the length of the ring. However, not enough mate discrimination occurs between adjacent populations until the ends of the ring meet in Europe.

A Paint Experiment in Gull Species Recognition

Work by Smith on a different species of sea gull demonstrates that the behavioral isolating mechanism which operate in gull species recognition may be quite simple. Smith investigated the signal that is used to isolate the two species into Larus hyperboreus, the glaucous gull and Larus thayeri, Thayer's Gull. The glaucous gull has a yellowish eye ring, and Thayer's Gull has a purple eye ring. Smith painted the eye rings of Thayer's males yellow, and this allowed Thayer males to successful pair with glaucous females. However, the males would not mate with the glaucous females until Smith painted the glaucous females purple. Such paint experiments are a powerful means of demonstrating the meaning behind signals and the role in species recognition.

Index

Ring Species in Salamanders and the Evolution of Mimicry

A ring species of salamander, Ensatina scholtzi, is found in higher altitude regions of California such as the sierra and coast ranges, but is conspicuously absent from the great Central Valley of California. Abundant evidence of gene flow along the cline is provided by comparing the frequency of electrophoretic alleles in adjacent populations of the salamander. Gene flow is present throughout the coast range from south to north, from the coast range to the ranges that cross northern California to the sierras, all down the length of the sierras, from the sierras to the transverse ranges of california, but gene flow abruptly ends in a sharp dividing line located in southern California. The ring of gene flow that traverses thousands of miles around the perimeter of California finally expires and speciation appears to have run its course.

The ring species of salamander is also noteworthy in terms of its morphology and behavior. Where Ensatina co-occurs with the California Newt in the coast range, Ensatina has evolved a color pattern that resembles the toxic Newt in several regards. It has a reddish cast to its belly, and its eye is colored a brilliant yellow, both are features that are thought to be warning coloration in the newt. In Ensatina, the coloration appears to be a ruse as Ensatina is not at all toxic. We will take up the causes of such warning coloration in chapter 12 which deals with predator and prey interactions.

Index

Sympatric speciation

Sympatric speciation is the origin of two new species with the two forms in the same place. Conditions for speciation by sympatric speciation are very stringent. Gene flow between incipient species must be eliminated or severely curtailed, or reproductive isolation is not possible. Whereas isolation by geographic distance as is seen in allopatric and parapatric speciation can limit gene flow, arriving at sympatric speciation when individuals are found in the same geographic locale is a challenge. Controversy surrounds all theories of sympatric speciation that involve a gradual speciation event in sibling species that are in a panmictic, or can freely interbreeding across the range.

A key aspect of the theory of sympatric speciation involves the process of natural selection that favors the evolution of discrimination on the part of the female and also the divergence of male traits that females use for mate discrimination. Dobzhansky suggested that natural selection should favor the more finely tuned discrimination mechanisms of the female, in response to the reduced viability that results from hybridization. Thus, a key component of this theory is some form of disruptive selection that selects against hybrid phenotypes, which then favors evolution of mate discrimination mechanisms. A process referred to as reproductive character displacement should accentuate differences in courtship behavior or mate preference in sympatric populations of the two incipient species compared to the differences between allopatric populations of the two incipient species. Butlin (1987) coined the term reinforcement to describe such evolutionary interactions between incipient species, in which hybrid unfitness promotes divergence in courtship behavior or mate preference in sympatric compared to allopatric populations.

The evolution of such divergence, in the face of gene flow is a tremendous challenge for the evolution of species by the process of sympatric speciation. Felsenstein showed that the process of sympatric speciation is a fairly unlikely event. He modeled this process in terms the evolution of two species attributes: 1) the phenotypic attributes that give rise to hybrid unfitness, and 2) the genes that give rise to mate discrimination. Generally, the genes for fitness, are not expected to be linked (either by pleiotropy or by physical linkage) to the genes that give rise to mate discrimination. This key fact places a tremendous genetic constraint on speciation by the process of reinforcement (see Side Box 5.4).

Polyploidy

Despite the apparent difficulties most theories of sympatric speciation, one form of sympatric speciation is universally accepted -- the instantaneous mode in which changes in ploidy lead to instantaneous reproductive isolation. While extremely common in plants, it is less common in animals and well documented cases are found in the lower vertebrate classes: fish, amphibians, and reptiles.

For example, Gerhardt has identified a case in which a single species of Hyla tree frogs has undergone tetraploid formations. The formation of tetraploid (4n) from the fusion of two diploid parents (2n) leads to a new species of treefrog that can only breed with another individual that has similar ploidy (Gerhardt 198X). The fusion of gametes from a diploid parent and tetraploid parent produces triploid offspring which die during larval stages and the few hybrids that reach adult size are sterile (Johnson. While production of triploids can be achieved in the laboratory, triploids are rarely found in natural populations of Hyla tree frogs (Gerhardt 1982).

Because the diploid parental species achieves reproductive isolation from the tetraploid daughter species in one fell-swoop, sympatric speciation is easy to achieve. Indeed Gerhardt and his colleagues have found that the production of new tetraploid species has occurred several times in this species complex. Diploid populations of Hyla chrysoscelis probably gave rise to tetraploid populations of H. versicolor (Tymowska 1991).

Character Displacement in Mate Calling and Female Preference

Gerhardt has tested female preferences for male call frequencies of the two species of Hyla frogs in areas of sympatry and allopatry. Males of each species from areas of sympatry are more divergent in key aspects of their calls compared to males from areas of allopatry. In areas of sympatry Hyla chrysoscelis males produce a longer call than co-occuring H. versicolor. In areas of allopatry the songs of both species are of medium duration.

The key test of character displacement via reinforcement involves assessing whether females have evolved more fine scale preferences. The hypothesis of character displacement would suggest that females from sympatric populations showed stronger preferences for acoustic stimuli of the males call than females from more remote populations. Gerhardt carried out a number of phonotaxis experiments to test whether females in allopatry have evolved more fine scale discrimination of the species differences in mate call. Gerhardt created synthetic calls which allowed him to vary specific attributes such as call length rather than the fine scale aspects of the song. He then played songs from two speakers. One speaker played a male with the short synthetic call typical of the males of the females area, the other speaker played a male with longer synthetic calls. Females from areas of sympatry were much more likely to discriminate the correct conspecific short song that females from more remote areas of allopatry.

Index

Production of Asexual Parthenogenetic Species

The production of polyploid species in lizards is also quite common. For example, species of Cnemidophorus or whiptail lizards from the southwestern Deserts have produced daughter species by polyploidy repeatedly. Not only have sibling polyploid species been produced, but reproductive isolation is achieved by another interesting evolutionary event -- many of the daughter species that are produced are asexuals which means they are all female and do not require males to produce offspring. The evolution of an asexual or parthenogenetic species can eliminate gene flow even more effectively than ploidy changes alone. If sperm is not required, the newly derived clonal species instantaneously begins an independent existence. The asexual daughter species of Cnemidophorus no longer needs males to produce, however, reproduction in females of the asexual species are facilitated by male-like behavior in a curious fashion. If females participate in pseudocopulation in which one female mounts another a female in the stereotypical male-female copulatory position, the female produce more eggs, and the interclutch interval is reduced .

Interestingly, the block to the formation of hybrids is not necessarily absolute in some parthenogenetic species. Genes from a facultative asexual species could be transmitted to a sexual species. Most of the time the asexual species can produce parthenogentically. For example, asexual species of flatworms from lakes in Europe have evolved repeatedly. Flatworms are also unique in that they are simultaneous hermaphrodites, which maintain active testes and ovaries. While the testes of the parthenogenetic species are much reduced, they still produce viable sperm. The asexual species of flatworm can produce haploid sperm that it can transfer back to the sexual species, which allows for one way gene flow from the facultatively asexual daughter species, back into the parental sexual species.

Index

Multiple-Niche Polymorphism in Insects and Host-Plant Preference

Without strong assortative mating of some form sympatric speciation is an impossibility. Most theories of sympatric speciation other than ploidy changes or the evolution of a parthenogenetic species are controversial. Any sympatric speciation theory that involves a gradual change in sympatry is difficult because gene flow would swamp out the build up genetic differentiation that forms the hallmark of speciation. There is another notable exception which involves the evolution of insect preferences for feeding on certain species plants. It is thought that the following mechanism may be why there are more insect species of insects on the planet than any other organism. If so, then this mechanism is truly important from the point of view of animal behavior and diversity. Part of the reason why it is so prevalent is because insects can evolve detoxifying specificity towards their host plants in a specific enzyme -- P450 (the details of which need not concern us). However, the acquisition of such specificity should lead to rapid acquisition of host plant preferences -- a behavioral trait.

One class of models maintains that there arise multiple-niche polymorphisms. Heterozygotes are inferior because extreme specialization in detoxifying capability is necessary to successful survive and reproduce on different species of plants. Assortative mating among homozygotes is a key component of this theory. This minimally requires two loci to evolve simulatneously, and thus we establish strong linkage disequilibrium for the niche polymorphism loci and the assortative mating loci. More loci makes it unlikely to occur.

One classic and well accepted form of sympatric speciation involving animal behavior has to do with host plant choice by insects. Many host plants of insects produce very nasty chemicals in repsonse to being eaten. The chemicals are meant to deter the herbivores, thus in order to continue feeding on the plants, the insects must develop mutations that allow it to detoxify the chemicals. Because different host species produce very different toxins, the mutations that allow an insect to detoxify one host-plants arsenal of toxins do not work with another. Here are the key issues involved in this mode of speciation.

An insect must evolve the detoxifying mutations, and also evolve host-plant preferences that allow it to become very choosy about the kind of plants it is feeding on. The mutations do not have to occur simultaneously, but the detoxifying mutation probably arises first and the behavioral preference evolves as a refinement to the detoxifying phenotye because individuals that match their own detoxifying abilities with the proper host plant would be strongly favored by natural selection.

In general, two such mutations are required, because it is difficult to imagine getting a single mutation that has a pleiotropic effect on detoxifying ability and host-plant preference. The two mutations should also arise on the same chromosome in close proximity to one another. This causes the behavior for host-plant specificity to be tightly correlated with the detoxifying genes. The linkage between the two loci forms a genetic correlation between the behavioral and physiological traits. An insect that ends up with the wrong allele for detoxifying matched up with the wrong preference is in trouble (recall the problem for stripped and spotted snakes). By having the mutations on the same chromosome, a super gene for the detoxifying ability and host-plant preference is produced. It is possible for the genome to become rearranged after the mutations arise. This would require a third chromosomal rearrangement mutation which makes it less likely.

Technically, the incipient insect species which feeds on a new host plant species is found in the same area as the ancestral species, but by virtue of different host plant preferences and mating microhabitats, they no longer exchange genes. The easiest way for many insects to find suitable hosts for their own progeny would be to oviposit their eggs on the host plant. If you oviposit on your host plant, chances are mating would also occur on the plant prior to oviposition.

This speciation process not only leads to adaptive plant choice, but because other members of the insects own incipient species also choose the same plants, the insects will mate with like members of their incipient species and not with other members, and reproductive isolation is achieved.

Sympatric versus Allopatric Speciation

Assortative Mating is a Stringent Condition for Sympatric Speciation

The formation of new sexual species by changes in ploidy and the formation of asexual species are both examples in which the new species becomes reproductively isolated from the parent. How likely is sympatric speciation in a group if there is any gene flow between incipient species. Lets re-visit an old system -- the beak morphs of African seedcrackers (Smith) and speculate about what might happen to Pyrenestes in the not too distant future. If S (small-billed) birds are much more successful in feeding on small seeds and L (large-billed) birds are much more successful in feeding on large seeds, why shouldn't these two morphs begin to form a new species, each specializing on a specific species of sedge? Natural selection on the two morphs appears to promote character divergence. Selection provides a key route for the rapid accumulation of genetic differences.

However, based on the calculations of observed and expected proportions of matings between small and large-beaked morphs, we have already seen that seedcrackers mate randomly (see Side Box 2.4). In order for there to be even a remote chance of speciation in seed crackers they would have to mate assortatively. Without assortative mating the gene flow between the finch morphs would continue unabated.

Table 5.1. Observed frequency of large and small-beaked morphs is random (A) which contrasts strongly with patterns expected under strong assortative mating (B)

A) Observed random mating in seed crackers. B) Assortative Mating required for speciation.

female of the the pair

S

L

male in the pair
S

34

14

L

14

6

female of the the pair

S

L

male in the pair
S

48

0

L

0

20

Index

Cultural Inheritance of Song and Speciation in Darwin's Finches

The natural selection observed in seed-cracking African Finches may form a model for how natural selection has acted on feeding behavior and performance of Darwin's Finches on the Galapagos Islands. For Darwin, the Galapagos islands shaped his thinking like no other place on the planet. These islands continue to be a source of inspiration for field biologists interested in the evolution of behavior. Peter and Rosemary Grant, Price, Boag and Schluter have carried out a series of elegant studies that document selection on different species.

Each of the twelve species of Darwin's Finches have evolved specialized morphologies and behaviors that are adapted to their ecological niche. On some islands one finds two or three species with drastically different beak morphologies that they use in feeding. One species of finch has even evolved a woodpecker lifestyle in that it feeds on insects under bark. Interestingly this species does not use its bill to extract the insects, rather it has evolved tool using behaviors. The species picks up twigs that it uses to probe the nooks and crannies of bark and extracts insects. The males of each species also have unique species-specific songs, and females strongly prefer song types of their own species over other species.

Most of the larger islands in the Galapagos archipelago have several species of finches which raises questions regarding the speciation events that have produced the diversity of finch species. Have the species of Galapagos finches evolved by sympatry on the same island? Alternatively, have the species evolved by founder effect speciation or speciation in allopatry on adjacent islands. To get more than one species on each island requires subsequent invasion events between islands where one species disperses to another island where a second species is found. Alternatively, The dispersing species then colonizes successfully and begins to grow in population size. The two species begin competing on some of the same resources, however, by a process of character displacement each species evolves specialized behaviors and morphology that allow the species to specialize on certain resources such as plants with different sized seeds.

A possible mechanism for the evolution of species in sympatry would be the origination of a gene of major effect that causes a new beak morph to evolve. Each morph would feed on a slightly different resource, much like the morphs of seed-cracker in Africa. However, matings between morphs of seed crackers from Africa produce perfectly viable young, and these morphs are considered a single species. In addition there are not hybrids in seed crackers as heterozygotes are hidden by the dominant effect of the Large beak allele. Finally the lack of assortative mating between morphs of African finches suggest that speciation is unlikely. Available data on the success of hybrids between different species of Galapagos finches suggests that hybrids have high viability (Grant and Grant 1996). Moreover, the hybrids are intermediate in terms of many morphological traits, including beak morphology. One of the key requirements for sympatric speciation would be a lower fitness of hybrids that would then favor the evolution of discrimination mechanisms between the two sympatric species of Galapagos finches. Where hybrids are formed between species, the hybrids appear to quite viable. In the absence of such a fitness cost, it is unlikely that species recognition mechanisms evolved in sympatry.

The evolution of the species in allopatry is a much more likely mechanism in light of the lack of hybrid inviability. On each island, species would have acquired differences in morphology, but also important differences in song also arise in allopatry. When the dispersal of birds occurs among islands, gene flow is minimized because of evolved differences in song type among males, and because females tend to choose males on the basis of song. In order to understand, speciation mechanisms in Darwin's Finches we need to understand the transmission of song between generations.

Figure. X Similarity of song between father to son, and between paternal and maternal grandfathers to son.

Darwins finches transmit song from father to son. Morevoer, the transmission of song from father to son is likely to be cultural and not genetic. If song is genetic, we would expect that a male bird could transmit genes for song to their daughter. While daughters do not sing their grandfathers song, they would still be expected to pass the genes for song on to their own progeny. Consequently, we would expect there to a positive correlation between maternal grandfather's song and the mother's son, but there is no relationship. In constrast there is a relatively strong and positive correlation between the paternal grandfather's song and the son, and even a stronger correlation between the father and son. This indicates that song has a patrilineal inheritance, and moreover, is culturally transmitted.

Figure X. Transmission of songs from father to son in Geospiza fortis across five generations. While most males copy their father's song (right hand lineage), some copy the song of another male.

The work of the Grants on the Galapogos Islands demonstrates that the transmission of dialect occurs in a paternal lineage by the process of cultural evolution. Mistakes are occaissional made in the transmission of song from father to song. In such cases the progeny may learn a song from a neighboring male that is from a different species and such misimprinting may increase the likelihood of a male of one species pairing with a female of the other species.

The cultural transmission of song has several important consequences for speciation mechanisms in Darwin's Finches. First, cultural transmission of song may allow for misimprinting of song to occur. If sons learn a song during a critical developmental period, and the father dies prior to this imprinting period, a nieghbor may fill in the role of song tutor, and the son learns the neighbors song (see Figure X). In some cases, the juvenile birds learn the song of another neighbor that is a different species. While rare, such cases of misimprinting allow for genes to flow from one species of sympatric Darwin's finch to another. Finally, While females of Darwin's finch do not learn sing a father's song, they learn to recognize the song of the father and prefer to mate with males that sing the species specific song that they learned during their youth. Most of the time, species of Darwin's finch correctly learn song as males, and females correctly choose males as mates based on their song. However, in rare cases mistakes are made, and some hybridization occurs.

Figure X. Misimprinting of song in Geospiza fortis sons that had G. magnirostris neighbors.

Index

Deme Recognition Female Choice for Natal Song

An additional requirement for speciation would be selection on females to strongly prefer males with the song that they heard their fathers sing. A complete answer to this question requires and experimental approach, which owing to the sensitive nature of Galapogos Islands, is impossible to perform on Darwin's Finches. However, many birds are only sexually responsive to mates from their own natal areas. The following experiment illustrates that female birds learn songs of appropriate mates in their youth and then choose rather males that

Baker first used tudor males to teach young birds a song dialect. Baker, found that female white-crowned sparrows responded positively to male courtship songs at a higher rate if the song was derived from their home dialect compared to an alien dialect.

The key feature of this study demonstartes that female birds learned the song during the nestling phase. In addition, it is possible to fool females into preferring a truly "non-natal" song, or one from a region other than their true source population. By playing them a song that is from a different source population during the critical song learning period. For example, if tudor songs had been from Gothic the females should have preferred Gothic males at maturity and rejected Sand Creek Songs.

Songs similar to the tudor songs were presented to the females and songs from a different area were presented to the females when they reached maturity:

References:

Baker, M. C. 1981. Early experience determines song dialect responsiveness of female sparrows. Science 21: 819-821.

Index

Founder Effect Speciation

Founder effect speciation refers to a mechanism very similar to Allopatric models of speciation. The primary difference relates to the size of the founding population which colonizes some region on the very edge of a species range. If this founding colony remains isolated for a long period of time then speciation might occur. A key component of founder effect speciation relates to the genetic material that is brought in by the initial colonists. Because the number of colonists is very low, the sample of genetic material is much reduced compared to the source population. In this case, the founding population is expected to diverge from the source population from the outset. Genetic drift facilitates divergence over the long term.

Index

Runaway Process, Sexual Selection, and Speciation

Recall the elements of Fisherian runaway process and sexual selection. The essence of the theory is that a mutation for female choosiness arises in a population and the females opt for males with some kind of elaborate trait. If Fisherian runaway process is to play a role in speciation mechanisms, then it must be related to the process by which reproductive isolation occurs. How do the female descrimination traits such as that seen in Gulls evolve? What about plumage variation seen in other bird species? How is it that lizards have evolved such markedly divergent head bobs.

The theory for runaway process as independently solved by Mark Kirkpatrick and Russel Lande has a very interesting second phase.

In the first phase we saw that:

Fisherian runaway will lead to ever increasing escalation in the male trait up to the point that natural selection begins to strongly act against the handicap that the male trait imposes on the male. Once this point is reached, the second phase of Fisherian runaway sexual selection takes hold.

The populations will change in frequency by the process of what is refered to as genetic drift. Genetic drift is simply the random processes that cause small fluctuations in the frequency of an allele simply because of limited population size. By chance alone, the frequency of alleles (for choosiness and the male trait) will fluctuate up and down in frequency. In small populations the effect of genetic is expected to be quite great compared to large populations. How small is small? Well generally populations that are less than about 100 individuals. What might happen is that certain alleles for choosiness might go to fixation in a small population, or perhaps genes for the male trait.

Up to this point we have considered runaway selection and its operation on a single male trait and a single mutation for choosiness. However, as discussed above sexual selection operates on the whole organism and females might choose on the basis of a number of male traits. For example, females might choose on the basis of both tail length as well as plumage coloration.

Now imagine a number of distinct sub-populations which are experiencing genetic drift and going to fixation. Because genetic drift is an entirely random process we would expect that each sub-population would fix for different choosiness or male traits. When this occurs, we have evolved strong assortative mating mechanisms that would lead to isolation.

Another scenario for such random effects is that the mutations that lead to choosiness or the male trait might differ in each sub-population. Fisherian runaway selection might lead to rapid evolution in each population according to such initial differences. And the populations would be expected to diverge rapidly in their mating preferences. The end result of runaway process is reproductive isolation by mating preferences -- the key ingredient for speciation.

Index

Back to the Syllabus