Reinforcement

In my final blog about the book Speciation, I will reflect on Coyne and Orr’s discussion of reinforcement. In case you don’t remember from earlier posts, reinforcement is the enhancement of prezygotic isolation between populations in sympatry by natural selection. If populations experience divergent evolution while in allopatry, and then come into secondary contact and hybridize, the hybrids may be less fit than offspring produced by individuals from the same population. If the low fitness of hybrid offspring leads to natural selection on individuals not to hybridize (or to mate with others with similar phenotypic characteristics) then this selection is called reinforcement. Reinforcement leads to the evolution of adaptations that increase the premating isolation of the populations.

Reinforcement is distinct from reproductive character displacement. This term was coined by Butlin to describe situations in which taxa come into secondary contact and produce hybrids with are completely unfit. In this case gene flow is not possible, as the hybrid will not be able to produce surviving offspring. This means the two populations are already good species, and so any further enhancement of prezygotic isolation would have nothing to do with speciation and should not be considered reinforcement. In contrast, in the case of reinforcement the hybrids are only partially unfit, so gene flow is possible, and the speciation process is not complete. Enhancement of prezygotic isolation in this case is relevant to speciation.

As Felsentein (1981) explained, it should be theoretically difficult for reinforcement to evolve. Reinforcement requires that individuals assortatively mate with others that share similar phenotypic traits. This means that there must be strong linkage disequilibrium between alleles underlying the characters that have undergone divergent evolution between the two populations, and the alleles underlying mating preferences. Felsenstein pointed out that although this linkage may be generated by selection, recombination often destroys it. Unless the two loci are tightly linked, it should be difficult to overcome recombination.

One way to look for evidence of reinforcement is to quantify the level of prezygotic isolation between populations in sympatry and in allopatry. The prediction would be that prezygotic isolation would be higher between the populations in sympatry. This is because there would only by hybridization between the two populations they come into contact, and so in sympatry there would be an opportunity for natural selection to increase the strength of adaptations leading to prezygotic isolation, where in allopatry there would not.

Could you use this kind of a prediction to look for reinforcement in white sands lizards? Or is reinforcement even a relevant concept to apply to this system? Upon first examination I thought it would make sense that reinforcement could be going on at the ecotone. After all, you have white populations of lizards in white sands, dark populations in the various surrounding desert habitats, and intermediately colored lizards at the ecotone. So it seemed to me that these intermediately colored individuals could be hybrids of dark and light populations that had diverged in allopatry and come back into secondary contact. Turns out it is likely that this isn’t the case. For one thing, there really aren’t any light populations and dark populations located close enough together to suggest that frequent hybridization of parental populations at the ecotone is common. In addition, the genetic basis for dark and blanched coloration in Aspidoscelis and Sceloporus is a simple dominant recessive relationship. In Aspidoscelis the mutation in Mc1r causing blanched coloration is recessive, while in Sceloporus it is dominant (Rosenblum et al 2009). So for example, if you had a mating between a dark (heterozygous or homozygous dominant) Aspidoscelis and light (homozygous recessive) Aspidoscelis, you might get either a light colored hybrid or a dark colored hybrid, but you shouldn’t see an intermediately colored individual. So, though it is interesting to contemplate where these intermediately colored individuals are coming from, their existence may not be relevant in the case of our discussion of reinforcement. It seems from the evidence we have currently that reinforcement of prezygotic isolation is not an important mechanism of speciation in the white sands system.

Behavioral Isolation

This week we read chapter 6 or Speciation, titled “Behavioral and Nonecological Isolation.” Coyne and Orr describe 4 main categories of “evolutionary forces” that lead to behavioral isolation. These include 1) selection on mate preference, 2) selection on certain traits, 3) genetic drift, and 4) nongenetic mechanisms.

1) In the first category, initial selection pressures act on mate preferences of individuals either directly or indirectly. Direct selection on mate preferences increases the immediate fitness of the chooser by improving the chooser’s ability to acquire resources, or allowing the chooser to avoid deleterious features associated with mating. Indirect selection on preference includes selection that does not increase the chooser’s immediate fitness. For example, in the runaway model of selection preference for a certain trait is genetically correlated with the gene conferring that trait.

2) In the second category there is selection on traits that improve the attractiveness of the bearer to the opposite sex, improve the ability of the bearer to overcome competition from members of the same sex, or that facilitate species recognition. This last category includes trait changes that evolve through natural selection with preferences coevolving. It also includes the evolution of reinforcement, which improves the ability of individuals to discriminate conspecifics from heterospecifics in order to reduce maladaptive hybridization.

3) Genetic drift may affect behavioral isolation if nonselective changes in allele frequencies affect a signal or a preference.

4) Nongenetic mechanisms of isolation include cultural drift, and host parasitism.

It seems to me that direct selection on traits would be most important for behavioral isolation of populations of white sands lizards and dark soils lizards, especially through species recognition. The two subcategories for the species recognition section seem quite similar to me. In the first scenario, natural selection causes a trait to change, and preferences for that trait coevolve because mating with individuals with that trait will improve the fitness of offspring. In the second scenario, evolution favors discrimination between individuals with different traits to reduce the chance of producing less fit offspring. It almost seems like two sides of the same coin; recognize a trait and mate with that individual to create more fit offspring, or recognize a trait and don’t mate with that individual to avoid making less fit offspring.

Coyne and Orr distinguish the two by emphasizing that in the first scenario there need not be a closely related sympatric species present. Divergent selection in different habitats could cause changes in traits and preferences for those traits, which could lead to behavioral isolation between locally adapted populations. The book cites the example of divergent selection on body size for resource acquisition in benthic and limnetic morphs of sticklebacks .

In the case of reinforcement, there must be a related sympatric species present, and the fact that hybrids are less fit leads to selection favoring the avoidance of hybridization through increased ability to recognize conspecifics. So I guess if the difference between these two ideas is just whether there is divergence of locally adapted populations or of sympatric sister taxa, then the lizards in white sands would fall under the first scenario. This distinction is still a little fuzzy to me so if anyone has any comments please feel free to let me know where I’m going wrong.

Ecological Isolation

We read two chapters this last week: Sympatric Speciation and Ecological Isolation.

Frankly, it is late in the night and I have no desire to frustrate myself talking about the former. so it'll be Ecological Isolation I discuss.

Well, what is Ecological Isolation? In my words, adapted from Coyne and Orr (citing Stebbins 1950), ecological isolation is the confinement of groups of individuals to different habitats and the selection against hybrids in either of these. So my question is, if we are saying that these groups of individuals are restricted to different habitats, but we make a point of defining habitats as different (but not mutually exclusive) from niches, then can we say that these two groups are sitting on different adaptive peaks? And, that hybrids fall somewhere in the valley between these peaks?

Habitat Isolation on the other hand refers to a similar spacial confinement of individuals to different habitats, but instead of the selection against hybrids that have already formed (post-zygotic isolation), it is unlikely that members of each parental group will breed to form these hybrids in the first place (pre-zygotic isolation). The formation of hybrids is unlikely because parental individuals are unable to live in the other habitat- gene flow is thus reduced, and speciation is possible between the parental lines.

So on to White Sands species. I feel like it is a dangerous area to get into by arguing that any of the White Sands lizards formed by some form of habitat or even ecological isolation. Yes, we can see today that White Sands and dark soils (and even the ecotone) form distinctive habitats more or less spatially separated from others. But it is difficult to know the qualities of initial selection after colonization of White Sands. Not to mention the whole 'hybrid' thing is a bit of a mess. What is a White Sands x dark soils 'hybrid' anyway? Are they selected against in either habitat? Are they found at the ecotone (i.e. is the ecotone a hybrid zone?)

If I may draw your attention to Rosenblum 2006, where on page 13 we can see the qualities of members of each species across the ecotone:

Holbrookia: ecotone lizards are phenotypically indistinguishable from White Sands individuals. Furthermore, population structure is high and there is no evidence of gene flow or changes in population size.

Sceloporus: ecotone lizards are intermediate between dark and White Sands individuals. Population structure is moderate, gene flow is evident, but population size is constant.

Aspidoscelis: ecotone lizards are indistinguishable from dark soils lizards. Population structure is weak, there is no evidence of gene flow but there are indications of increasing population size.

This might tell us something about what we may call 'hybrids' in each species. Both Holbrookia and Aspidoscelis show little or no evidence of gene flow between dark and White Sands populations; however, ecotone individuals are composed of individuals from opposite areas in either species. This suggests that populations are not presently continuous across the ecotone: hybrids between both dark and White Sands populations do not (often) form.

But what about Sceloporus? Evidence suggests that there is gene flow across the ecotone. Furthermore, individuals at the ecotone are actually intermediate phenotypically (in terms of colour at least)! But is this local adaptation to ecotone conditions? The ecotone generally has white gypsum sand, but is characterized by denser vegetation than the heart of the dunes (see above pictures). But what does this matter anyway if most Sceloporus are found basking on yucca stalks (which, consequently, are the same colour everywhere). So are these hybrids between dark soil and White Sands populations? If their population size is constant, they aren't being selected against, are they?

This summer we plan on initiating a mark recapture study. I think that it is especially important to sample individuals across the ecotone. First, because populations of each species are structured so differently here, and second, because we might be able to resolve some questions regarding selection at this crucial spot.

Thanks for tuning in this week! There was something else I was going to say about ecology, but I forget it now. Maybe it'll come to me in my sleep.

~ Simone

The "Species" of White Sands

Ecotone Sceloporus undulatus

Dear attentive blog followers. As may be obvious, I have been slacking a bit lately. As such, this blog will focus on last-last week's reading on species concepts. I think that the three species in White Sands provide interesting examples of how we might define species. I'll start by posing the question:

Could any of the three White Sands lizard populations be considered distinct species from their dark-soils counterparts? If so, which ones?

Here I'll consider each of the three White Sands lizards in the light of the major species concepts illustrated on page 27 of SPECIATION. I will focus mainly on:

1. The Biological Species Concept (BSC)

2. The Genetic/Phenotypic Cohesion Species Concept (RSC, CSC)

3. The Evolutionary Cohesion Species Concept (EcSC, EvSC)

4. The Evolutionary History/Phylogenetic Species Concept (PSC).

Now... to the species:

Sceloporus undulatus

The strongest line of evidence that Sceloporus form good species is based on the fact that local White Sands Sceloporus do not recognize 'invading' males in their territory as members of their own species. Indeed, instead of engaging in antagonistic behaviour, the local males seem to get confused... trying out courtship, exploratory and other confused displays (Robertson and Rosenblum in prep). We don't know yet what confuses these males, but this does indicate that some degree of speciation has occurred. A further confounding factor is that Sceloporus are continuously distributed across the ecotone separating dark soils and White Sands. This discourages allopatric speciation between the two habitats as the presence of a 'hybrid-zone' allows gene flow.

Thus, we generally can conclude that White Sands Sceloporus forms a pretty poor species. The presence of 'hybrid-like' individuals at the ecotone rules out the group as good species according to the Genetic cluster species concept (GCSC). It is difficult to rule out the possibility of multiple invasions of White Sands from dark soils, such that we cannotconclude that White Sands Sceloporus are monophyletic (largely ruling out the Evolutionary and phylogenetic species concepts).

Furthermore, Kayla will be looking at the further possibilities of mate choice and recognition in Sceloporus, thereby in a way, examining elements of the Recognition species concept. I personally am looking at the ecological roles of White Sands and dark soils Sceloporus... thereby determining if either makes good Ecological species.

Aspidoscelis inornata

Less work has been completed on this species, but it's useful to consider if the White Sands populations make a good species compared to the other two. At the ecotone, unlike Sceloporus, Aspidoscelis individuals are actually dark. Again, it's likely that multiple colonizations of White Sands occurred. We have not performed recognition and mate choice studies, so it is difficult to quantify the degree of behavioural isolation between dark and white lizards. As such, it seems as though Aspidoscelis do not form very good species by any criteria.

Holbrookia maculata

Of the three White Sands colonist lizards, Holbrookia seems the most likely candidate for a new species, distinct from dark soils populations. Following are two main lines of evidence in support of this statement.

a. White Sands Holbrookia males more often choose local females, and both sexes recognize local mates over non-local mates (see Rosenblum 2008). This suggests behavioural (pre-mating) isolating mechanisms in effect.

b. White Sands Holbrookia populations are disjunct from ancestral dark soils populations. That is, there is no 'hybrid' zone at the ecotone (individuals at the ecotone are indistinguishable from inter-dune individuals). This limits gene flow into the White Sands populations from dark soils populations.

But are White Sands Holbrookia monophyletic?

Were there multiple invasions from dark soils?

How does this compare to the other species?

A new member of our lab, Tyler Hether will be looking at landscape genetics and perhaps we can begin to uncover some of these questions...

Again, Kayla and I will be looking at subjects related to the recognition and ecological species concepts (respectively)... hopefully we can find out more about the nature of species formation at White Sands.

Speciation models

In chapter three Coyne and Orr focus on allopatric and parapatric models of speciation. Allopatric speciation refers to the divergence of populations that are separated by some kind of physical barrier, so that there is no gene flow between populations. There are two models of allopatric speciation- vicariant speciation and peripatric speciation. The main difference between these two models is the relative sizes of the populations involved in speciation. In the vicariant model, a population is split into two or more large populations by a barrier or by extinction of intermediate populations. In the peripatric model, a small population becomes isolated from the main population either by physical barriers or when a new area is colonized by a few individuals.

Parapatric speciation refers to the divergence of populations between which there is some gene flow. One model of parapatric speciation is the clinal model. In this model a species is distributed continuously across a variable environment. Subpopulations adapt to local habitats, but adaptation is hindered by gene flow from nearby ecologically distinct populations. When populations have become differentiated enough, reinforcement may play a role in generating reproductive isolation. In the stepping stone model of parapatric speciation, there are discrete populations with restricted gene exchange. Since populations are discrete it is easier for selection or drift to result in reproductive isolation than in the clinal model.

Though white sands and dark soils lizards have not speciated, we can still think of which of these models is most applicable to divergence of these populations. The first question we should answer is how much gene flow is occuring between white sands and dark soils lizards. Previous research has shown that there is some gene flow, though the amount differs between the different species of lizards (Rosenblum 2006). This means that the parapatric models are more applicable than the allopatric models. Next question: are populations distributed continuously over a variable environment, or are there discrete populations? While the environment is definitely variable from white sands to dark soils, the amount of connectivity between populations is also different between different species of lizards. For example, Holbrookia maculata show the most genetic differentiation between populations, and also the largest color difference between white sands and dark soils populations. Sceloporus undulatus shows an intermediate level of phenotypic and genetic difference between populations, and Aspidoscelis inornata shows the least of both (Rosenblum 2006).

This evidence shows that Holbrookia populations are probably the most discrete, with Sceloporus populations occurring in a less discrete distribution, and Aspidoscelis distributed almost continuously. So maybe the stepping stone model is most likely for Holbrookia, clinal for Aspidoscelis, and some combination for Sceloporus? And if the models are different, does this mean that different types of reproductive isolation could be important for different species? Reinforcement is thought to be important for the clinal model, but in the stepping stone model premating isolation might evolve more easily as a byproduct of selection or drift. Hmmmmmm…

Barriers leading to reproductive isolation

Last week we read the first two chapters of Coyne and Orr’s book Speciation. Chapter 1 of Speciation spends quite a bit of time discussing and classifying reproductive isolating barriers. Though I have already focused on reproductive isolation in past blogs based on information given in The Ecology of Adaptive Radiation, I will spend some time now giving a more comprehensive overview of possible barriers leading to reproductive isolation. Coyne and Orr define pre- and postmating isolation, as well as the barriers that cause them, as follows:

1. Premating isolation barriers- barriers that impede gene flow before transfer of gametes to members of other species.
a. Behavioral isolation: Members of different species do not initiate courtship or copulation with each other due to a lack of “cross-attraction”.
b. Ecological isolation: Barriers to reproduction are direct byproducts of adaptation to the local environment. This category includes isolation because of differences in habitat preference, timing of breeding, and pollinator interactions.
c. Mechanical isolation: Members of different species do not reproduce because they have incompatible reproductive structures.
d. Mating system isolation: This barrier refers to the evolution of partial or complete self fertilization or the asexual production of offspring, which can result in the formation of new species.

2. Postmating, prezygotic isolation barriers- barriers that act after transfer of gametes but before fertilization takes place.
a. Copulatory behavioral isolation: The behaviors of individuals during copulation do not allow fertilization to occur.
b. Gametic isolation: The transferred gametes cannot effectively cause fertilization.

3. Postzygotic isolation barriers- barriers resulting in hybrid sterility and inviability.
a. Extrinsic barriers: Sterility and inviability are results of the biotic or abiotic environment. This category includes ecological inviability, where hybrids have a lower fitness because of a failure to find an appropriate ecological niche. Behavioral sterility, where hybrids cannot obtain mates, is also included.
b. Intrinsic barriers: Sterility and inviability are results of developmental problems. Hybrid inviability is a result of developmental problems causing full or partial lethality. Hybrid sterility may be physiological or behavioral; physiological sterility leads to problems in the development of gametes or reproductive organs, while behavioral leads to developmental problems causing hybrids to be incapable of successful courtship.

Which of these potential barriers could be acting currently among populations of lizards in white sands and dark soils? Of the premating isolation category, behavioral and ecological isolation both seem possible. In fact, it seems like a combination of these two types would be most likely. As discussed previously, behavioral isolation could be a result of either a preference for local mates that are phenotypically similar, or a lack of recognition of mates from populations adapted to living in different environments. Ecological isolation is certainly important in terms of habitat preference; if white lizards and dark lizards don’t encounter each other because they live in different areas, this will result in reproductive isolation. I think ecological isolation would be important over most of the habitat, while behavioral isolation may be important at the ecotone.

In terms of postzygotic isolation, extrinsic barriers seem most relevant to the white sands system. Depending on the genetic basis of dorsal coloration (which is different among different lizard species in white sands) hybrids could suffer from ecological inviability. The mutation that causes blanched coloration is dominant in one white sands species and recessive in another. Either way, if a hybrid finds itself not well matched to its substrate and can’t effectively choose a better niche, it won’t do as well (it will probably end up as food for someone else).

From reading about all these different isolation barriers I have come to the realization that as I continue on with my plans to investigate behavioral premating reproductive isolation in these lizards, I can’t overlook the possibility that other reproductive barriers could be playing an important role as well, and I need to be aware that many factors have likely shaped the system as it is today.

Genetic lines of least resistance, continued

Good explanation of the genetic lines of least resistance, Simone.

I just wanted to add a few comments on the subject. Schluter mentions that when traits covary (in the ways Simone described earlier) the amount that one trait covaries with another will affect how far evolution will be pulled away from the direction in which selection is at its strongest. Take the finch beak example that Simone gave, and imagine that selection is strongest in the direction favoring an extremely long, narrow beak. If beak width and length depend on each other to a great extent, so that selection on one automatically changes the other dramatically (selection for a longer beak leads to a wider beak too), then this interaction would lead to evolution in a direction that is greatly skewed off of the path of strongest directional selection. If the traits did not affect each other as much, the direction of evolution would deviate from the direction of strongest selection to a smaller extent.

Schluter mentions that over time this effect of genetic constraint can be overcome and a population may make it to the top of an adaptive peak. But the stronger the covariance between the traits is, the stronger the skew will be away from the direction of strongest selection, and the longer it will take to make it up the peak.

One cool result of this whole covariance thing is that the degree of covariance between traits can help you more accurately predict the direction of evolution. If you realize that selection is strongest in a certain direction on an adaptive landscape, you can examine phenotypic traits that would be affected by selection, and based on how these traits covary you could make predictions about the direction of evolution along the landscape. Neat!

Now that’s really it for the Ecology of Adaptive Radiation…Speciation here we come!