Speciation? II

Response to Section 5.1.2, "Animals"


Summary: Animal hybrids are "rare organisms" because hybridization is not a viable mechanism for animal diversification. Animal hybrids are typically unisexual, where part of the genome is not heritable. For example, in female fish hybrid clones, the parent species is always required to provide the male portion of the genome, meaning a truly new independent species is not formed. Darwinian evolution requires heredity, but this does not entail the origin of anything new that is heritable. Asexual animal hybrids are thus often called "evolutionary dead ends" where they do not produce new diversity and instead "[e]xtant asexual ‘species’ are little more than scattered twigs at the tips of major phylogenetic branches." Moreover, hybrid asexual females are so similar to the parent species that "sexual males of the progenitor species are unable to distinguish hybrid females from those of their own species," implying little morphological change arises in this process. This mechanism is largely "irrelevant" to sexually reproducing animals.



The FAQ suggests animals may also speciate through hybridization, though one primary paper cited by the FAQ (Vrijenhoek 1994) admits that animal hybrids are "rare organisms."
20 Gregory and Mable (2005) concur, observing that "recent polyploidy is far less common in animals than in plants."21 Indeed, there are good reasons why animal hybrids are much rarer than plant hybrids: the problems encountered with animal hybridization appear far more severe than those raised by plant hybridization.

Darwinian evolution can only operate when there is variation, selection, and heredity. Vrijenhoek (1994) discusses hybridogenesis within fish but finds that the hybrid genome is not entirely heritable since, "Paternal B genes are expressed in the hybrids but are not heritable. Only the A’ ‘hemiclonal’ genome is transmitted between generations." This is why hybrid fish like
Poecilia formosa are often called "clones," because the heritable portion of the genome is simply cloned from generation to generation, and males from the parent "species" are always required to maintain the line. Thus, this "species" cannot be maintained without the constant presence of the parent line, meaning the hybrids are not a truly independent species. Since hybridization in vertebrates typically involves asexual clonal reproduction, this poses a problem for those who would cite hybridization as a mechanism of animal evolution.22 As the Monterey Bay Aquarium Research Institute website explains regarding hybrid fish:

"In hybridogenesis, the female mates with a male, forming a female offspring with both the maternal and paternal genomes. When that female offspring produces eggs however, the male genome is discarded."
23

Again, in other words, these hybrids cannot persist without both species present. These hybrids always require the parent species to be present in order for them to originate and persist. The review by Vrijenhoek (1994) explains that such animal hybrids are typically "evolutionary dead ends": 15


"Asexual species are often considered evolutionary dead ends because of their presumed genetic inflexibility. Among vertebrates and insects. only 0.1% to 0.2% of species are strictly asexual. This rarity suggests a ‘mutation/selection-like’ balance. New asexual lineages arise infrequently and go extinct rapidly.
Extant asexual ‘species’ are little more than scattered twigs at the tips of major phylogenetic branches. Except for bdelloid rotifers. asexual lineages have not speciated and diversified into rich asexual clades."24

If anything, animal hybridization would seem to entail collapse, loss, and decrease of pre-existing diversity rather than the generation of new diversity. Indeed, in this case these asexual animal hybrids are not genetically viable in the long term:

"Genetic decay provides a final challenge to the persistence of clones. Muller suggested that mutations would accumulate like a ‘ratchet mechanism" in asexual lineages. Recombination in sexual lineages produces offspring with higher and lower mutational loads than the parents and purifying selection effectively maintains a low load. An asexual population cannot reduce its load below that of the ‘least loaded’ clone. If by chance that clone is lost, the load has increased one step. Excluding back mutations. it cannot be reduced."
25

It’s not clear that this can be a viable long-term mechanism for evolution in vertebrates because this new "species" could never exist on its own in the wild without the parent species constantly available to continually generate it. In the long term, animal hybridization appears to be a byproduct of existing species, not a mechanism for producing new species.

Little more is said in the FAQ regarding hybridization in animals, but it is clear that hybrids result from crossing highly similar species within the same genus, and nothing new is created that is heritable. In fact, regarding
Poecilia formosa, Gregory and Mable (2005) observe that "sexual males of the progenitor species are unable to distinguish hybrid females from those of their own species,"26 implying that little morphological change has occurred. This is not a viable mechanism for evolution in animals, as Dobzhansky recognizes: "Sudden emergence of new species by allopolyploidy is … irrelevant to Drosophila and most bisexual animals."27

Response to Section 5.2 "Speciations in Plant Species not Involving Hybridization or Polyploidy"

Response to Section 5.2.2, "Maize (Zea mays)"


Summary: Corn breeders produced "almost complete reproductive isolation" between two "races" or "varieties" within the same species but a new species was not claimed to have emerged. The partial reproductive isolation results from a premating mechanism—changes in flowering timing—not a large-scale change which might produce fundamentally new types of organisms. The isolation was produced via strong artificial selection; whether this could occur in the wild is not established. 16


According to the paper cited by the FAQ (Paterniani, 1969), humans have bred the corn species
Zea mays into many races over the past 4000 years, and "all races cross readily one with the other giving progenies with normal fertility."28 Through selective breeding that destroyed hybrid crosses, this study sought to achieve reproductive isolation between two species of corn. However, even at the end of the experiment some intercrossing between the varieties still occurred. As Jonathan Wells writes, "Paterniani noted ‘an almost complete reproductive isolation between two maize populations’ but did not claim that a new species had been produced."29

So what exactly generated the reproductive isolation? It turns out that by destroying hybrids, the experimenters were selecting for individuals that flowered at different times and did not produce hybrids. Reproductive isolation was probably achieved by little more than slight changes (a few days) in the timing of flowering in the two varieties:

"Data on days to flowering show that some change occurred, thus indicating that this mechanism is playing a role in the isolation obtained. Both original populations flowered in the same days with identical averages for days-to-tassel and days-to-ear flowering. Cycle IV of the two populations, already showing a great degree of reproductive isolation, have a marked difference in days to flowering. The white flint maize became about 5 days earlier, for tassel and ear, than the corresponding original population. The yellow sweet maize changed less, but in the opposite way; it became, on the average, 2 days later. As a result, the two cycle-IV populations have about 1 week of difference in flowering time. This difference is of sufficient magnitude to explain most of the reproductive isolation obtained."
30

The paper concludes: "The data show that the number of days from planting to flowering was probably the main factor."
31 Thus, what the paper shows is that the plants are flowering at different times and thus don’t have the opportunity to pollinate one-another.32 Other than that, there’s no indication of biological changes. This does not amount to the kind of large-scale change which can produce fundamentally new types of organisms. And it was done by controlled artificial selection; whether this could be achieved in the wild is not established.

Response to Section 5.2.3, "Speciation as a Result of Selection for Tolerance to a Toxin: Yellow Monkey Flower (Mimulus guttatus)"


Summary: Two populations within the same species of a flowering plant "developed partial postmating isolation between some races" where "total postzygotic reproductive isolation between two populations, in the sense that inviable zygotes are formed, can be produced by a comparatively simple genetic mechanism." The cause of the reproductive isolation is thought have "a simple genetic basis" entailing changes in "a single gene." Large-scale differences have not evolved and speciation is only claimed to have been "initiated," not complete.



According to the paper cited by the FAQ (Macnair and Christie, 1983), the yellow monkey flower,
Mimulus guttatus, has "developed partial postmating isolation between some races."33 While offspring can be produced between the two populations, they are inviable, as the paper 17

later reports "total postzygotic reproductive isolation between two populations, in the sense that inviable zygotes are formed, can be produced by a comparatively simple genetic mechanism."
34 The investigators do not fully understand the genetic cause of reproductive isolation but believe it may be linked to pleiotropic effects caused by the gene responsible for copper tolerance/intolerance. They suggest that the isolation has "a simple genetic basis" and is caused by changes to "a single gene."

On the one hand this shows that reproductive isolation may be achieved. On the other hand the fact that offspring can be produced shows large-scale biological differences have not evolved. In fact, the paper suggests that speciation was only "initiated" and not complete.

In this case, we have seen two races within the same plant species are essentially identical apart from one being tolerant to copper and other intolerant. The two races are so similar that they can produce offspring, but those offspring are not viable. Speciation is not complete and significant biological change is not observed. The exact genetic mechanisms which are causing such reproductive isolation are unknown, but they might result from "a simple genetic basis" entailing changes in "a single gene." This does not show significant biological change.


Response to Section 5.3, "The Fruit Fly Literature"

Response to Section 5.3.1, "Drosophila paulistorum"


Summary: This study showed that if you start with "semispecies" within a fruit fly species which are "indistinguishable morphologically," and then subject the strains to artificial breeding experiments, "in none has anything like complete isolation been achieved." Moreover, there is no suggestion that the populations were no longer "indistinguishable morphologically" after the experiments. At best, only a "new race or incipient species" was created. Some authorities have challenged even the partial isolation, claiming the results "may have been due to contamination of cultures by other subspecies."



In this example, the FAQ discusses whether reproductive isolation has been achieved between various strains, or "semispecies," of the fruit fly
Drosophila paulistorum. The paper cited by the FAQ (Dobzhansky and Pavlovsky , 1971) states that in the wild, "The semispecies are indistinguishable morphologically,"35 and "not different enough" to be considered "fully differentiated species." They observed that after a certain amount of time breeding in captivity, crosses between two particular strains only produced sterile males. (Female hybrid crosses were apparently still fertile.) The investigators claim that contamination was "ruled out" but Jonathan Wells notes that those claims may be incorrect, for "Coyne and Orr wrote in 2004, however, that Dobzhansky and Pavlovsky’s result ‘may have been due to contamination of cultures by other subspecies.’"36

After conducting artificial breeding experiments where hybrid crosses between the strains were destroyed, the authors produced some assortative mating. (This process does not mimic natural conditions.) The authors admit that the proportion of hybrids merely "decreased" and "in none has anything like complete isolation been achieved."
37 Jonathan Wells thus notes that in this example: "Dobzhansky and Pavlovsky reported only a ‘new race or incipient species,’ not a new 18

species."
38 Another paper in the FAQ (Halliburton and Gall, 1981) lists this study among various studies where "none has succeeded in establishing complete sexual isolation."39 In any case, speciation is not claimed to have occurred, and there is no suggestion whatsoever that the semispecies are no longer "indistinguishable morphologically." In fact, after reviewing this example, Dobzhansky concludes that sometimes "reproductive isolation and speciation precede differential adaptedness,"40 suggesting the populations had not diverged.

Response to Section 5.3.2, "Disruptive Selection on Drosophila melanogaster"


Summary: Artificial selection for the number of chaetae (hairs) on a population of fruit flies produced partial, though not complete, reproductive isolation. The extent of morphological variation is small-scale changes in the number of chaetae. Later attempts to reproduce these results were unsuccessful. The authors explicitly state that natural speciation has not been demonstrated. This experiment does not show complete reproductive isolation, speciation, or significant morphological change.



In this example, the paper cited by the FAQ (Thoday and Gibson, 1962) reports experiments with a small population of wild fruit flies (
Drosophila melanogaster), which over a series of generations selected those with both the highest and lowest numbers of chaetae, or hairs. Through artificial selection over the course of successive generations, they were able to select for flies with more, or less chaetae than in the original population. This is the extent of the variation bred by this experiment.

Even after multiple generations, the flies with high or low numbers of chaetae could still hybridize. However, hybrids did become less common as the experiment progressed, although the reason for this was not known. The experimenters proposed that perhaps it "arises from mating preferences or from an inability of hybrid flies to compete as larvae."
41 The ability for the high and low chaetae populations to hybridize, although rare, was not impossible. Reproductive isolation was thus not complete. Additionally, since artificial (rather than natural) selection was used to create and maintain the two populations, the authors warn:

"We do not, however, wish it to be thought that we regard this as a demonstration that sympatric speciation occurs in Nature, for such a conclusion cannot be drawn from the results of laboratory selection experiments."
42

At best, this experiment shows there is pre-existing variation among fruit flies for numbers of chaetae, and artificial selection for this trait in a non-natural laboratory setting can result in partial reproductive isolation. But changes in the numbers of hairs was all that was produced: it does not show anything close to large-scale evolutionary change.

Thoday and Gibson (1962) open their paper by admitting that "the key demonstration that a single wildtype population can be converted by selection into two populations that are mutually isolated in the conditions in which they have to maintain themselves has not hitherto been made."
43 But Jonathan Wells observes, "Not only did Thoday and Gibson not claim to have produced a new species, but also other laboratories were unable to replicate their results."44 The 19

FAQ likewise admits, "In the decade or so following this, eighteen labs attempted unsuccessfully to reproduce these results." The summary of Halliburton and Gall (1981) is striking:

"Several attempts to repeat these results have failed (e.g., Scharloo
et al., 1967a; Chabora, 1968; Barker and Cummins, 1969). Similar experiments, selecting for other quantitative characters or using other organisms, have usually failed to produce assortative mating (Scharloo, 1964; Robertson, 1966; Scharloo et al., 1967b; Grant and Mettler, 1969; Bos and Scharloo, 1973), but a few have succeeded (Coyne and Grant, 1972; Soans et al., 1974)."45

As far as this example goes, it would seem this "key demonstration" of speciation has not been made.


Response to Section 5.3.3, "Selection on Courtship Behavior in Drosophila melanogaster"


Summary: This experiment sought to induce changes in the mating preferences of two strains of fruit flies. Only "partial" reproductive isolation was achieved, and the extent of change observed was small changes in courtship initiation behaviors (e.g., licking and vibrations). The two strains were "similar" before the experiments, and apart from slight changes in mating behaviors, remained very similar after the experiments.



This experiment took two pre-existing strains of fruit flies from within the same species—
Drosophila melanogaster—and sought to determine whether changes in mating preferences could be induced. This included artificially killing hybrids between the strains (a process that does not necessarily mimic nature). Incomplete reproductive isolation was observed, which one paper cited by the FAQ (Knight et al. 1956) called only "[p]artial sexual isolation."46 Another paper in the FAQ (Halliburton and Gall, 1981) lists this study among various studies where "none has succeeded in establishing complete sexual isolation."47 The most biological change that this example documented was small-scale behavioral differences pertaining to courtship, specifically changes in the amount of "licking" that males do to females to initiate mating. One paper cited by the FAQ (Crossley, 1974) showed just how unimpressive the sort of change observed in this experiment was:

"Quantitative analysis of male and female behavior revealed the underlying causes of changed mating preferences and faster mating. In the LS experiment male courtship became more stimulating because percentage licking of both males and percentage licking plus vibration of males increased."
48

Thus, all that was observed were changes in the courtship initiation behaviors (licking and vibrations) between the strains. This is small-scale change. The two strains were "similar" before the experiments, and apart from slight changes in mating behaviors, remained very similar after the experiments.

These experiments were conducted in the laboratory, but Crossley (1974) observed why laboratory experiments do not match natural conditions: "One difficulty in relating these results to selection against hybrids in nature is that in the laboratory, selection against hybrids was total 20


but in the wild some hybrids would survive to breed in spite of their disadvantages compared with pure bred offspring."
49

Response to Section 5.3.4, "Sexual Isolation as a Byproduct of Adaptation to Environmental Conditions in Drosophila melanogaster"


Summary: This fruit fly study found partial reproductive isolation after selection experiments on fruit flies. No significant morphological change was reported, and any reproductive isolation which did exist stemmed from premating factors. This paper thus serves as a good example of how speciation need not entail significant morphological or genetic change.



In this experiment, the investigators changed temperature and humidity conditions for populations of fruit flies (
Drosophila melanogaster) in the laboratory. Originally the populations came from the same genetic stock, but after about 5 years of artificially exposing them to different environmental conditions, the experiment found that some reproductive isolation was established. According to the paper cited by the FAQ (Kilias et al., 1980), "The highest isolation index detected was 0.388 ± 0.108,"50 meaning reproductive isolation was far from complete.

But was any significant morphological change found? None was reported, and in fact Kilias
et al. (1980) said the reproductive isolation resulted simply from pre-mating factors:

"Since, females from either population mate equally, the reproductive isolation detected in the present investigation seems to be due to changes in behavior (different mating preferences or discrimination) of our populations. … In the present study we failed to detect significant postmating isolation."
51

The change that was observed pertained to the sexual isolation and changes in ovipositional rhythms, and thus they observed that "speciation" (meaning mere reproductive isolation) "may occur with relatively little genetic change in structural genes."
52 This led them to the final conclusion that genetic divergence between populations often occurs after "speciation" (again, mere reproductive isolation) because such divergence was not observed here: "The genetic distance observed between species probably results from post-speciational divergence."53 This paper thus serves as a good example of how speciation need not entail significant morphological or genetic change.

Response to Section 5.3.5, "Sympatric Speciation in Drosophila melanogaster"


Summary: After two populations of fruit flies were selected for various food-finding behaviors, incomplete reproductive isolation was observed. The populations could still produce "fertile offspring" and speciation was only claimed to be "incipient. No significant morphological change arose.



This is another study where partial reproductive isolation was established between populations of
Drosophila melanogaster, and the differences between the populations were minor, and of ambiguous importance. 21

The experimenters forced fruit fly pupae to navigate a maze to find food where they might choose to go towards the light/dark, up/down, or choose between two different scents. Flies which made opposite choices were then separated, allowed to breed, and then subsequent generations were selected according to which flies made the same choices. When the fly populations were then allowed to mix, partial reproductive isolation was achieved.

Complete reproductive isolation was not found. As one of the papers cite by the FAQ (Rice and Salt, 1988) reports:

"It might be argued that incipient speciation has not occurred in this experiment for two reasons. First, trace amounts of gene flow occurred between the population using habitats SE and 4L, since a small fraction of flies switched habitats between generations. … Second, forced matings between the two populations produced fertile offspring in the F1 and F2, and thus reproductive isolation was mediated only by habitat-preference behavior. … [I]rreversible reproductive isolation did not occur in this experiment."
54

The fact that gene flow between the populations occurred, and that they could be forced to produce fertile offspring, shows how similar the strains remained. They claim that any speciation is merely "incipient."

But what evolved? At most they might have selected for preferences for when seeking food. As they conclude: "In these experiments, the only barriers to gene flow were gaps that gradually developed in the distribution of spatiotemporal habitat preference."
55 It’s important to realize that again, we have seen no significant morphological change. As Jonathan Wells observes, "Within thirty generations the flies had sorted themselves into two populations that did not interbreed, but Rice and Salt claimed only ‘incipient speciation that we believe to have occurred.’"56

Response to Section 5.3.6, "Isolation Produced as an Incidental Effect of Selection on several Drosophila species"


Summary: Three fruit fly studies were reported: they showed "slight" or "incipient" or "not complete" sexual reproductive isolation, but none showed complete reproductive isolation or speciation. None showed significant morphological change.



In this example, the FAQ discusses a paper (del Solar, 1966) which reported experiments that artificially selected "positively and negatively geotactic" and "positively and negatively phototactic" strains of both
Drosophila melanogaster and Drosophila pseudoobscura.57 The paper reports that this produced what was called "slight" sexual isolation, or "incipient reproductive isolation," due to "changes in sexual behavior."58 The paper thus reports that complete reproductive isolation was not found:

"Whether selection for geotaxis and phototaxis always and necessarily produces a change in the sexual behavior, and whether continued selection may carry the sexual divergence
anywhere near complete isolation, can only be decided by further experiments."59 22

Not only was "anywhere near complete isolation" not achieved, but significant biological change was also not achieved. As the paper reports: "The geotactically and phototactically positive and negative strains appear to be
indistinguishable in external morphology."60

Another example discussed by the FAQ in this section pertains to Dodd (1989) which reported experiments on populations of the fruit fly
Drosophila pseudoobscura. Four populations were given a starch-based medium, and the other four were given a maltose-based medium. The paper reported that individuals raised on the starch medium preferred to mate with other starch-fed fruit flies; likewise flies fed maltose preferred to mate with other maltose-fed individuals. Interestingly, these traits arose independently in each of the four populations in each medium.

Because the experiments controlled for food source rather than mating behavior, they concluded that the sexual isolation was "a pleiotropic by-product of the adaption of the populations to the two media" but "The mechanism of the isolation in this system is as yet unknown."
61 Indeed, it was not that the two populations were incapable of interbreeding or never interbred, it was just that they did so less than would be expected under normal random mating. Another paper cited by the FAQ (Schluter and Nagel, 1995) described the findings by stating that only "some premating isolation evolved," and "Reproductive isolation between divergent lines was not complete."62 Speciation was also not said to have occurred.

Aside from mating and food preferences, there were no claims of biological change between the populations. Again, we see that not only has reproductive isolation not been demonstrated, but significant biological change did not evolve.

Despite the aforementioned underwhelming results, the FAQ then discusses another paper which it says reported "Less dramatic results." According to the paper cited by the FAQ (de Oliveira and Cordeiro, 1980), different populations of
Drosophila willistoni were given food at different pH levels. Like the other studies in this section, some individuals preferred to mate with other individuals fed food at the same pH. But the different populations were still capable of interbreeding. When the offspring were fed alkaline food, "their hybrids are not less fit."63 However, the paper reported that "on the acid substrate the hybrids are inferior to their parents which are adapted to this food."64 The paper thus only claimed to find "incipient isolation,"65 not complete reproductive isolation.

As for the degree of morphological change, aside from a preference for a certain pH level in food, no significant biological change was reported. In fact, the paper notes that among three long-standing natural races of
D. willistoni, "These flies are morphologically indistinguishable." This study certainly did not change that observation: Once again we have not seen complete reproductive isolation, nor have we seen significant biological change.

Response to Section 5.3.7, "Selection for Reinforcement in Drosophila melanogaster"


Summary: Again, fruit fly experiments found only "partial" reproductive isolation and did not report significant biological change. One paper boasted that "[t]he evidence here presented shows … that natural selection can act to strengthen isolation." But since the ‘destroy the hybrid’ experiments simulated processes that would never occur in nature—23


the artificial destruction of all hybrid flies for no biological reason other than experimental curiosity—it obviously confused natural selection with artificial selection.



In this section, the FAQ acknowledges that Rice and Hostert (1993) do not find evidence for the reinforcement model of speciation, where, according to the paper, "the physical barrier breaks down before complete reproductive isolation has evolved in allopatry" but yet "matings between previously separated subpopulations are presumed to produce low-fitness hybrid offspring, and this selects for positive assortative mating."
66 If hybrids cannot survive and reproduce, then arguably the two populations are reproductively isolated.

The FAQ then discusses two older papers which purportedly support the reinforcement model. Keep in mind again that the most important question here is not whether two populations can fail to produce viable hybrids, but whether the two populations show some non-trivial degree of evolutionary change.

The first study cited by the FAQ (Ehrman, 1971) took two strains of
Drosophila melanogaster and sought to test for sexual isolation. The paper never claims that significant morphological change evolved, but it does note that this experiment had results similar to Knight et al. 1956, a paper which only found "[p]artial sexual isolation," and never claimed significant biological change arose. Likewise, Ehrman (1971) reports that after breeding experiments, in males there is only "some sexual isolation," and the author thus hopes that "the degree of reproductive isolation evolved to be enhanced by the passage of time."67 The paper cites no significant changes in the fly populations after the experiments. The experiment did not find cross-mating is impossible, placing limits on the degree of change which arose. Not only was complete reproductive isolation not achieved, but there is no report whatsoever that significant biological change emerged. Another paper in the FAQ (Halliburton and Gall, 1981) lists this study among various studies where "none has succeeded in establishing complete sexual isolation."68

In a similar study, another paper cited by the FAQ (Koopman, 1950) mixed two similar species of
Drosophila, D. pseudoobscura and D. persimilis in an attempt to induce reproductive isolation in the lab. Complete reproductive isolation was not established.

Normally one would think that if the two populations are already classified as members of different species then perhaps they are already completely reproductively isolated; but this is not the case here, as the two species are "closely related" and in fact
D. persimilis was "formerly known as D. pseudoobscura, race B."69 Complete reproductive isolation between the two species does not exist in the lab: the two groups can form hybrids, and "Hybrids seem to have the same viability as the pure species," although hybrid males are sterile and females, when backcrossed with parent species, tend to have eggs with "poor viability."70 While they form hybrids in the lab, however, "in nature, not a single hybrid has been found, even from localities where both species occur together."71

To help investigate whether complete reproductive isolation could emerge, the experimenters used a tactic that would not be present in nature: they artificially killed any hybrids. As the paper stated: 24


"The experiments herein described were made in order to determine whether, in artificial populations consisting of the two closely related species … an increase in the reproductive isolating mechanisms could be detected if in each generation the hybrids between the two species were systematically eliminated."
72

The paper concluded that "[t]he evidence here presented shows … that natural selection can act to strengthen isolation between species."
73 But was it "natural selection" or artificial selection? The paper shows astonishment at how "in a surprisingly short time," increased reproductive isolation was established. But they really should not be so surprised since they recognize that "This change, of course, was aided by the practice of removing the hybrids entirely each generation, in this way simulating complete hybrid inviability."74 Thus, natural selection was not at work; rather artificial selection caused these changes.

In any case, complete reproductive isolation did not arise as hybrids still formed, albeit at a "low level." So the experiments started off with two similar populations of flies that have partial reproductive isolation, and they ended with two highly similar populations of flies that have "partial"
75 (though slightly more) reproductive isolation. The paper makes no report of significant morphological differences between the populations at the beginning of the experiment, and the end, so this experiment once again shows that (1) complete reproductive isolation was not achieved, and (2) significant biological change did not arise.

Response to Section 5.3.8, "Tests of the Founder-flush Speciation Hypothesis Using Drosophila"


Summary: Three papers testing the founder-flush model of speciation using fruit flies failed to produce complete reproductive isolation. Reproductive isolation was called "partial" and / or "weak," and no significant morphological change was reported.



The first paper cited by the FAQ in this section (Powell, 1978) investigated a hypothetical "founder-flush" mechanism of speciation where a small number of individuals found a new population, which then goes through various cycles of expansion in population size ("flush"), followed by a "crash," where "[a]t each crash the bottleneck population is small and genetic drift is strong."
76 After the crash, another small group of individuals found the new population, and the cycle repeats. Powell (1978) attempted to simulate this process for strains of fruit flies within a population of Drosophila pseudoobscura.

The paper reported that "neither isolation nor inbreeding by themselves lead to reproductive isolation. Only among populations which were inbred (four founder events) and allowed to flush did reproductive isolation evolve."
77 When reproductive isolation did evolve, it was called "partial"; at one point the author claimed only "some degree of reproductive isolation."78 He claimed to only have observed the "first stages of speciation,"79 not complete speciation.

What is also significant is the type of reproductive isolation that evolved. Powell (1978) reports that "no post-mating factors were detected,"
80 indicating that when crossbreeding between the populations did occur in the experiment, it produced viable and fertile offspring. This implies that significant biological change between the populations did not arise over the course of the 25

experiments. The only type of reproductive isolation that was observed was "pre-mating (ethological) isolation,"
81 where behavioral factors reduce cross-mating.

Finally, the paper observes the extreme degree of repeated flushes and crashes simulated in the experiment would require "rather special circumstances" not necessarily common in nature. Either way, once again neither complete reproductive isolation nor significant biological change was observed.

Another paper cited by the FAQ (Dodd and Powell, 1985) repeated this type of experiment and found very similar results to Powell (1978). Reproductive isolation was called "significant" but was far from complete. The overall finding was that: "in general, it appears that some weak ethological isolation exists."
82

Also like Powell (1978), all reproductive isolation that did arise was due to behavioral (premating) mechanisms and "no post-mating isolation could be detected."
83 This implies that significant morphological or genetic change did not arise between the separated populations over the course of the experiment, because fertile and viable offspring could be produced. The paper thus notes that these results counter the common evolutionary presumption that speciation occurs because populations diverge biologically:

"Many scenarios for the formation of new species envision post-mating isolation factors to evolve before pre-mating isolation evolves (e.g., Dobzhansky, 1940). Pre-mating barriers are thought to be secondary, reinforcing mechanisms of isolation. Here they appear to be primary; that is, they have evolved in the apparent absence of post-mating isolation."
84

Like the others, this experiment did not report that significant biological change evolved.

A final paper cited in this section of the FAQ (Ringo
et al., 1985) used populations of Drosophila simulans test the founder-flush model (which fosters genetic drift) against the classical model of speciation, where certain traits undergo selection, gradually leading to a new species. The lines experiencing selection were artificially selected for various "arbitrary traits."85

While some "partial reproductive isolation" in the various lines did arise, they found it was "much weaker than that typically found between sibling species of
Drosophila."86 They thus hope that:

"More-complete reproductive isolation might be established by the same forces at work over a longer time span, perhaps reinforced by direct selection for premating barriers to gene flow."
87

They thus lament that "A large gap lies between the degree of isolation between any experimental populations and the degree of isolation observed between species."
88

As far as premating isolation goes, the paper reports that "Weak sexual isolation was observed between BASE and the drift lines, for the experiment as a whole."
89 However, the paper did find 26

some postmating isolation. Interestingly, "reproductive isolation was stronger in drift lines than in selection lines" and "postmating isolation increased over time in drift lines but not in selection lines" and overall "there was only a 5% reduction in hybrid fitness." This implies that even artificial selection for certain traits did not produce sufficient biological change to prevent viable and fertile hybrids from forming between the selected lines and the original base population. Significant biological change is not reported.

Even the FAQ admits regarding this study that "only weak isolation was found and that there was little difference between the effects of natural selection and the effects of genetic drift." In other words, even when certain traits are selected for in the laboratory, only weak isolation arises viable and fertile offspring between crossed lines can still be produced. Once again, incomplete reproductive isolation (which the FAQ admits is "weak") and only very limited small-scale change observed, even when there is artificial selection for many traits.

Finally, it’s worth noting that Ringo
et al. (1985) and Powell (1978) "have been criticized (Charlesworth et al. 1982) because the base populations were derived from geographically diverse stocks, ‘so it is not clear whether their results are representative of what might happen in a natural population.’"90 Other critics (Meffert and Bryant, 1991) observed that in Ringo et al.’s experiment, "Because the populations contributing to the base population exhibited differences in mating activity that were presumably genetic, the experimental protocol created an artificially high genetic variance for traits affecting mating behavior and exaggerated the divergence among experimental lines."91

The response from the authors to such criticisms is that they wanted to study "the extent and the speed of establishment of reproductive isolation under optimal conditions; that is, we hoped to maximize reproductive isolation among lines by maximizing genetic variation"
92 in the initial population. They did this because when more realistic natural conditions are modeled in experiments, the results might have "[n]egative."93 In other words, even when they gave speciation mechanisms their best shot—better than would likely exist in nature—complete reproductive isolation was not achieved and significant biological change was not observed.

In any case, Meffert and Bryant (1991) observe that due to weaknesses in Ringo
et al. 1985:

"Hence the critical issue in the founder-flush theory of speciation has not been addressed: can bottlenecks in a natural population cause permanent alteration of courtship behavior in founder lines that would lead to premating isolation."
94

Would Meffert and Bryant thus argue that the papers cited in this section of the FAQ don’t even establish what the FAQ claims they do?