Evolutionary Biology

8 posts / 0 new
Last post
ronald bertram's picture
Evolutionary Biology

There was a thread on the Evolution Myth. I was on a website that brought up the issue of eye evolution. It had been thought that the invertebrate eye and vertebrate eye had evolved independently. I found this 2004 publication that reveals the invertebrate and vertebrate eyes may be linear evolution.


Subscription Note: 

Choosing to subscribe to this topic will automatically register you for email notifications for comments and updates on this thread.

Email notifications will be sent out daily by default unless specified otherwise on your account which you can edit by going to your userpage here and clicking on the subscriptions tab.

David Killens's picture
Interesting stuff. Because

Interesting stuff. Because the eye structure on different species is so mechanically different, I was in the camp of" at least two separate paths". But this new evidence just adds to knowledge and understanding.

Science is so cool.

ronald bertram's picture
@David Killens

@David Killens

I enjoyed Vertebrate Comparative Anatomy where you looked at the phylogeny from an organ-system perspective. I mentioned Ernst Mayr as the father of Speciation. Another great contemporary of Mayr was Alfred S. Romer. His text, The Vertebrate Body, was used in Comparative Anatomy.


Attach Image/Video?: 

Lion IRC's picture
Yes, I agree. Science is cool

Yes, I agree. Science is cool.

Old man shouts at clouds's picture
@ OP

@ OP

Bloody interesting that, thanks.

Cognostic's picture
Black Raven: Vertebrate,

Black Raven: Vertebrate, invertebrate,??? We are still miles away from the evolutionary development of the eye as even the initial photoreceptor cells may have evolved more than once from molecularity similar chemoreceptor cells.

Not saying it won't be interesting if they can one day pin down the development of the eye. It may also be interesting that vertebrate and invertebrate eyes come from the same evolutionary chain (It has been debated for years) still, it is just one small step closer to understanding the complexity. Nice article, and we have miles to go before we sleep.

Calilasseia's picture
However, it is apposite to

However, it is apposite to notice that the same gene is implicated as the master gene controlling eye evolution in both invertebrates and vertebrates. Time for me to reprise a past post on eye evolution (with paper references) that should prove illuminating here ...

The Grand Eye Evolution Exposition

Eye evolution is one of those vexed topics that continues to resurface on forums such as this, and indeed, will probably continue to resurface, courtesy not only of the wilful ignorance (not to mention ideologically motivated discoursive duplicity) of critics of evolutionary theory (Darwin quote mines, anyone?), but because incredulity still persists with respect to this topic. Therefore, I thought it apposite to collect, in one place, as large a body of scientific work on eye evolution, with reference to as many relevant scientific papers as possible, that can be fitted into the confines of posts within these forums. Among those papers are a brace of papers on blind cave fishes, which are particularly apposite with respect to eye evolution, and the illumination of relevant processes and relevant genes.

Having engaged in a literature search on the Mexican Blind Cave Characin, Astyanax mexicanus, for another thread elsewhere, I thought I would bring the material over here, as it makes compelling reading. I found several interesting papers, and provide links to those that can be downloaded in full, and quotes from abstracts where appropriate in order to highlight specific points. Where possible, I shall also provide detailed exposition of the contents of some papers. As a consequence, this is going to be a long post.

Basically, the development of the eye in Astyanax mexicanus is controlled by the genes Pax6, shh and twhh among others, though these are, thus far, the ones principally implicated. I suspect that somewhere along the lines, a linkage with HOX genes and possibly even bmp signalling may prove to be part of the total picture once the requisite research is performed - it wouldn't surprise me if this was the case, given how HOX genes and bmp signalling turn up in a diverse range of other developmental mechanisms, but for now, shh and twhh appear to be the prime movers signalling wise.

So, on to the papers! First ...

Evidence for Multiple Genetic Forms With Similar Eyeless Phenotypes In The Blind Cavefish, Astyanax mexicanus by Thomas E. Dowling, David P. Martasian and William R. Jeffery, Molecular Biology & Evolution, 19(4): 446-455 (2002), which can be downloaded in full and read at leisure as a PDF document from here, and the abstract reads thus:

A diverse group of animals has adapted to caves and lost their eyes and pigmentation, but little is known about how these animals and their striking phenotypes have evolved. The teleost Astyanax mexicanus consists of an eyed epigean form (surface fish) and at least 29 different populations of eyeless hypogean forms (cavefish). Current alternative hypotheses suggest that adaptation to cave environments may have occurred either once or multiple times during the evolutionary history of this species. If the latter is true, the unique phenotypes of different cave dwelling populations may result from convergence of form, and different genetic changes and developmental processes may have similar morphological consequences. Here we report an analysis of variation in the mitochondrial NADH dehydrogenase 2 (ND2) gene among different surface fish and cavefish populations. The results identify a minimum of two genetically distinctive cavefish lineages with similar eyeless phenotypes. The distinction between these divergent forms is supported by differences in the number of rib-bearing thoracic vertebrae in their axial skeletons. The geographic distribution of ND2 haplotypes is consistent with roles for multiple founder events and introgressive hybridization in the evolution of cave-related phenotypes. The existence of multiple genetic lineages makes A. mexicanus an excellent model to study convergence and the genes and developmental pathways involved in the evolution of the eye and pigment degeneration.

Other interesting papers include:

Genetic Analysis of Cavefish Reveals Molecular Convergence in the Evolution of Albinism by Meredith E. Protas, Candace Hershey, Dawn Kochanek, Yi Zhou, Horst Wilkens, William R. Jeffery, Leonard I. Zon, Richard Borowsky and Clifford J. Tabin, Nature Genetics, 38(1): 107-111 (January 2006) which can be downloaded in full from here;

Hedgehog Signalling Controls Eye Degeneration in Blind Cavefish by Yoshiyuki Yamamoto, David W. Stock and William R. Jeffery, Nature, 431: 844-847 (19 July 2004) (not a free download but the abstract is online).

From the latter paper about Hedgehog signalling, I provide the abstract:

Hedgehog (Hh) proteins are responsible for critical signalling events during development but their evolutionary roles remain to be determined. Here we show that hh gene expression at the embryonic midline controls eye degeneration in blind cavefish. We use the teleost Astyanax mexicanus, a single species with an eyed surface-dwelling form (surface fish) and many blind cave forms (cavefish), to study the evolution of eye degeneration. Small eye primordia are formed during cavefish embryogenesis, which later arrest in development, degenerate and sink into the orbits. Eye degeneration is caused by apoptosis of the embryonic lens, and transplanting a surface fish embryonic lens into a cavefish optic cup can restore a complete eye. Here we show that sonic hedgehog (shh) and tiggy-winkle hedgehog (twhh) gene expression is expanded along the anterior embryonic midline in several different cavefish populations. The expansion of hh signalling results in hyperactivation of downstream genes, lens apoptosis and arrested eye growth and development. These features can be mimicked in surface fish by twhh and/or shh overexpression, supporting the role of hh signalling in the evolution of cavefish eye regression.

So it transpires that if you transplant an embryonic lens taken from a surface-dwelling Astyanax mexicanus with normal eyes into the optic cup of an embryonic blind cavefish, normal eye development resumes. Interesting is it not? And, by manipulating the shh and twhh gene expression in surface-dwelling eyed fishes during embryonic development, the scientists were able to reproduce the eye apoptosis seen in the cave dwelling fishes.

An additional paper (downloadable in full from here) is this one:

Early and Late Changes in Pax6 Experession Accompany Eye Degeneration During Blind Cavefish Development by Allen G. Strickler, Yoshiyuki Yamamoto and William R. Jeffery, Development, Genes & Evolution, 211(3): 138-144 (March 2001)

in which the role of the Pax6 gene (which is common to eye development across a wide range of organisms) is examined with respect to the differences in development between the surface-dwelling form of Astyanax mexicanus and the cave-dwelling forms.

From here, you can download the following paper:

Eyed Cave Fish In A Karst Window by Luis Espinasa and Richard Borowsky, Journal of Cave and Karst Studies, 62(3): 180-183 (2000)

which describes the co-existence of eyed and eyeless forms in a cave with a karst window, and the abstract makes VERY interesting reading indeed - namely:

We considered four hypotheses for the origin of Caballo Moro eyed cave fish. The RAPD data rule out that the mixed population represents a transitional stage of evolution, or that the eyed fish are unmodified surface immigrants. We cannot rule out that the eyed fish are the direct descendants of surface fish that have acquired markers from blind fish by hybridization, although the apparent distinctness of the two sub-populations suggests otherwise. An alternative hypothesis, that the eyed fish of the cave are direct descendants of blind cave fish that re-acquired eyes with the opening of the karst window, is consistent
with the data and tentatively accepted.

So here, we have the first cited evidence that when a blind cave population was, by a serendipitous accident, granted readmission to daytime light sources, some of the blind cave fishes regained their eyes over time.

Indeed, I'll cover this paper in more detail, as it makes interesting reading to put it mildly. The paper opens as follows:

Caballo Moro, a karst window cave in northeastern Mexico, supports a mixed population of cave Astyanax mexicanus : eyed and eyeless. The relationships of these sub-populations to one another and to other populations of Mexican tetras were examined using RAPD DNA fingerprint markers. The eyed tetras of Caballo Moro Cave are genetically closer to blind tetras from Caballo Moro and other caves in the region than they are to eyed tetras from the surface. The two forms are not genetically identical, however, and may represent distinct sub-populations.

Eyed and eyeless fish have a distributional bias in the cave, with eyed fish preferentially in the illuminated area and blind fish in the dark zone. Aggression of eyed towards blind fish in the illuminated area contributes to this bias and may serve to stabilize the eye-state polymorphism.

We considered four hypotheses for the origin of Caballo Moro eyed cave fish. The RAPD data rule out that the mixed population represents a transitional stage of evolution, or that the eyed fish are unmodified surface immigrants. We cannot rule out that the eyed fish are the direct descendants of surface fish that have acquired markers from blind fish by hybridization, although the apparent distinctness of the two sub-populations suggests otherwise. An alternative hypothesis, that the eyed fish of the cave are direct descendants of blind cave fish that re-acquired eyes with the opening of the karst window, is consistent with the data and tentatively accepted.

So, there exists a cave in Mexico called Caballo Moro, that has a karst window admitting light, and within this cave, within reach of the light admitted by the karst window, there is a population of Astyanax mexicanus. This population contains fishes that have lost their eyes, conforming to the phenotype that was once described via the taxon Anophthichthys jordani, that taxon now recognised as a junior synonym of Astyanax mexicanus. However, the population contains fishes with functioning eyes. It is tempting to think that the eyed fishes are members of a surface-dwelling population that have become intermingled with the cave fishes, and, courtesy of still having access to light, retained their eyes. The population genetics of Astyanax mexicanus have been extensively studied, and as a consequence, a great deal is known about the surface-dwelling and cave-dwelling populations of these fishes, including the fact that there exist distinct genetic markers for distinct populations, allowing scientists to alight upon the fact that the eye-loss phenotype has arisen in multiple separate populations independently, via a range of acquired mutations. The relevant paper containing evidence for this is one I've already cited above, namely:

Evidence For Multiple Genetic Forms With Similar Eyeless Phenotypes In The Blind Cavefish, Astyanax mexicanus by Thomas E. Dowling, David P. Martasian and William R. Jeffery, Molecular & Biological Evolution, 19(4): 446-455 (2002)

I'll leave that paper aside for the moment, as I've dealt with it elsewhere in the past, and can always return to it in detail in another post. However, that paper establishes that different cave populations of Astyanax mexicanus possess an eyeless phenotype arising via different sets of mutations in the genes responsible for eye development (namely Pax6, shh and twhh, about which I have posted in the past, including the paper covering Pax6 as a master gene in eye development). Likewise, populations of the surface dwelling eyed phenotype have genetic markers identifying them as belonging to particular populations, where those populations experience little or no gene flow with other populations, and consequently, the provenance of a fish can be determined with reasonable precision by appropriate genetic analysis. The authors of the paper I am covering here have established that the eyed phenotype fishes in the Caballo Moro karst window cave possess genetic markers identifying them as having derived from ancestral eyeless stock. Which means that these fishes had eyeless ancestors, and consequently regained functioning eyes once light was present.

So, it remains to cover the present paper in more detail, and examine the evidence presented therein. Let's do that shall we?

The Mexican Tetra, Astyanax mexicanus, is a visually orienting, schooling fish widely distributed in surface streams of northern Mexico. In addition to the epigean populations, numerous cave forms of this species occur in the Sierra de El Abra region of northeast Mexico (Fig. 1; Mitchell et al. 1977). In contrast to the surface fish, these troglobitic forms have rudimentary, non-functional eyes, and their melanin pigmentation is reduced or absent.

Generally, caves with troglobitic Mexican tetras do not contain eyed tetras, except for the occasional doomed individual swept underground. One exception is El Sótano de El Caballo Moro, which contains an apparently stable, mixed population of A. mexicanus, both eyed and eyeless.

The entrance of Caballo Moro Cave (CMC) is a karst window. Karst windows are habitats within cave systems that are exposed to light, and typically result from cave passage collapse. The 50-m deep entrance pit of CMC is found at the bottom of a 60-m doline, and leads directly to a large “lake” of approximately 18 m x 90 m. (Cave “lake” in this case, is a wide stream pool). Light reaches only the upstream half of the lake, while the downstream half remains in darkness. The lake contains both blind depigmented and eyed pigmented forms of A. mexicanus. The distribution of fish in the lake appears to be biased, with over-representations of blind fish in the dark area and eyed fish in the light area.

Mitchell et al. (1977) observed that the source of the eyed fish of Caballo Moro cave was a mystery. The cave’s entrance pit is 11 km away from the nearest potential resurgence and does not capture a surface stream. Furthermore, there is no permanent water nearby. The nearest recorded surface fish locality in the Río Boquillas system is 4 km distant. They hypothesized that seasonal flooding of Río Boquillas tributaries affords occasional access to the cave through, as yet undetected, sinks.

As part of a larger study of the evolutionary history of the Mexican cave tetra, we investigated the relationships of the eyed fish of CMC. If they represent an unmodified surface population recently captured from a nearby sink, their presence in the karst window would be unremarkable. If, on the other hand, the population were of long standing, it would raise the question of the maintenance of its integrity in the face of potential hybridization with, and introgression of genes from, the troglobites. Alternatively, if the eyed fish of the cave originated from blind cave progenitors, they would make a good model for study of the effects of the reversal of selection pressures on populations.

And thus, the groundwork is laid for what follows. Namely, that there is no obvious source of eyed fishes from surface or epigean populations, with the cave running for 11 Km underground, without capturing a surface stream between the cave's entrance pit and the karst window illuminating the population of interest. Moreover, the nearest population of epigean fishes is 4 Km distant from the cave, and there is no obvious connection between the body of water containing that epigean population, and the mixed population of fishes in the karst window lake, which comprises a mixture of epigean and hypogean (cave-phenotype) fishes. So, the possibilities are:

[1] The epigean phenotype fishes (possessing pigmentation and functional eyes) are a recent arrival, courtesy of an as yet unknown connection between the cave system and a surface body of water supplying these fishes;

[2] The epigean phenotype fishes have coexisted with the hypogean phenotype (eyeless and depigmented) fishes for an extended period of time with little or no interbreeding;

[3] The epigean phenotype fishes have arisen from hypogean ancestors.

[1] is considered unlikely by the authors, given the known geography of the cave system, but is required to be ruled out evidentially. [2] poses problems with respect to the appearance of an isolating mechanism between the two phenotypes, given that prior breeding experiments have established that epigean and hypogean fishes are capable of mating and producing offspring. [3], meanwhile, would provide an extremely interesting example of evolution reversing a character change that had previously occurred in these fishes, but requires evidential support before the postulate can be considered valid. So, let's see what the authors discovered upon further analysis! First, the authors outline their experimental procedures:


The relationships among representative surface and cave populations of Astyanax mexicanus from the El Abra region were studied using RAPD data. RAPD (synonymous with APPCR) technique generates a DNA fingerprint from genomic DNA using the polymerase chain reaction (Welsh & McClelland 1990; Williams et al. 1990). RAPD fingerprints are species and population specific and carry significant amounts of taxonomic information (Borowsky et al. 1995).

The following populations were sampled (Fig. 1): caves: Molino, Vasquez, and Caballo Moro: surface: Río Frío, Río Boquillas and Río Comandante. Astyanax aeneus from the Río Granadas, a tributary to the Río Amacuzac, northeast of Taxco, Guerrero, Mexico, were used as the outgroup for phylogenetic analyses. Two individuals each were examined from Molino cave, Vasquez cave, all surface localities, and A. aeneus. Five blind individuals and six eyed individuals were examined from CMC. RAPD amplification procedures followed Borowsky et al. (1995). Two primers were used: Mey7 (5’ggagtaggggatatgatcgatgga3’) and Mey8 (5’cagcaaacagaaaccagtcag3’). Reactions were cycled five times in a Hybaid thermocycler: 94°C for 70 s, 40°C for 5 minutes, and 72°C for 3 minutes, followed by 35 cycles at higher stringency: 94°C for 70 s, 50°C for 1 minute, and 72°C for 90 s. Reaction products were run on 6% polyacrylamide gels (29:1) and silver stained (after Gottlieb & Chavko 1987). RAPD fragment distributions were compared among individuals using a size match criterion. Each uniquely sized band was assumed to be a character, and character states were scored as “present” or “absent.”

Phylogenetic analysis of the data was done using Paup 4.0b2 software (Swofford 1999). Maximum parsimony analysis (character states unordered) was done by bootstrapping the data (1000 replicates) using full heuristic search to produce a 50% majority-rule consensus tree. For analysis of distance (“mean character difference”), neighbor-joining trees were generated from bootstrapped data (1000 replicates) and used to obtain a 50% majority-rule consensus tree.

A supplementary analysis was done using a Monte Carlo procedure to estimate the variance of distances among individuals within and between the sets of eyed and eyeless fish from CMC. Individual phenotypes for distance comparisons were created by sampling, based on the frequencies of bands in each set. Twenty such pairs of phenotypes were generated for each simulation and the calculated distances were used to estimate means and their standard deviations. For this analysis, distances were calculated as the sum of the absolute differences in band frequencies among taxa or individuals divided by the number of bands.

Now, comes the analytical results!


One hundred and fifty-eight bands were scored, of which 127 were variable and of value in distance analysis, and 58 were parsimony informative. The number of bands observed in any individual ranged from 55-69. The raw data matrix presented as table 1, is organized in the style of a “sequence alignment.” As such, it arrays the character states of the outgroup species along the top row (+, -, and “P” for polymorphic). The character states for the other taxa are arrayed below, using “.” to denote state identity with the outgroup, and the other symbols, where different from the outgroup. The data were sorted by character states in the cave fish, putting “-“ towards the left and “+” towards the right. This arrangement makes apparent a series of derived bands shared among all cave fishes or among all individuals of Caballo Moro cave. These synapomorphies imply a closer relationship of the eyed fish of Caballo Moro cave to other cave fish than to epigean fish.

Indeed, the accompanying figure is quite impressive (see Table 1 charting the RAPD bands for the various populations). The Caballo Moro fishes are manifestly members of a well-defined and genetically distinct grouping, exhibit a well-defined clustering of bands from the DNA analysis that are only partially shared with individuals from the Molino and Vasquez caves (the other two cave populations sampled), and there are marked differences between the Río Frío, Río Boquillas and Rio Comandante fishes and those from Caballo Moro.

Moving on:

This implication is supported by both parsimony and distance analyses, which gave essentially the same result: consensus trees with two clusters, one consisting of the epigean populations and the other of the cave populations. The tree produced by distance analysis (Fig. 2) had a little more structure than the one based on parsimony and may be more appropriate for analysis of populations that can hybridize. The relationship of the eyed and blind fish of Caballo Moro cave is strongly supported by a bootstrap value of 0.83 as is the clustering of all fish of Caballo Moro cave with the other cave fish (bootstrap value of 0.82). The tree also shows a clustering of four of the five blind fish within Caballo Moro, which suggests that the eyed and blind fish of the cave may comprise two distinct sub-populations, in spite of their closeness.

The supplemental distance analysis lends some support to these hypotheses. Distances calculated among populations showed the eyed and eyeless fish of CMC to be closer to each other (0.101) than either was to surface fish (0.337 and 0.359, respectively) or to the other blind cave fish (0.253 and 0.240, respectively). The distance between the eyed and eyeless fish of CMC was investigated in more detail by Monte Carlo simulation. The average distance between simulated eyed and eyeless individuals (0.095 ± 0.015) was significantly greater than the average distance between simulated eyed individuals (0.067 ± 0.016, t(38) = 3.83, p < 0.05) or simulated eyeless individuals (0.031 ± 0.012, t(38) = 9.60, p < 0.05). The means and standard deviations of all the distances measured among the real individuals in the two groups are very similar to those in the simulation (between sets: 0.1226 ± 0.0264; eyed: 0.0965 ± 0.0215; eyeless: 0.0513 ± 0.168) and the t values are high (t(38) = 9.60 and t(43) = 3.24), but only the t tests in the simulation are valid.

Of 41 fish collected from Caballo Moro Cave, 21 were eyed and pigmented, and eighteen had eye rudiments completely covered by muscle and scales and were depigmented. Two were intermediate in phenotype. The collection made from the dark side of the lake had eight fish, one with eyes. The collection made from the illuminated side of the lake had seventeen fish, ten with eyes (locations of other specimens had not been recorded). The biased distribution is statistically significant (Fisher’s exact test, p < 0.05). We observed eyed fish nipping and chasing blind fish on the illuminated side, and this behavior may contribute to the distributional bias within the lake.

Basically, the above tests establish that the Caballo Moro fish form a genetically distinct group, and that furthermore, there exists an interesting set of relations between the eyed and eyeless fishes, which closely matches that of a Monte Carlo simulation of the emergence of eyed and eyeless fishes in that group.

With that, it's time to move on to the authors' discussion of their results:


At least four hypotheses could account for the presence of eyed fish in Caballo Moro cave. The first is that the eyed individuals are surface fish recently swept underground. As such, their residency might be short-lived and they would not necessarily be part of the troglobitic population. A second hypothesis is that the eyed fish represent one phenotypic extreme of a variable cave fish population in evolutionary transition towards eyelessness. A third is that they are the descendants of surface fish swept underground that had interbred with the blind fish and acquired their RAPD marker set by hybridization. A fourth is that the eyed fish are descendants of blind, depigmented cave fish that reacquired eyes and pigmentation through an evolutionary process. The reacquisition of eyes and pigment in troglobites reintroduced to light has been suggested before, for karst window populations of the amphipod Gammarus minus (Culver et al. 1995).

We reject the first hypothesis because it predicts that the eyed fish of CMC should be genetically closer to surface fish than to the blind cave fish. Our results showed the opposite to be true; both distance and parsimony analyses clustered the eyed fish of the cave with blind cave fish rather than surface fish. This clustering was well supported by bootstrap analysis (Fig. 2).

What of the second hypothesis? Is the CMC population in transition from an eyed to a blind condition? Wilkens (1988) hypothesized such a situation in the isolated cave populations of the Micos area, to the west of the El Abra. Micos fish have reduced eyes, but the rudiments are better developed than in the cave tetras of the Sierra de El Abra region, and Micos fish are not fully depigmented. Wilkens suggested that the Micos cave tetras are in transition because they are “phylogenetically younger” than other populations of troglobitic Mexican Tetras, and our (unpublished) RAPD data support this contention.

Nevertheless, we think it unlikely that the CMC population is in transition between the eyed and blind conditions, as in the Micos fish. First, Caballo Moro cave is centrally located within the range of other populations of cave tetras, none of which appear to be in a transitional state. Second, the fish of the Micos caves are uniformly intermediate in eye size and pigmentation phenotype according to Wilkens (1988) and our unpublished observations, while most (95%) of the Caballo Moro cave fish fall into two distinct morphological groups — eyes functional versus blind. Thus, any intermediate “transitional” quality of the CMC population exists primarily as a statistical average of two phenotypic extremes.

We cannot yet distinguish between the third and fourth hypotheses: the eyed fish of the cave may have descended from a captured surface population having interbred extensively with the blind fish or it may have descended from blind cave ancestors by reacquisition of eyes and pigment. Both hypotheses predict extensive sharing of character states among eyed and eyeless fish from CMC and might prove difficult to distinguish in practice.

A test based on distance data may be possible. Our results show that the average distance between eyed and eyeless individuals of CMC is significantly greater than the average distances within these sets. A biologically significant genetic distance between the two groups of fish would arise in different ways according to the two hypotheses. Hypothesis three is one of introgressive hybridization, and would view distance as evidence of a mixing process not yet complete. Hypothesis four is one of centripetal evolution and would view distance as a derived state, as one subset splits from the other. Thus, hypothesis three predicts the eyed fish of CMC to be closer than their eyeless companions to the fish of the surface and more distant from the fish of the other caves. Instead, our data show both groups in CMC to be equally far from surface fish and equally far from the other cave fish. Thus, the current data support hypothesis four, but more will be necessary for a definitive test.

The data presented here confirm the status of the CMC population as one worth further study for the light it can shed upon evolutionary processes. Karst windows, in general, should provide unique opportunities to study the effects of the alteration of selective pressures on troglobites and the ecological and evolutionary interactions between troglobitic and surface species.

Now the authors are being appropriately cautious here, with respect to the data that they have obtained, but, that data is more consistent with the hypothesis of the eyed fishes of Caballo Moro having arisen from eyeless ancestors, than it is with competing hypotheses. Which means, if confirmed by more in-depth study involving larger data sets, that the eyed specimens of Astyanax mexicanus resident in the Caballo Moro karst window lake are fishes that have regained functional eyes, courtesy of appropriate mutations being positively selected for in their lineage. It would be interesting to examine the genetic data for the Pax6, shh and twhh genes for these fishes, as, given their known role in the appearance of the eyeless phenotype in other hypogean lineages of Astyanax mexicanus.

Now, aside from the fact that the above refutes wholesale any notion that selection cannot affect the dissemination of particular genes, or shape the inheritance thereof (which as susu.exp has already noted on numerous occasions elsewhere, is based upon a singularly woeful lack of understanding of basic biology - some critics of evolution apparently hasn't heard of meiosis, apart from anything else), the above findings also drive a tank battalion through creationist quote mining of Crow's paper, because here we have an instance of purported 'genetic deterioration' being thrown into full reverse by evolutionary processes, something which creationist assertions about "genomic entropy" claim simply cannot happen. Once again, the real world demonstrates that blind creationist assertion is nothing more than that - blind assertion.

So, looks like the evidence for the active evolution of these fishes is pretty compelling.

Meanwhile, it's time to move on from the blind cave fishes somewhat, and concentrate upon Pax6. The papers extant in this area are very interesting. Indeed, as if yet more evidence for the importance of Pax6 was needed, here is the Ensembl page covering the Pax6 gene and the oculorhombin protein that it codes for. That page notes that the following diseases are caused by mutations in Pax6:

[1] Aniridia type II - partial or complete absence of the iris, absence of the fovea and malformations of the lens (among other structural malformations). Approximately 67% of these defects are familial, and the inheritance mechanism is autosomal dominant;

[2] Peter's Anomaly - the site describes this condition thus:

Peter's anomaly consists of a central corneal leukoma, absence of the posterior corneal stroma and descemet membrane, and a variable degree of iris and lenticular attachments to the central aspect of the posterior cornea.

In other words, more severe eye defects;

[3] Ectopia pupillae - failure of the pupil to be properly centred;

[4] Foveal hypoplasia - failure of the fovea to develop fully during embryogenesis, inheritance again autosomal dominant;

[5] Autosomal dominant keratitis - opacity of the cornea with accompanying vascularisation, often associated with foveal hypoplasia above;

[6] Ocular coloboma - abnormal development of the optic cup and stalk, accompanied by holes appearing in various eye structures;

[7] Bilateral optic nerve hypoplasia - failure of the optic nerve to develop properly, again with autosomal dominant inheritance;

Here's the human Pax6 gene, formatted using my nice Visual Basic applet:


Here's the protein it codes for, again nicely formatted using my applet:

LQ Ochre

(The legend "Ochre" at the end refers to the fact that the gene ends with an Ochre stop codon, TAA - no amino acid is coded for by this codon).

It's instructive to look at some variants for this gene. Here's one associated with Aniridia Type II:


Now already we know something is wrong here because the gene is a different size. But the BIG surprise is what happens when we look at the protein it codes for ...

YR Ochre

Oops. MAJOR malfunction here!

Basically, this mutant form of the gene fails to code for a working protein full stop. The transcription process hits an Ochre stop codon at the third coding triplet.

Furthermore, the extant online gene databases inform me that there are variations as follows associated with Peter's Anomaly:

[1] Substitution of codon triplet for W (tryptophan) replacing G (glycine) at codon position 18 (bp 52-54);

[2] Subsititution of codon triplet for R (arginine) replacing G (glycine) at codon position 26 (bp 76-78);

[3] Substitution of codon triplet for V (valine) replacing D (aspartic acid) at codon position 53 (bp 157-159) - found principally in Japanese human lineages manifesting the disease (ethnospecific), also associated with congenital cataract and foveal hypoplasia in affected individuals;

[4] Substitution of codon triplet for S (serine) replacing P (proline) at codon position 363 (bp 727-729);

So, it looks once again as if real science knows a LOT more about eye evolution than mendacious propagandists for creationist fantasies dare to even imagine it is possible to know.

Meanwhile, I'll also reprise this material - apologies if I repeat citations of papers cited above here:

Adaptive Evolution of Eye Degeneration in the Mexican Blind Cavefish by W. R. Jeffrey, journal of Heredity, 96(3): 185-196 (Jan 2005) - explains how selection is a key factor in the evolution of eye degeneration in cave fishes

Cavefish as a Model System in Evolutionary Developmental Biology by William R. Jeffrey, Developmental Biology, 231:, 1-12 (1 Mar 2001) - contains experimental tests of hypotheses about eye evolution

Hedgehog Signalling Controls Eye Degeneration in Blind Cavefish by Yoshiyuki Yamamoto, David W. Stock and William R. Jeffery, Nature, 431: 844-847 (14 Oct 2004) - direct experimental test of theories about eye evolution and the elucidation of the controlling genes involved

The Master Control Gene For Morphogenesis And Evolution Of The Eye by Walter J. Gehrig, Genes to Cells, 1: 11-15, 1996 - direct experimental test of hypotheses concerning eye evolution including the elucidation of the connection between the Droso gene and eye morphogenesis, and the experimental manipulation of that gene to control eye development

Why cavefish are blind by Natasha .M. Tian & David .J. Price, Bioessays, 27: 235-238 (Mar 2005) - also reports on the connection between the Pax6 and hedgehog signalling genes and how these are subject to selection over time

Let's have a look at some of the contents of those papers shall we? Starting with the Jeffrey et al paper ...


The evolutionary mechanisms responsible for eye degeneration in cave-adapted animals have not been resolved. Opposing hypotheses invoking neural mutation or natural selection, each with certain genetic and developmental expectations, have been advanced to explain eye regression, although little or no experimental evidence has been presented to support or reject either theory. Here we review recent developmental and molecular studies in the teleost Astyanax mexicanus, a single species consisting of a sighted surface-dwelling form (surface fish) and many blind cave-dwelling forms (cavefish), which shed new light on this problem. The manner of eye development and degeneration, the ability to experimentally restore eyes, gene expression patterns, and comparisons between different cavefish populations all provide important clues for understanding the evolutionary forces responsible for eye degeneration. A key discovery is that Hedgehog midline signaling is expanded and inhibits eye formation by inducing lens apoptosis in cavefish embryos. Accordingly, eyes could have been lost by default as a consequence of natural selection for constructive traits, such as feeding structures, which are positively regulated by Hh signaling. We conclude from these studies that eye degeneration in cavefish may be caused by adaptive evolution and pleiotropy.

Oh dear. The hard evidence from the real world supports evolution. Let's look at the next paper:


The Mexican tetra Astyanax mexicanus has many of the favorable attributes that have made the zebrafish a model system in developmental biology. The existence of eyed surface (surface fish) and blind cave (cavefish) dwelling forms in Astyanax also provides an attractive system for studying the evolution of developmental mechanisms. The polarity of evolutionary changes and the environmental conditions leading to the cavefish phenotype are known with certainty, and several different cavefish populations have evolved constructive and regressive changes independently. The constructive changes include enhancement of the feeding apparatus (jaws, taste buds, and teeth) and the mechanosensory system of cranial neuromasts. The homeobox gene Prox 1, which is expressed in the expanded taste buds and cranial neuromasts, is one of the genes involved in the constructive changes in sensory organ development. The regressive changes include loss of pigmentation and eye degeneration. Although adult cavefish lack functional eyes, small eye primordia are formed during embryogenesis, which later arrest in development, degenerate, and sink into the orbit. Apoptosis and lens signaling to other eye parts, such as the cornea, iris, and retina, result in the arrest of eye development and ultimate optic degeneration. Accordingly, an eye with restored cornea, iris, and retinal photoreceptor cells is formed when a surface fish lens is transplanted into a cavefish optic cup, indicating that cavefish optic tissues have conserved the ability to respond to lens signaling. Genetic analysis indicates that multiple genes regulate eye degeneration, and molecular studies suggest that Pax6 may be one of the genes controlling cavefish eye degeneration. Further studies of the Astyanax system will contribute to our understanding of the evolution of developmental mechanisms in vertebrates.

Oh look. More hard evidence from the real world supporting eye evolution. Namely that:

[1] Different cave fish populations evolved the eye apoptosis mechanism independently, and have different mutations coding for this;

[2] Other senses, particularly those connected with feeding efficiency, are enhanced in the blind cave populations of Astyanax mexicanus, and the underlying genetic mechanism for this is being elucidated, with special reference to the Prox 1 gene;

[3] In embryonic fishes, eye formation begins normally, but then undergoes reversal because of cell apoptosis controlled by signalling from the lens tissues, and experimental transplantation of a normal lens from a sighted embryo into an optic cup belonging to a cave dwelling embryo results in the restoration of normal eye formation;

[4] The genes involved in this process are now known, and the Pax6 gene, which has been demonstrated experimentally to be the master control gene for eye morphogenesis, is involved in the differential formation of eyes in cave dwelling Astyanax mexicanus populations.

However, one of the best papers I can present is this - the very paper that supports the statement I have just made above about the role of Pax6, namely:

The Master Control Gene For Morphgenesis And Evolution Of The Eye by Walter J. Gehrig, Genes To Cells, 1: 11-15, 1996.

I quote:

Abstract. The human aniridia, the murine small eye, and the eyeless mutations of Drosophila affect homologous (Pax-6) genes that contain both a paired- and a homeobox. By ectopic expression of these genes, functional eyes can be induced on the legs, wings and antennae of the fly, indicating that eyeless (Pax-6) is the master control gene for eye morphogenesis. The finding of Pax-6 from flatworms to humans suggests that eyeless is a universal master control gene and that the various types of eyes in the various animal phyla may have evolved from a single prototype.

However, the best part is when we look at the article in detail ...

Hox and Pax genes

Homeotic mutations in Drosophila have led to the isolation of master control genes specifying the body plan. Loss- and gain-of-function mutations in these genes lead to opposite homeotic transformations: in Antennapedia (Antp) for example, loss-of-function mutations lead to the partial transformation of middle legs to antennae, whereas gain-of-function mutations induce the transformation of antennae into middle legs. These transformations in opposite directions suggest that Antp is a switch gene inducing the leg development pathway. We have tested this hypothesis by expressing the normal ANTP protein and asking whether leg structures can be induced in other parts of the body. As predicted, antenna-to-leg transformations can be induced in transgenic flies carrying an Antp cDNA gene under the control of a heat shock promoter (Scheuwly et al., 1987). Heat induction of this transgene during the early third larval stage, just before the antenna becomes determined, leads to the induction of middle legs, indicating that Antp is a master control gene switching on all the genes required for leg morphogenesis. This experiment was the first attempt to redesign the body plan of the fly. Even though the heat shock promoter induces the ANTP protein all over the animal the morphogenetic effect is restricted to the antennae. In the more posterior body segments, Antp has to compete with the homeotic genes of the bithorax complex that specify the more posterior body segments, each segment being specified by a particular combination of homeotic proteins. Hameotic genes are characterised by the homeobox, a 180 bp DNA segment encoding the homeodomain, the DNA binding domain of the respective proteins (McGinnis et al., 1984a,b; Scott & Weiner 1984). The homeotic proteins serve as transcription factors controlling a large number of subordinate genes involved in morphogenesis.

The paired box encodes another DNA binding domain and characterises the Pax genes (see Noll 1993 for review). The Pax genes are a perfect example of what has been called evolutionary tinkering (Jacob 1977). Some Pax genes have a paired box only, some have both a paired and a homeobox, and some have a paired and a partial homeobox. This indicates that in the course of evolution new genes can be generated by putting together bits and pieces from pre-existing genes by recombination, strongly resembling tinkering.

The Pax6 Gene

Using Drosophila probes, a family of mammalian Pax genes has been cloned, including Pax-6 which includes a paired and a homeobox (Walther & Gruss 1991). Subequently it was shown that the murine Small eye (Hill et al., 1991) and the human Aniridia (Ton et al., 1991) mutations affect the respective Pax-6 genes.Mice heterozygous for Small eye (Sey) mutations have reduced eyes, whereas homozygous carriers of the mutation are lethal and lack eyes as well as the nose. The human Aniridia (An) syndrome has a similar phenotype with heterozygotes having reduced eyes sometimes lacking the iris, and a putatively homozygous, lethal foetus lacking eyes completely has also been described. Pax-6 is expressed in the spinal cord, parts of the brain and particularly, at all stages of eye morphogenesis, first in the optic sulcus, then in the optic vesicle, the pigmented and the neural retina, the iris, in the lens and finally in the cornea. This expression pattern led to the suggestion that Sey might control eye induction (Walther and Gruss 1991). The induction of the lens by the optic cup had been demonstrated in frogs by Spemann (1901) and Lewis (1904) who deserve credit for the first experimental documentation of a case of embryonic induction: when the optic vesicle is transplanted under the flank epidermis, an ectopic eye with a lens is induced. However, Spemann and Lewis did not consider the possibility that genes might control eye induction.

On the basis of a comparison between the Pax genes of mammals and Drosophila, Noll (1993) proposed that certain genes are homologous, but no homologue for the mammalian Pax6 gene had been found in Drosophila. This gene was discovered in my laboratory quite accidentally in a control experiment (Quiring et al., 1994). Even though a Pax-6 homologue was expected in Drosophila, it came as a great surprise that mutations in this gene have an eyeless phenotype. The first eyeless (ey) mutation was discovered as early as 1915 by Hoge (Hoge 1915). Homozygous ey mutant flies have strongly reduced eyes or they lack eyes completely. The cloned Pax-6 homologue maps to section 102D on chromosome 4, at the ey locus. In two independent spontaneous mutations, ey(2) and ey(R), the cloned gene carries two different transposon insertions, and in both mutants the cloned gene is neither expressed in the eye primordia of the embryo nor in the eye imaginal discs of the larva, strongly suggesting that the cloned Pax-6 gene represents ey. The transposon insertions disrupt an eye specific enhancer preventing gene expression in the eye primordia. Furthermore, the sequence conservation between the mammalian and insect gene are very high: 95% amino acid sequence identity is found in the paired boxes and 90% between the homeoboxes plus some scattered identity outside of the boxes. Also, two out of three intron splice sites in the paired box, and and one of the two splice sites in the homeobox are conserved, indicating that these genes are true homologues.

This is an unexpected finding since the single lens eye of vertebrates was generally considered to have evolved independently of the compound eye of insects because these two eye types are morphologically completely different. Since homologous organs share variations of the same genetic programme, the possible homology between the insect and the vertebrate eye has to be reconsidered.

Eyeless is the master control gene for eye morphogenesis

The high degree of sequence conservation between ey and Sey or An and the similarity of the mutant phenotypes, as well as the patterns of expression, suggested that these genes play a key role in eye morphogenesis and evolution. However, it was not obvious that they are master control genes since the loss-of-function mutations lead to a loss of eye structures, rather than their homeotic transformation. Thus, the mutational block might occur at the initial steps of the eye development pathway, which is compatible with a possible tole of ey as a master control gene, but it may also occur at a lower level of the hierarchy. This is exemplified by other mutations blocking eye development, like eyes absent, which act downstream of ey. In order to find out whether ey is a master control gene, I planned to construct a gain-of-function mutant and to express the normal EY protein ectopically in other body parts of the fly. The prediction was that the ectopic expression of EY protein would induce ectopic eye structures, if ey were a master control gene. I was encouraged to try out this bold experiment since I knew from transdetermination experiments that wing imaginal disc tissue that is cultured continuously in the abdominal cavity of female flies (Gehrig et al., 1968) can eventually give rise to eye facets (Fig. 2A). This raised the possibility that ey[sup]+[/sup] might induce eye structures at least in wing discs. My collaborators Georg Halder and Patrick Callaerts used both the heat shock vector for ubiquitous expression of ey(+) (as in the case of Antp mentioned above) and the GAL4 system for targeted gene expression (Brand & Perrimon 1993) by means of enhancer detector strains expressing GAL4 and/or leg and antennal discs. As indicated in Fig. 1, GAL4 drives the expression of an ey(+) gene carrying several GAL4 upstream activating sequences (UAS). By crossing the GAL4 enhancer detection stock with the UAS-ey[sup]+[/sup] target gene stock, EY protein can be targeted onto the wing, leg and antennal discs. In contrast to the heat shock vector which requires precise timing of gene expression in relatively short pulses, the GAL4 system allows continuous expression.

As shown in Figs 2B and 3, ectopic eye structures can be induced by switching on the ey(+) c-DNA in the wing and antennal discs, and also in leg discs (Halder et al., 1995). The ectopic eyes are morphologically normal with normal photoreceptors, lens, cone and pigment cells and an electroretinogram as it is typical for photoreceptor cells can be recorded, when the ectopic eyes are exposed to light (P. Callaerts, unpublished data). Thus the formation of an ectopic eye can be induced by switching on a single gene. Therefore, we consider ey to be a master control gene for eye morphogenesis. In addition, ey has other functions in the brain, the nose and the ventral nervous system that remain to be determined.

Oh dear. We know rather more about the genetic processes involved in eye formation and evolution than creationists think. The requisite genes, as I've already established above, are found right across a whole brace of animal phyla from flatworms to mammals, including you and I.

One question I've never seen creationists answer with a substantive answer (as opposed to vacuous apologetics) is this: why has their magic man chose to produce cave fishes that have all the genetic and molecular machinery for eye formation, which initiates normal eye development to begin with but then goes into reverse, and moreover exhibit different mutations for this in different populations? If their magic man knew that these fishes were going to end up in caves, why bother giving them the genetic and molecular machinery for eyes in the first place? Bit of a cock up there, creating these fishes in such a manner as to provide evidence for evolution.

Also, I'm reminded of the following video clip (with bonus appearance by Stephen Jay Gould):

Eye Evolution Video

Note that all of the postulated intermediate stages exist in real living organisms today.

I think this should cover all of the relevant scientific bases with respect to eye evolution and the role of specific genes therein.

ronald bertram's picture


Posted: "Oh dear. We know rather more about the genetic processes involved in eye formation and evolution than creationists think."

Oh dear! There is more to the evidence behind evolution that the creationist think!

I often wonder if the debates are worth the effort. It is impossible to have a legitimate debate with folks who don't have the prerequisite knowledge. The juice just ain't worth the squeeze IMO.

Thanks for your effort.

Donating = Loving

Heart Icon

Bringing you atheist articles and building active godless communities takes hundreds of hours and resources each month. If you find any joy or stimulation at Atheist Republic, please consider becoming a Supporting Member with a recurring monthly donation of your choosing, between a cup of tea and a good dinner.

Or make a one-time donation in any amount.