Donkeys, Horses, Mules




and Evolution




 The Phenotypic Effects of Chromosome Variation



Sean D. Pitman M.D.

© August 2003










       Much is made of DNA or protein sequence analysis as a way of determining evolutionary ancestry.  Branching phylogenetic trees are based on such sequence analysis and are used to suggest various evolutionary relationships between different creatures on the "Tree of Life."  Even so called "pseudogenes" are often used to show supposed evolutionary relationships. However, although sequence analysis is interesting by itself and may in fact say a lot about common ancestry, such analysis does not necessarily favor the theory of common descent in many cases.  Certain differences between the DNA sequences of various creatures can easily be explained by the mindless naturalistic processes of random mutation and natural selection while other differences cannot be so easily explained.

      Not all differences are created equal.  Some differences are neutral while others are functional.  Natural selection is not required to achieve neutral genetic variations.  Because of this, neutral variations arise relatively rapidly.  Functional variations, depending on the complexity of the functions in question, might take a lot longer to evolve since there are a so many more non-functional, detrimental, or neutral sequences that could evolve when compared to the relatively few beneficial functions available to a given creature in a particular environment.  So, a simple statement that chimps and humans are 94% to 98% identical as far as DNA is concerned is deceptive in that it says nothing about the type of differences involved (neutral or functional) or if such differences are statistically possible to achieve in the amount of time allotted. 

After presenting this argument, I am occasionally asked why I would lean toward classifying horses and donkeys as the same "kind" or as having a common ancestor while at the same time classifying humans and chimps as different "kinds" not having a common ancestor?  This is not an easy question.  I do not believe that there is sufficient evidence, at this point, to definitively exclude chimps and humans from the possibility of common ancestry within an evolutionary time frame (5 to 8 million years from a "common ancestor").  I just do not think there is enough evidence to support such a theory either, to a statistical significance, over other possibilities.  More information is needed that details the types of genetic differences that exist between humans and chimps.   Some differences are obvious.  For example, gross chromosomal differences are easily noted.  Humans and chimps differ not only in chromosome number, but also by nine pericentric chromosomal inversions and one centric fusion. We also know that both humans and chimps do have unique genetic sequences, but our knowledge seems rather limited as far as exactly how the many differences affect the overall phenotypic function of the two species. Because of this limited knowledge, it is difficult to say whether or not such differences in genetic function can be achieved via random mutation and natural selection since there certainly are limits to the level of functional complexity that can be achieved by such "evolutionary" processes.

I am not a believer in design theory because of I have a problem with humans sharing a common ancestor with chimps.  This possibility really does not bother me at all.  What bothers me is that I know of various functions in different creatures that are separated by large neutral gaps in function.  Such neutral gaps cannot be crossed except via random drift.  This drift cannot be guided by the forces of natural selection toward any particular genetic sequence over any other particular genetic sequence - be it beneficially functional or not.  Such random drift simply takes too long to produce the various independent functions that we do observe in living things.  It is this problem that has convinced me of the truth of design theory and of the implausibility of common decent as an overriding explanation for the huge range in life forms that we see in the natural world.  A selection mechanism that is phenotypically based cannot direct neutral genotypic changes.  This is my problem with the theory of evolution.  I do not have a problem with the idea of the common decent of anything, but I do have a problem with the mechanisms that have been suggested as a driving force for such changes.

So, if such evolutionary "boundaries" exist as I think they do, how can they be delineated?  Granted, it is very difficult to know exactly where and how to draw genetic boundaries between various "kinds" or gene pools.  It seems that such boundaries do exist since there is a fair amount of evidence to support this hypothesis, but determining exact boundaries seems to be a difficulty in many cases owing to a lack of sufficient knowledge at this time.  

However, there are clues as to which groups might have had an easier time crossing the genetic distances that exist between them.  For one thing, those "kinds" that can breed and actually produce viable offspring are obviously very closely related as far as their gene pools are concerned.  Horses and donkeys can interbreed, while humans and chimps cannot.  For horses and donkeys, this ability to produce viable offspring seems to be very good evidence of common ancestry.   In other words, I believe that horses and donkeys share the same genetic information and/or genetic loci on their respective sets of chromosomes.  All the information appears to be the same, but the order of that information, as it is written on the chromosomal blueprints, is different.



The Same Information in Different Places


Looking a bit more into the genetics involved, horses have 32 pairs of chromosomes while donkeys have only 31 pairs.  When a male donkey and a female horse mate, they produce a mule (a male horse with a female donkey produces a different creature called a hinny).  A mule has 31-paired chromosomes as well as an extra chromosome from its horse mother that is not paired.  Many think that this extra non-paired chromosome is what makes the mule infertile.  The suggested reason for this is that during meiosis, the extra chromosome would not have a partner to match up with.  When the paired chromosomes are split apart into separate haploid gametes, the split would be uneven, creating sterility.  However, this is not an entirely satisfactory explanation for several reasons.  The most obvious reason that this hypothesis cannot be such an automatic assumption is the fact that other matings of odd-paired chromosomes do give rise to virile offspring.  For example, hybrids of the wild horse (33 pairs) and the domesticated horse (32 pairs) are fertile, and yet they have an odd chromosome out, just like mules do.  

Clearly then, something more than a difference in chromosome number is contributing to the species-interbreeding barrier.  If one thinks about it, it is not really all that hard to imagine how the pairing of odd numbers of chromosomes could be realized during meiosis so that fertility would not be lost. During meiosis all the chromosomes double themselves.  A cell undergoing gametogenesis in a diploid organism, such as horses and donkeys, will become tetraploid during the first part of meiosis.  At this point genetic recombination occurs between the four copies of a given chromosome.  After this, the four copies are eventually split apart to occupy four different cells called gametes.  Each gamete, containing only one copy of a given chromosome, is termed “haploid”.  However, if there is an odd chromosome out when a cell starts gametogenesis, this single chromosome with no sister chromosome can only double to make two chromosomes during meiosis instead of the usual four chromosome copies.  Still, all is not lost.  These two copies will also be sorted out into gametes.  But, since there are only half as many copies as "normal" diploid chromosomes, only half of the gametes will get the extra chromosome.  That means that 50% of the gametes will be viable, having the complete set of genetic information needed to code for that particular creature. 

So, since hybrid fertility is clearly possible for those having odd chromosome numbers, why are mules sterile?  Horse-donkey hybrids (mules and hinnies), show a "block" during testicular meiosis at the primary spermatocyte stage.  This is caused by incompatibility of synaptal pairing between paternal and maternal chromosomes - resulting in a total arrest of spermatogenesis.  But what exactly causes this block?

In order to understand this problem a bit more, it might be interesting to look into how interbreeding creatures can have variations in chromosome numbers and yet have the same information in those chromosomes.  As it turns out, many interbreeding creatures having different chromosome numbers have undergone chromosomal translocations.  What happens here is that a piece or "arm" of one chromosome becomes detached from its normal position on one particular chromosome and reattached to a different chromosome.  There are several different types of translocations and chromosomal fusions. Sometimes chromosomes can fuse with each other to form much longer chromosomes or they can split at the centromere to form two shorter chromosomes. One such rearrangement is known as a Robertsonian rearrangement and is the result of the fusion of two centromeres into one or the splitting of one centromere into two. A tandem fusion, on the other hand, is a fusion of two chromosomes in which one end of a chromosome fuses with the end or the centromere of another chromosome. 

Comparisons between the chromosomal banding patterns of these rearranged chromosomes show that the information is still the same, only rearranged a bit.  Moreover, it appears that certain types of rearrangements are quite group-specific for a certain type of creature.   One type of rearrangement that occurs in one type of animal does not necessarily occur in another type of animal.  The problem with this is that during meiosis the various portions of translocated chromosomes must still match up with their normal counterpart on different chromosomes during meiosis in order for genetic recombination to proceed without lethal mistakes.  Of course, one can only imagine that some types of complicated chromosomal rearrangements would make correct lineups impossible.  Obviously then, if a matching lineup is impossible, interbreeding impossible. 

However, with some kinds of chromosomal rearrangements, the various segments are still able to pair up and crossover even though there might be more than two centromeres involved.  For example, two chromosomes might only match up half way while the rest of each of these two chromosomes, might match a third chromosome (see visual illustration).6  This would make it possible for various chromosome numbers to maintain the same information and still give rise to viable as well as fertile offspring.  And, various interbreeding creatures do in fact maintain a wide range of possible chromosome numbers in their respective gene pools.  Besides the wild and domestic horse mentioned above, there are several other fairly well known examples. 

Of particular interest are domestic dogs and wolves of the genus canis.  They have 78 chromosomes while foxes have a varied number from 38-78 chromosomes. The uniformity of chromosome number in canid dogs can be due to free interbreeding over a wide range, whereas foxes live in small family groups and smaller territories so that new arrangements will persist. If the "kind" is penned at the level of the family Canidae, then the implications in terms of the number of animals required to produce the present varieties are not as daunting as many fear. Indeed, numerous chromosome homologies have been identified in animals today.  The differences between many species can often be explained with chromosomal rearrangements, as in the case of kangaroos, where Robertsonian fusions can account for much of the variation that exists between the different kangaroo species.



Types of Chromosomal Alterations





When complex eukaryotes like humans replicate, specialized cells are created ("gametes", known as sperm and ovae) that have half as many chromosomes as the rest the cells. Instead of having two of each chromosome, these gametes only have one of each chromosome. Gametes from male and female partners fuse and recombine their single chromosomes into pairs of chromosomes again where half of each pair came from a sperm or male gamete and the other half from an ovum or female gamete. On very rare occasions, when new chromosomes are being made, the new copies will fail to separate (or "disjoin") from the original copy. When this happens, one gamete will get an extra copy of the chromosome that failed to disjoin, and the other gamete will get no copy. This is called a non-disjunction.  In such a case if a sperm that retained two copies of a given chromosome fused with an ovum that had one copy of that chromosome the resulting cell would have three copies of that chromosome instead of the normal two copies. Usually, the gain of an extra copy of a chromosome is fatal, and the cell dies, but not always.

Another way to increase chromosome number is by duplicating all of the chromosomes in a sort of whole genome non-disjunction resulting in polyploidy. This seems to work alright for plants, but for animals, this is very rare.  However, there are many examples of tetraploid (4n) and hexaploid (6n) insects, triploid and tetraploid fish (Catfish, Cyprinids and Carp), and even a hexaploid fish (a subspecies of the tetraploid Carp). There are also several tetraploid species of claw-toed frogs, and even a tetraploid rat (the red viscacha rat, or Tympanoctomys barrerae).  In the case of this rat species, the number of chromosomes (102) is much lower than one might expect by looking at closely related rat species. Based on these relatives, one would expect the tetraploid rat to have 112 chromosomes, which brings us to the other mechanisms for changing chromosome number.




Robertsonian Fusion:


Robertsonian fusion changes the chromosome number, but not the arm number. When chromosomes line up during meiosis I, a metacentric chromosome lines up with two acrocentric chromosomes (see illustration). For example, the house mouse Mus Musculis has 40 chromosomes, and a population of mice form the Italian Alps was found to have only 22 chromosomes. This population differs slightly from the normal house mouse in morphology as well and is classified as a different species named Mus poschiavanus. Other populations have been discovered with chromosome numbers varying between 22 and 40. For example, over 40 Robertsonian "races" of Mus musculus domesticus have been found in Europe and North Africa.1  The number of chromosome arms are the same and banding studies reveal the genes to be homologous. Obviously, in terms of their relationship, these different species are all one group. 

Another example of a Robertsonian polymorphism is the house musk shrew that lives in the central region of West Malaysia and has a variation in chromosome numbers from 36 to 40.  Also, in Southern India and Sri Lanka musk shrews can be found with having between 30 and 32 chromosomes due to Robertsonian-type changes.2

A new Robertsonian translocation has been found in cattle. A bull from Marchigiana breed (central Italy) was found to be a heterozygous carrier of a centric fusion translocation involving chromosomes 13 and 19 according to RBA-banding and cattle standard nomenclatures. CBC-banding revealed the dicentric nature of this new translocation, underlining the recent origin of this fusion. In fact, both the bull's parents and relatives had normal karyotypes. In vitro fertilization tests were also performed in the bull carrying the new translocation, in two bulls with normal karyotypes (control) and in four other bulls carrying four different translocations.3

The significance of centric fusions (Robertsonian translocations) in domestic animals, with special reference to sheep, is also interesting. A study was done that considered 856 ewes with either a normal chromosome number 2n = 54 or carrying one or more of the three different translocations (centric fusions) t1, t2 and t3 in various heterozygous and homozygous arrangements. Rams, which were used in the matings, were homozygous for one of the translocation chromosomes (2n = 52), double heterozygotes (2n = 52), triple heterozygotes (2n = 51), or were carriers of four translocation chromosomes (2n = 50) and five translocation chromosomes (2n = 49). A remarkably even distribution of segregation products was recorded in the progeny of all combinations of translocation ewes plus translocation rams in those groups in which sufficient animals were available for statistical analysis. Forty-eight chromosomally different groups of animals were mated. Further, the overall fertility of the translocation sheep (measured by conception rate to first service via the lambing percentage and number of ewes that did not breed a lamb) was not significantly different from the New Zealand national sheep breeding data where this study was done. In some groups the poorer reproductive performance could be explained by the age structure of the flock and inbreeding depression, which probably affected the performance of some animals. Sheep with progressively decreasing chromosome numbers, due to centric fusion, 2n = 50, 2n = 49 and 2n = 48, were reported. The 2n = 48 category represents a triple homozygous ewe and a triple homozygous ram and is the first report of the viable evolution of such domestic animals. Less than 1% of phenotypically abnormal lambs were recorded in a total of 1,995 progeny born over 10 years. It is now considered that there is little or no evidence to suggest that centric fusions in a variety of combinations affect the total productive fitness of domestic sheep. It is suggested that future research should be more actively directed to understanding their genetic significance.4

A detailed investigation of testicular meiosis in a mule, a hinny and a Przewalski horse/domestic horse hybrid were made. Abnormalities of pairing were observed in the mule and hinny in most germ cells at the pachytene stage of meiotic prophase, and spermatogenesis was almost totally arrested. A few mature spermatozoa were recovered from the ejaculate and epididymal flushings of the hinny. The Przewalski horse/domestic horse hybrid was fertile and showed normal spermatogenesis. Chromosome banding studies showed a close homology between the karyotypes of the Prezwalski horse (Equus przewalskii, 2n = 66) and the domestic horse (E. caballus, 2n = 64), and it is evident that a single Robertsonian translocation occurred that transformed four acrocentric chromosomes of E. przewalskii into two metacentric chromosomes in E. caballus. The investigations showed that a trivalent is formed at meiosis in the hybrid (2n = 65), segregation which gives two classes of genetically balanced spermatozoa. Both of these are capable of producing normal offspring if they fertilize the eggs of a domestic mare.5



Tandem Fusion:


Tandem fusion changes arm number and chromosome number. Tandem fusions have been found in some antelope species where a sex chromosome is fused with an autosome. Although rare, one can assume that the organisms probably had a common forerunner. The antelope displaying this fusion range in size from the eland (the largest of all the antelopes) to smaller species such as the sitatunge and the bushbuck. However, they all share common features, such as similar shapes of the horns and stripes on the body which may be prominent as in the case of the Bongo or less prominent as in the case of the eland. Species with this type of fusion are: the eland, bongo, lesser and greater kudu, bushbuck, sitatunge and nilgai (Indian antelope) where the y-chromosome is fused to an autosome. Tandem fusions are also found in Malaysian swamp buffalo and Asian river buffalo. Another very interesting example of this type of fusion is also found in the Asian deer. In the species Muntiacus muntjac, the females have only 6 chromosomes and the males have 7 chromosomes (this is the smallest chromosome number in mammals). However, in a different "species" of the group, Muntiacus reevesi, both the males and the females have 46 chromosomes.  However, banding studies show that the same genetic material is present in both species, the chromosomes in M. muntjac are just fused together to form very long chromosomes. Once again no new information is added, it is just reshuffled, thus providing differential expressions and increased variety.





A translocation is what happens when two chromosomes that are not part of a pair get stuck together as if they were a pair, and exchange segments. If the segments that get exchanged are large enough most of both chromosomes can move onto one single chromosome.  In this figure, the blue and red figures are cartoons of two different chromosomes. The indentions in their centers represent the "centromeres" where the cell attaches filaments that drag the chromosomes around. When these chromosomes cross-over, the result is two chromosomes of very different sizes. One larger chromosome, contains almost all of the genetic material of the two chromosomes, while the other smaller chromosome contains almost none. In this example, the material from one chromosome has been moved or "translocated" onto another chromosome. In dramatic cases where one chromosome is significantly smaller after the translocation than the other chromosome the smaller chromosome may be lost resulting in a reduced chromosome count.

 Translocations can also lead to reduced fertility.  In humans a well-known example of a translocation of part of chromosome 21 results in Down Syndrome.  Humans with such translocations generally experience many problems as well as reduced fertility depending upon when the translocation occurred and how much genetic information was involved.  However, in some insects and plants that have meiotic drive, healthy, viable, and even fertile offspring can be produced via such translocations.6



Pericentric Inversions:


These provide changes in arm number but not chromosome number. The number of arms depends on the position of the centromere. If it is located at the end, then there is one arm and if in the middle there are two arms. The inversion can change acrocentric chromosomes to metacentric chromosomes. The rodents Neotoma and Peromyscus differ by this inversion.



Paracentric Inversion:


In this type of inversion the centromere is not included. This inversion is relatively uncommon, but has been proposed for some bats, hares and apes.2



Drastic Rearrangements:




Under certain circumstances of severe environmental stress, drastic rearrangements can produce greater varieties, which could enhance survival. These changes can be rapid when new adaptive zones are entered (canalization model). Such rearrangements have been proposed for the Spalax mole rat.2





So, the potential problems with chromosomal rearrangements are obvious.  There can be multiple translocations involving the same chromosome as well as chromosomal inversions and a number of other interesting cuttings and splicings of chromosomes.  These can result in a difficult or even impossible situation where chromosomes in such involved hybrids simply cannot match up properly during meiosis.  This lack of matching capability results in sterility.  For example, if a piece of donkey chromosome is inverted relative to its counterpart in horses, then gene-by-gene pairing cannot occur without elaborate looping and twisting. The chance of a successful cell division is very much reduced. So, the mule cannot make egg or sperm cells. Thus, the mule is sterile.  However, a mule can still be born healthy because such growth of non-meiotic cells occurs via a different process called mitosis.  Mitosis does not have to match things up; it only has to make copies. So, inversions and translocations do not prevent the mule from growing up to be an adult.7

Rearrangements can and do account for differences in insectivores, bats, primates, marine mammals, rodents, rabbits and hares, ungulates etc. The potential for change certainly exists.  Nevertheless, there are certain barriers that cannot be transgressed.  Such changes, as occur with reshuffling of genetic information, do not add to that information.  The information itself is still the same.  True, there are a few documented cases where mutations do actually result in unique or novel functions that are beneficial to the host, such as the galactosidase evolution experiments performed by B.G. Hall, the nylonase evolution experiments performed by Kinoshita, et. al. and of course the all to familiar everyday examples of antibiotic resistance by bacteria.  Surprisingly though, these very examples of evolution in action are the ones that make the boundaries of evolutionary processes most clearly delineated. 



1.      Hauffe HC, Searle JB, Chromosomal heterozygosity and fertility in house mice (Mus musculus domesticus) from Northern Italy. Department of Zoology, University of Oxford, Oxford OX1 3PS, United Kingdom.

2.      Sharma et al. 1969;. Aswathanarayana and Prakash 1976; Yosida 1982, Ishikawa et al. 1989, Young, 1970, 1971, 1974; Sam et al., 1979 (

3.      Molteni L, De Giovanni-Macchi A, Succi G, Cremonesi F, Stacchezzini S, Di Meo GP, Iannuzzi L. A new centric fusion translocation in cattle: rob (13;19). Hereditas 1998;129(2):177-80  Institute of Animal Husbandry, Faculty of Agricultural Science, Milan, Italy. (

4.      Bruere AN, Ellis PM., Cytogenetics and reproduction of sheep with multiple centric fusions (Robertsonian translocations).  J Reprod Fertil 1979 Nov;57(2):363-75. (

5.      Chandley AC, Short RV, Allen WR., Cytogenetic studies of three equine hybrids.  J Reprod Fertil Suppl 1975 Oct;(23):356-70. (








. Home Page                                                                           . Truth, the Scientific Method, and Evolution   

. Methinks it is Like a Weasel                                                 . The Cat and the Hat - The Evolution of Code   

. Maquiziliducks - The Language of Evolution             . Defining Evolution    

. The God of the Gaps                                                           . Rube Goldberg Machines  

. Evolving the Irreducible                                                     . Gregor Mendel  

. Natural Selection                                                                  . Computer Evolution  

. Which Came First                                                                 . Antibiotic Resistance  

. The Immune System                                                            . Pseudogenes  

. Genetic Phylogeny                                                                . Fossilized DNA  

. DNA Mutation Rates                                                            . Donkeys, Horses, Mules and Evolution  

. The Fossil Record                                                                . The Geologic Column  

.  Early Man                                                                                . The Human Eye  

. Carbon 14                                                                              . Radiometric Dating  

.  Quotes from Scientists                                                           . Ancient Ice

 . Meaningful Information                                                           . Limited Evolutionary Potential

 . J Harlen Bretz                                                                               . The Flagellum






Stacking the Deck


God of the Gaps


The Density of Beneficial Functions


All Functions are Irreducibly Complex


Ladder of Complexity


Chaos and Complexity


Confusing Chaos with Complexity


Evolving Bacteria


Irreducible Complexity


Scientific Theory of Intelligent Design


A Circle Within a Circle


Crop Circles


Mindless vs. Mindful


Single Protein Functions


BCR/ABL Chimeric Protein


Function Flexibility


The Limits of Functional Flexibility


Functions based on Deregulation


Neandertal DNA


Human/Chimp phylogenies




The Geologic Column


Fish Fossils


Matters of Faith


Evidence of Things Unseen


The Two Impossible Options




Links to Design, Creation, and Evolution Websites






    Search this site or the web powered by FreeFind
   Site search Web search



Since June 1, 2002





AddFreeStats AddFreeStats AddFreeStats