(This article was originally published in "Cichlids Yearbook Volume 3, Cichid press; 1993; pp. 24-27. It is here reproduced with the permission of Ad Konings, Cichlid Press).
In August 1992 Sturmbauer & Meyer published an article in which they claimed to have deduced a phylogenetic tree (a kind of pedigree of populations and species) for Tropheus by comparing the sequence of two small pieces of the cichlids' DNA. Although it has received considerable attention in the scientific as well as in the aquaristic press, opinions regarding the acceptability of their findings differ. Persons who have ample experience with the techniques involved largely agree with the authors; persons who are well versed with the cichlids in question find it difficult to accept their ideas. Before I try to explain their results and give my personal ideas about their usefulness, some terms must be defined.
A phylogenetic tree is a graphic representation of the descent of populations, species and groups of species. It shows which of these are closely related and how they evolved. A phylogenetic tree shows the evolution of related populations and species from a single ancestor.
Speciation is the evolution of new species and may take place in various ways (Wiley, 1981). A species is rather difficult to define because it is a natural group of individuals which recognise each other as belonging to this group. This definition is useless to us because we don't know whether an individual from population A recognises an individual from population B as being conspecific or not. The species-specific recognition is expressed if and when individuals from population A, under natural circumstances, mate with individuals from B. If interbreeding does not occur we are unable to tell with certainty whether A and B are conspecific or not. When an apparent species consists of several populations which are geographically separated then the assignment of two or more of these populations to one species is the personal opinion of the author. DNA is the abbreviation for deoxyribonucleic acid. The DNA molecule is found mainly in the nucleus of a cell and is responsible for the transmission of hereditary characteristics.
The nucleus of a cell contains chromosomes, each of which consists of a single DNA molecule wrapped up in a coat of proteins. Almost all vertebrate organisms have a double set of such chromosomes, one received from the mother and the other from the father. Although the DNA molecules are extremely complex they are built from only four different building blocks. It is the sequence of these blocks (bases) which is very important.
A gene is a section of the DNA molecule whose sequence, after decoding, is responsible for a protein. A change to a single building block in the sequence, a mutation, may completely abolish the production of the protein; or it may produce a different protein. Most individual mutations, however, have no or only minor effects on the final product. Mutations take place regularly. They are caused mainly by (cosmic) radiation and chemical agents.
Mutations can be present also in the DNA of germ cells (spermatozoid and ovum) and are thus transferred to offspring. These mutations are then added to the gene-pool of the species. A gene-pool is the total of all the DNA sequences of all the sexually reproducing individuals of an interbreeding population. Mutations occur continually and when organisms no longer participate in the gene-pool from which they were once derived their DNA sequences gradually deviate from those of the mother gene-pool. It thus follows that not only species but also geographically isolated populations of a single species will have different gene-pools. The variability of a particular piece of DNA may be greater between populations of one species than between two sympatric species!
It is self-evident that any mutation which alters a protein so that the organism cannot sustain itself in the community will be selected against; for instance, when a mutation would change the colour of a cichlid from yellow to blue. The blue fish may no longer be recognised by its conspecifics and therefore be excluded from reproduction. This means that its genes are lost from the gene-pool. Thus although mutations occur (almost) randomly in the DNA they do not all become fixed in the gene-pool of the species. There is selection against certain mutations in certain genes. It must be stressed that not all mutations in a gene produce an altered protein and even if a protein is altered there is not necessarily any selection against it (there are several polymorphic populations known among cichlids). Using certain techniques, termed DNA sequencing, the sequence of a piece of DNA can be resolved into its individual building blocks. The corresponding piece of DNA from other species can likewise be sequenced and compared. It is erroneous to conclude that differences in the DNA sequences found during such comparisons are derived solely from the fact that DNAs from different species are being compared: speciation and the continuously occurring mutations are two different processes. Speciation could not occur without genetic variability (caused by mutations) but mutation does not necessarily result in speciation. One million mutations might not make a new species whereas a single mutation in the right gene might.
Sturmbauer & Meyer (1992) deduced their phylogenetic tree from the number of mutations in a small section of a gene of the various Tropheus populations compared with that of Oreochromis tanganicae. (They sequenced a part of the cytochrome b gene and a control region on the mitochondrial DNA.) As has been discussed above the number of mutations says little about speciation (because it is not a continuous process). However, owing to the fact that mutations occur continually, it may indicate the age of a particular gene. The more differences one finds between two gene-pools the longer these pools have been separated. Sturmbauer & Meyer remarked: "The morphology of 'living fossils', like horse-shoe crabs, has remained essentially unchanged for millions of years, although these organisms exhibit normal levels of molecular evolution." These authors thus note that mutations (molecular evolution) do occur but that fhe horse-shoe crab still looks like it did a hundred million years ago. And, as I would understand their remark, the present-day horse-shoe crab is still the same species as it was hundred million years ago. DNA sequencers compare numbers of mutations and thus look at the length of the period during which different populations (gene-pools) have been isolated.
On the basis of their sequencing data Sturmbauer & Meyer cannot realistically say much about speciation among Tropheus and thus cannot give a true phylogenetic tree. However, their investigations do suggest that the populations of the Tropheus species are likely to be old, much older than the species in Lake Malawi and Lake Victoria.
In my opinion the discovery of the antiquity of the Tropheus species is very interesting and deserves more attention than the proposed new classification based on the number of DNA mutations.
There are a few other points in the publication that need to be discussed. First is the material used in the comparisons. Owing to the fact that it is virtually impossible to extract DNA from formalin-fixed fishes DNA sequencers need fresh material or fishes that have been fixed in alcohol. Sturmbauer & Meyer obtained most of their fresh material from Laif DeMason (Old World Exotic Fish in Florida) who, in turn, imported it from several stations in Africa. Among the fishes used there were also some bred in ponds in Burundi which are thus derived from a limited number of females. When comparing individuals from the same pond one expects to find very little difference between the DNAs. This is likely to be true in the case of the group of 10 individuals of the Bemba variant eight of which were identical. The other group used in the comparisons stemmed from Mpulungu and consisted of wild caught specimens (DeMason, pers. comm.). The natural variation among these ten fishes was demonstrated by the finding that eight out of ten individuals appeared to be different for the gene tested.
The part of the paper that really disturbed me was the assumption that the "Kirschfleck" or "Double Spot Moorii" was the species most closely related to T. polli! I regard, on grounds of coloration, morphology, distribution, and behaviour, T. polli as a population of T. annectens but Sturmbauer & Meyer found more mutations between these two populations than between T. polli and the "Kirschfleck." As mentioned before, the number of mutations has little effect on speciation, as is nicely demonstrated by these comparisons. What the results may suggest is that the sections of DNA tested of the "Kirschfleck Moorii" and T. polli are of similar antiquity but not that these species are closely related (within Tropheus). Personally I feel that the composition of a small fraction of DNA from the two samples of each species used should not be given the importance inferred by Sturmbauer & Meyer. Not only the "Kirschfleck" and T. polli were regarded as closely related (within the genus) but also the "Bemba" or "Orange Moorii" and T. brichardi from Nyanza! T. brichardi from Kavalla, according to these authors, finds its most closely related populations at Mpulungu, Kala, and Kasanga! For DNA scientists this may sound irrelevant but for aquarists well versed with these species it sounds ridiculous.
The crux of the matter is that the variability of a DNA section can be greater in two populations of a single species than in two sympatric species. It is not the number of mutations that make a species but where in the genes such mutations have occurred.
Results of Sturmbauer & Meyer DNA analysis 1992
T. annectens from Zaire
T. annectens from Bulu point
According to Sturmbauer & Meyer, the population of T. annectens from Bulu point, Tanzania (Right) is more closely related to the "Kirschfleck Moorii" (Left) than to T. annectens from Zaire (middle)!
Tropheus sp. "Black" from Bemba
Tropheus sp. "Black" from Kiriza
T. brichardi from Nyanza-Lac
According to Sturmbauer & Meyer, the raze of Tropheus sp. "Black" from Bemba (Left) is more closely related to T. brichardi from Nyanza-Lac (The choco moorii) (right) than to Tropheus sp. "Black" from Kiriza (center)!
Distribution of Tropheus in Lake Tanganika
I thank Dr. Irv Kornfield, who, although not sharing my views, critically read the manuscript and gave valuable comments.