Cichlidophiles: Discus Wars: How Many Species Are There?
Author: Wayne S. Leibel
In a recent article that appeared in the December 2006 issue of TFH, I asked the question “How Many Discus Species Are There?” The answer, according to fish collector/explorer Heiko Bleher in his new discus book (Bleher’s Discus Volume 1, 2006, Aquapress, Italy), which appeared this past summer, and which I reviewed in that article, was three—Symphysodon discus, S. aequifasciatus, and S. haraldi. Just when you thought it was safe to go back into the discus tank, however, recent developments suggest otherwise.
In a recently published scientific study out of noted South American ichthyologist Sven O. Kullander’s lab in Sweden (Ready, J. S., E. J. G. Ferreira and S. O. Kullander. 2006. Discus fishes: mitochondrial DNA evidence for a phylogeographic barrier in the Amazonian genus Symphysodon (Teleostei: Cichlidae). Journal of Fish Biology 69 (Supplement B): 200-211), data are presented that confirm that the answer still is three. However, only two of the three species recognized by Bleher are supported (S. discusHeckel, 1840 and S. aequifasciatus Pellegrin, 1904, but not S. haraldi Schultz, 1960) and a third, new species of discus, S. tarzoo, is described.
A Quick Recap
To summarize Bleher’s (2006) proposed three species, he recognizes the so-called “Heckel discus” Symphysodon discus Heckel, 1840, which is distinguished from the others principally on the basis of their unique black vertical barring pattern in which the first, fifth, and ninth of nine vertical bars are wider and darker than the remaining six. They also sport a conspicuous body striping pattern of parallel, longitudinal iridescent vermiform stripes.
Bleher also recognizes the green discus: Symphysodon aequifasciatus Pellegrin, 1904, again distinguished principally by their distinctive color pattern which, in addition to the nine black vertical bars of equal intensity (aequifasciatus means “equal bars”), consists of small red dots of varying number on the body. These sometimes occur over the entire body and may include red dotted lines, and almost always occur on the anal fin region.
He also recognizes as a third distinct species, the blue discus Symphysodon haraldi Schultz, 1960, a lumping of “brown” and “blue” discus types. This species has the typical aequifasciatus equal vertical black stripes but otherwise is highly variable in body coloration from totally brown to all-over blue parallel striping (in the best specimens, not unlike that of the Heckel discus).
In an attempt to resolve the systematic confusion surrounding discus species and coloration morphs, Bleher collected discus widely throughout the Amazon, photographed them alive, and took fin clips (a good source of DNA when gently preserved in ethanol) from them and returned these to Germany for genetic evaluation by the well-respected fish molecular systematics team of Dr. Axel Meyer and associates. While the paper is, according to Bleher’s book, currently “in press” (Stolting, K., Salzburger, W., Bleher, H. and A. Meyer, 2006. Preliminary Revision of the Genus Symphysodon Heckel. Aqua, Journal of Ichthyology and Aquatic Biology: Special Publication), his recent presentation at the July 2006 American Cichlid Association convention of genetic haplotype maps using various DNA markers seemed to indicate three major genetic groupings corresponding to the three morphologically distinct species he is recognizing/promoting. All, scientists and hobbyist alike, eagerly await publication of these results.
In the meantime, Kullander’s group (Ready et al., 2006) has published their results based on a similarly extensive (though not as exhaustive) sampling of discus from along their near-linear Amazonian floodplain distribution. Discus are restricted to areas where seasonal flooding occurs, therefore they are found only near the main stem of the Amazon River itself and in the lower reaches of its tributaries on the Amazonian floodplain. In this study, discus were sampled widely from east to west (Brazil through Peru), both from the Amazon proper and nearby tributaries at locations some of which may be familiar to discus enthusiasts: (from east to west) JuÁ, Boim, Alenquer, MauÉs, Abacaxis, Madeira, Manaus, Rio Bauana, TefÉ, Jrua, JutaÍ, and Ica. Many of these populations that have unique color patterns have previously been described as unique sub-species, or even species, e.g., S. aequifasciatusaequifasciatus Schultz, 1960 (green); S. a. haraldi Schultz,1960 (blue); S. a. axelrodi Schultz, 1960 (brown); S. discus willischwartzi Burgess, 1981; and S. discus tarzoo Lyon, 1959.
As in Bleher’s pending study, newly collected discus were photographed alive and tissue samples taken from them for later DNA analysis. Molecular data (DNA sequencing) were obtained from 23 individuals, qualitative data on color patterns and body shape characteristics were determined from photos of 263 live fish, and morphological characteristics—fin and scale counts (meristics) and rhomboid body measurements—were taken for 94 fish.
Color pattern characters that were scored included: anal fin red color (absent to much); extension of anal fin color into body (absent to extended); anal fin pattern (none, red spots or small streaks on blue base, thin red lines on blue base with lines thinner than base, thick red lines on blue base of equal width, thick red lines reticulating over blue base); body base color (light brown, brown, green/brown, gray green); forehead marking (absent, few blue stripes, moderate stripes, many stripes continuing to dorsal fin); and red spots on body (absent/present). Body shape was scored with respect to forehead shape (flat slope, slightly rounded, mainly rounded, completely rounded) and overall body shape (elongate, intermediate, discoid).
Analysis of Data
Before this most recent study, Kullander (2003. Family Cichlidae. In: Checklist of Freshwater Fishes of South and Central America, Reis et al. editors, Porto Alegre) only recognized as valid the two aforementioned species which are easily discriminated based on color/barring pattern: S. discus, found in the central Amazon Basin in the Rio Negro, Abacaxis, and Trombetas drainages and S. aequifasciatus, found along almost the entire length of the Amazon from 49 to 70 west longitudes.
In this two-species scheme S. aequifasciatus showed great variation. While the general color of S. aequifasciatus is brown with some blue and red markings on the forehead and near the anal fin, some individuals in all populations have brighter colors including greater coverage or strength of red and blue stripes, spots, or reticulations, and some populations differed from others in coloration. So one of the goals of this study was to see if consistent coloration differences could suggest that several distinct species comprised nominal S. aequifasciatus.
Testing these qualitative data (color pattern and body shape) statistically with a mathematical approach known as Principal Component Analysis (PCA) placed specimens from the same population in the same cluster, but didn’t really meaningfully discriminate between S. aequifasciatus populations. Likewise, PCA testing of meristics and rhomboid measurements failed to distinguish populations: the within-population variation was greater than the between-population variation.
However, the main western (TefÉ and west) and central/eastern groupings of S. aequifasciatus were largely separable from each other by the presence or absence of red spots and for anal fin red coloration. The darker-bodied fish from central and eastern populations always lacked red-pigmented spots on the anal fin and body and were distinct from the western fish. Two sub-groups appeared to be present in the central/eastern cluster too, though the clustering was not strong enough to support a distinct far-eastern lineage or species. Remember that S. discus and S. aequifasciatus, with overlapping, largely non-defining meristics, are separated from each other principally on the single morphological character of vertical black barring.
With respect to the molecular data, the DNA extracted from 23 individuals (including nominal S. discusand S. aequifasciatus from along its extensive distribution) was sequenced with respect to two genes/gene segments: mitochondrial cytochrome b and nuclear rhodopsin.
Cytochrome b is an important protein in the energy-generating respiratory pathway of mitochondria, the powerhouses of the cell. Mitochondria have their own DNA which codes for many, but not all, of the mitochondrial proteins. Here, 1134 nucleotide bases were sequenced, and the resulting sequences from all 23 discus sampled were aligned and compared with each other in pairwise fashion.
Rhodopsin is one of the vertebrate visual pigments (proteins) important for color vision and is coded for by a nuclear gene (i.e., found on the chromosomes of the cell’s nucleus). Here, a 514-base-pair segment of the gene was sequenced from each of the same 23 sampled discus.
The strategy of DNA sequencing, in a nutshell, is to see how much sequence divergence—change in the nucleotide base sequence order—there is between individuals. No or little difference in sequence indicates close relatedness or even genetic identity (same species and/or interbreeding populations), whereas moderate to much difference indicates separation, genetic divergence, and ultimately speciation.
So, one would expect that morphologically distinct species would have non-identical DNA sequences and the degree of sequence change between them to be indicative of how long ago they became genetically isolated from each other, either due to physical separation leading to genetic divergence or due to their inability to interbreed (genetic or other incompatibility). Reproductive isolation and interrupted gene flow are signatures of speciation and can indicate distinct species even in the absence of recognizable (to us) morphological differences.
The DNA sequence data were puzzling. Rhodopsin showed no variation between the 23 sampled individuals, which included S. discus and S. aequifasciatus, the latter from all along its extensive east-west range. This high level of sequence conservation (i.e., no differences at all) indicates that either rhodopsin is a very important protein and mutations (base changes) are not tolerated (are eliminated by natural selection), or the divergence of the morphological species has occurred relatively recently, too recently for mutational base changes to have accumulated in this gene.
The cytochrome b sequence data were also unexpected. Ready et al. found sequence differences that sorted all discus into two distinct genetic lineages. One of these cytochrome b lineages was found in samples from the western populations that would be formerly morphologically identified as S. aequifasciatus. The second and larger lineage included cytochrome b sequences from central and eastern S. aequifasciatus but also included S. discus from Manaus and from the Rio Abacaxis. One S. aequifasciatus individual from the Madeira River, at the midpoint of the geographic distribution of sampled discus, had the phenotype (overall appearance) and cytochrome b sequence (haplotype) typical of western populations.
The sequence data also suggested that a genetically distinct far-eastern lineage may also exist, however these data comprise only two samples that were collected from sites. It would seem that while looking dramatically different, S. aequifasciatus and S. discus are genetically indistinguishable at the mitochondrial cytochrome b gene, whereas the western S. aequifasciatus, which can be separated and recognized easily from central/eastern S. aequifasciatus on the basis of their red spotting, are significantly different in cytochrome b sequence.
A Different Third Species
The DNA data along with the reliability of the red-spotting character suggest strongly that the western populations of S. aequifasciatus are reproductively isolated and genetically divergent enough for them to be described as a distinct species, which Ready et al. choose to do in this paper. It turns out that these discus had been described as a distinct subspecies earlier by Lyons in an aquarium hobby magazine in 1960 as Symphysodon discus tarzoo (Tropicals Magazine 4 [3; “Holiday Issue”]: Cover, 6–8, 10).
According to this article, Al Wechsler, owner of Suwannee Water Gardens in Chicago, arranged for the export of these discus. He hired a pilot who flew directly from Chicago to Leticia, Colombia to pick them up. The name “tarzoo” is in recognition of the Tarpon Zoo, in Tarpon Springs, Florida, whose co-owner Mike Tsalickis was the contact person living in Leticia who apparently had located and arranged for the collection of these discus. Lyons gave them the “unofficial” subspecies name of S. discus tarzoo, but the verbal description and photos accompanying the article are clearly sufficient to make him the author of these species and the species name tarzoo a valid one. Regrettably no type specimens were preserved, although the Chicago Natural History Museum was given six with this, and a formal description, in mind.
Whereas other ichthyologists (e.g., Schultz 1960) found this description insufficient, apparently Kullander (1996) does not, and now names (Ready et al. 2006) and re-describes the western population of discus with red spots on the anal fin and in the body as Symphysodon tarzoo, elevating it from subspecific rank and distinguishable from the other two species of discus, S. aequifasciatus and S. discus, which have reticulations (a network of lines or stripes) and not spots. In so doing, they have designated a type specimen (neotype) which had already been deposited and recorded (as S. aequifasciatus) at INPA in Brazil.
Symphysodon tarzoo and S. aequifasciatus overlap in distribution only slightly, as represented by a few individuals of S. tarzoo collected along with S. aequifasciatus from the Rio Madeira. Ready et al. (2006) also declare S. haraldi (Bleher’s third species) and S. axelrodi as synonyms of S. aequifasciatus.
Is your head hurting yet? I applaud you if you have stuck with me until now. There is much more interesting biology to report about these discus, namely how and why speciation may have occurred and why there apparently are so few species of discus, and these are also addressed in the Ready et al. (2006) paper. But I will defer this discussion to my next column (April). Meanwhile, the Kullander group concludes there are three species of discus: S. discus (“Heckel”), S. aequifasciatus (“brown and blue”), and S. tarzoo (“green, spotted”), whereas Bleher supports S. discus (“Heckel”), S. aequifasciatus (in his scheme, this is the “red-spotted green,” now tarzoo), and S. haraldi (“brown and blue”).
Pending publication of the aforementioned and eagerly awaited Bleher/Meyer’s genetic marker (haplotype) data, the discus wars continue—may the best science win!
Special thanks to my bibliophilic friend and co-columnist Lee Finley who supplied the article and to my book collector friend Tom Tilles who graciously sent me the copy of Tropicals in which the “Blue Tarzoo” were announced.