Variation in Taxonomic and Systematic Characters

There are three major types of character variation within and between species that is typically observed in systematic and taxonomic studies. These include:



Variation in one or more characteristics over space.



Variation in one or more traits between or within a sex.



Variation in one or more characteristics within the lifetime of an individual organism.

There may also be variation observed within the lifetime of an individual or within or between demes, populations, and species that is not under genetic control and thus would not be considered heritable variation. However, this type of variation can cause many problems in some studies where the life cycles or histories of organisms are unknown or where environmental "factors influence" character variation.

Regardless of the "type" of variation that may be encountered there is no substitute for having experience with the group of organisms that you wish to study. Through this experience you gain valuable knowledge as to what type of variation observed in an individual organism is important (relative to the question at hand) and what variation is not important.


1. Geographic Variation

As implied by the name this type of variation occurs over geographic space. This includes, but is not limited to, latitudinal, longitudinal, and altitudinal variation of characters.

Usually, researchers examine variation in demes, populations, and species and look for any "geographic" correlation with any of this variation. Within logical boundaries, if a "geographic" component to the variation can be identified then one may hypothesize that the observed variation within the taxon may be clinal and may be correlated with an environmental component. If so, such a character may have a selective advantage in some geographic areas or, developmentally, characters may exist in some geographic areas because of temperature, light, or other physical influences early in ontogeny.

Alternatively, in some instances, what may appear to be clinal variation in a taxon may be the remnants of an ancestral cline or may be clinal variation in characteristics that are not involved in species identification.

Below, are a few examples of geographic variation within some species. Other examples are provided in the previous lecture regarding different types of characters. Finally, there is no substitute to reading the literature of the groups of organisms that you wish to study (as well as other groups) to gain insight into character observation and interpretation and analysis and interpretation of variation observed.

Observed variation in breast squamation observed across the geographic range of the banded darter, Etheostoma zonale.

Observed variation in cheek squamation observed across the geographic range of the banded darter, Etheostoma zonale.

Percent composition of the various types of breast squamation (see above) in populations across the range of the banded darter, Etheostoma zonale. (Note that those samples from Mississippi and western Tennessee are now referred to as the brighteye darter, Etheostoma lynceum.)

Distribution of the various subspecies, races, and intergrade populations of the banded darter, Etheostoma zonale. (Note: Populations Y, H, P, and R all represent the brighteye darter, Etheostoma lynceum.)

Example of Geographic Variation in the Number of Lateral Line Scales in Populations of the Banded darter, Etheostoma zonale.


2. Sexual Variation

As implied by the name of this type of variation males and females frequently vary from one another for characteristics. This is most often known as sexual dimorphism and the characteristics that do differ are often referred to as primary or secondary sexual features. These latter terms will be discussed below.

Sexually dimorphic characters may not vary in larval and juvenile individuals; variation may be obtained only later in life. Adults may display these characters throughout the year or may only possess them during the breeding season. In the former case, usually the dimorphic characters are most extreme during the breeding season. Some species do not vary in secondary sexual features and sex is determined only by examination of gonadal tissue.

Aside from the obvious differences that will be observed in the gonads between sexes there seems to be an unlimited variety of different types of characters that differ between sexes. In vertebrates most sexual dimorphism is displayed by the males being more brightly colored relative to females; however, this is reversed in Phalaropes.


Coloration in Fishes


Breeding Tubercles of Fishes


Second dorsal fins of breeding males of fishes in the subgenus Catonotus, genus Etheostoma.


Swollen head of Nocomis micropogon during breeding season.

3. Individual Variation

Morphological Variation

I. Age variations

Common in many groups of organisms to have different looking juveniles or larvae from adults. Many synonyms have resulted from this phenomenon.

For example, Linnaeus described the immature striped goshawk as a different species from adult. This was obviously the result of molting pattern differences.

In some groups of fishes some have immature forms that are so different that they have been placed in separate genera or families. The lamprey larval forms referred to as ammocetes were at one time placed in the genus Ammocetes; or fishes of the order Elopiformes (tarpon, eels) have a leptocephalus larval form and these have been placed in the genus Leptocephalus. This can be a real problem in groups where larvae do not look anything like adults (e.g., larvae of coelenterates, echinoderms, mollusks, etc.). This problem is also common among fungi and lower plants.


Allometric Variation

This type of variation is typically thought of as being under genetic control. Allometric growth or variation results when the size of some particular structure or number of structures is disproportionate relative to other structures or the rest of the body. Thus, individuals of different sizes may have different sized traits. For body measurements this is solved with multivariate statistics or use of ratios. In fishes sometimes the number and sizes of breeding tubercules of adult males may be correlated with their body size.

With ontogenetic variation there can be variation in some traits through development. Other traits, like meristics, do not vary after they have formed. In some birds the size of the bill increases with age; horns of some mammals also increase with age.


Seasonal Variation in Individuals

In species that survive for more than one year or more than one reproductive season, characteristics may vary depending upon the season. For example, the plumage of birds, antlers of some mammals, and general breeding coloration.

In some invertebrate groups that produce multiple batches of young during a year may produce different types of offspring depending upon season. For example, in "cool" periods (spring) individuals may have different attributes relative to those produced during the summer. Butterflies in Spring are typically larger than those in the Fall.

Cyclomorphosis - season variation process that occurs in some invertebrates like rotifers and cladocerans. The variation in Daphnia can be quite extreme depending upon the temperature, turbulence of H2O, and other properties of H2O. In rotifers the type of food may alter morphology.


II. Social Variation

In some social insects (bees and wasps, termites) certain castes are developed (reproductive, workers, soldiers). The individuals may be males, females or both. The different structural types that are observed may be the result of different larval food or may be due to hormonal or other controls. Obviously these different forms should not be considered different species.


III. Ecological Variation

Habitat Variation

Populations of a single species may occur in different habitats in the same region and are often visibly different depending upon the habitat that they are found in. Taxonomic treatment of local variants of this nature have fluctuated between two extremes. Some researchers have considered them to be different species while others have considered them to be non-genetic variants. Obviously, one can only determine the status of such entities with additional information derived from controlled growth studies or genetic analyses.

Some species are particularly plastic, such as mollusks (snails and mussels). In these species those in the upper parts of rivers where there is cooler water and more flow have different forms from those in lower reaches with higher temperatures and lower flow. In limestone areas their shells may be thicker and of a different shape from those in other areas. In some cases it may be important to transplant populations or rear them in laboratories to solve the problem. As above, however, some of this variation may be due to ecophenotypic variation while other aspects of it may be heritable variation that can be detected using genetic studies.

Example: In France, Schnitter (1922) recognized 251 species of Anodonta; now considered to be 2 species.


Temporary Climatic Conditions

Some species have tremendous phenotypic plasticity and for some traits a different phenotype is produced in years of extreme conditions (drought, cold, warm weather) relative to those from other year classes under normal conditions. Fishes are commonly dwarfed in bad years. Stunted growth or periods of exceptionally rapid growth can be displayed in different proportional traits.


Host-Determined Variation

Parasitic species may display different traits dependent upon the host on which they feed. Cocoons can vary in color depending upon wasp host. Some wasps may be winged or wingless, depending upon host. Some hosts may display different traits when parasitized. Color patterns may vary with fishes sometimes if they are parasitized (usually this is obvious).


Density-Dependent Variation

Crowding can influence morphological variations. This can be a result of reduced food supply or not. Under crowded conditions the phenotypes may vary from those reared under less crowded conditions: This phenomenon is particularly common with locusts.


Neurogenic or Neurohumoral Variation

Color change in individuals due to regions in environment. Accomplished through the concentration or dispersal of color bearing bodies known as chromatophores. This has been observed in chameleons, some lower vertebrates, crustaceans, cephalopods, and flat fishes.

A common example of this phenomenon is that of colored fishes begin placed into a white bucket. After being in the bucked for a few minutes the coloration is "washed out" and the organism looks different to you than it did 5 minutes previously.

Another common example includes individuals captured over different colored substrates. In fishes those collected over a light sand will be lighter in coloration (and lose some of their darker colored markings) relative to conspecifics captured over a darker sand or gravel at the same location.


IV. Traumatic Variation

This type of variation occurs with varying frequency depending on the group. It is usually obvious, but in some cases may be subtle and misleading.


Parasite individual variation

Typical patterns discovered in a host individual will include swelling, distortion, and perhaps mechanical injury. With insects parasites can alter head size, wing venation, and other structural features. Parasitized fishes may appear pale and soft, have darks spots on the body, have weak fin rays, pop-eyed and pot-bellied appearance, have small mouths, nostrils that are joined, no lateral lines, an increase in scale numbers and other abnormalities.


Teratological or accidental

Alterations in development. Usually these are externally induced but can be developmental and may be from hormonal control. External stimuli for aberrant morphologies may be mechanical, physical, or chemical. Usually obvious because individuals appear as freaks!


Post-mortem Changes

Common in some museum specimens that have been fixed or preserved or pinned. Colors are often lost or fade, preserved bodies may appear odd in some cases. In some instances some color patterns do not appear until after specimens are fixed. Some insects that are yellow turn red in cyanide.

Thus, in taxonomic and systematic studies there is no substitute for observing your taxa in life from various locations within its range. Also, be sure to take careful notes of life colors.


Genetic Variation

Before, the same individual is actually or potentially subject to change in appearance. In addition to this non-inherited variation, there is much interpopulational variation which is primarily due to differences in genetic constitution. This variation can be more or less arbitrarily divided into two such classes.

I. Sex-Associated Variation

Among the genetically determined variants within a population, there may be some that are sexually associated. They may be sex-linked (expressed in one sex only) or be otherwise associated with one or the other sex.

  • Primary sex differences - Those that involve primary sex organs used in reproduction (gonads, genitalia). Where the sexes are otherwise quite similar, these will rarely be a source of taxonomic confusion.

    Secondary sex differences - Many groups display pronounced sexual dimorphism. These differences can be quite striking. Different sexes have frequently been described as different species until more work has been done on a group.

    Alteration of Generations - In some groups there may be an agamic stage that looks quite different from a reproducing stage. In aphids the parthenogenetic females are wingless whereas the sexual females have wings.

    Gonadromorphs and Intersexes- Gonadromorphs display male characters on one part of the body and female on the other. Due to unequal somatic distribution of sex chromosomes. Spiders

    Intersexes - exhibit a blending of male and female traits. Thought to result from upset in balance of male tendency and female tendency genes. Can be from irregularities in fertilization on mitosis or physiological disturbance due to parasitism. Occur most frequently in areas of interspecific hybrids.


  • II. Non-Sex Associated Individual Variation

  • Continuous Variation - Most common type of variation due to slight genetic differences which exist between individuals. No two individuals are exactly alike in a population genetically or morphologically. This is one of the foremost tasks of the taxonomist. No single individual is "typical" of the characters of a population. Only with statistics of the whole population can we arrive at the true picture of the whole population.

    Each character is likely to show different degrees of variation in a population. Likewise there will be differing degrees of variation between species for a character.

    Discontinuous Variation - Differences between individuals in a population are, in general, slight and intergrading. In some species, however, can be grouped into different classes determined by some characters. This discontinuous variation is frequently termed polymorphism. Frequency such polymorphisms may be controlled by a single gene.

  • Cichlasoma minckleyi - This is a cichlid species found only in some small bodies of water in northern Mexico. In these populations biologists have identified two different morphologies in tooth structure. After careful study, this appears to be an instance of a polymorphism in tooth structure.

    Peppered moth - industrial melanism.

    Many bird species have been proposed to demonstrate this type of polymorphic variation within populations for morphological characteristics.

    Some butterflies which mimic poisonous species may have more than one morphotype in a population. By possessing multiple mimics to poisonous butterfly species the polymorphic species has an advantage when it comes to predation. This is special form of Batesian mimicry.


  • Delineation of Species

    The following discussion concerning decisions on the level of species is derived primarily from Wiley (1981). The discussion includes some general strategies or "rules of thumb" that one may use in addressing question concerning the diversity among populations or demes and the naming of any of these as distinct species.

    Of course, an understanding of

    (1) the different types of characters,
    (2) the possible types of variation, and
    (3) the importance of interpreting character variation within a phylogenetic context

    are requisite to making informed decisions as to delineating species.


    Before we begin, some general background information and definitions are necessary.


    Species Concepts

    As we have discussed, various secondary, operational species concepts exist that are consistent with the primary concept of species, the Evolutionary Species Concept. These different secondary concepts optimize different criteria necessary for the recognition of species.


    Diversity of Taxonomists

    Expert taxonomists and systematists working with different groups of organisms frequently employ different criteria in determining the species status of a population or group of populations. They do this because the biological attributes of various groups may be quite different. Different types of characters exist for different groups of organisms, AND the degree of character variation across groups vary. What may seem like local demic variation to an ichthyologist may indicate distinct species to an entomologist. A gene may be highly variable for one group of organisms but very conserved for another group. Thus, there is no substitute for knowing as much as you can about characters, character variation and the life history or biology of the organisms that you study as possible.


    Some Definitions



    The geographic or temporal distributions of organisms can be important in making decisions regarding the taxonomic status of species in some species concepts.


    Sympatric distribution
    Two or more populations or taxa occupy the same geographic area.



    Syntopic distributions
    When two phenotypes are found in the same habitat (they are microsympatric). You may find these phenotypes in the same sweep net, seine haul, fish trap, etc.



    Allotopic distributions
    When two phenotypes are found in the same geographic area (sympatric) but are not found in the same habitat (they are microallopatric).



    Partly sympatric distribution
    Two or more populations or taxa have geographic distributions that overlap in the same geographic area but not in other areas.



    Allopatric distribution
    Two or more populations or taxa occur in different geographic areas.


    Contiguous/Parapatric distributions
    A type of allopatric distribution where the ranges abut one another.


    Disjunct distributions
    A type of allopatric distribution where the ranges are separated by geographic space.

    Cryptic species.

    These are species that cannot be diagnosed using traditional morphological data but act as independent evolutionary lineages in nature and, thus, qualify as evolutionary species. These species can, however, be diagnosed using other types of characters (e.g., chromosomes, DNA, behavior, etc.) and are valid evolutionary species.

    The practical problems of a species not being identifiable using "traditional" characters for a group of organisms or "convenient" characters must be secondary to the biological importance of recognizing and conserving this diversity. Accommodations can be made for these difficult-to-identify species. In keys and in collections these species can be included in "complexes" of species.

    Cryptic species have sometimes been referred to as sibling species because of their overall similarity -- referring to being siblings and implying that they are closely related. However, in such cases rarely if ever has one demonstrated that they are in fact sister species. Thus, the term cryptic species is probably a better term for this biological phenomenon.

    Two species of Sculpins in the Genus Cottus, Family Cottidae


    Wiley (1981:12) comments on the term epiphenotype as follows. The epiphenotype of an organism is "its morphology at any particular time it is inspected during its life. The term is largely synonymous with the term phenotype although the connotation of the prefix implies that the epiphenotype is the result of an array of genetic and ontogenetic phenomena." On page 61 Wiley (1981) further notes that he uses "the term 'epiphenotype' to connote morphological, ontogenetic, and genetic components of organisms." and that the "terms 'morphotype' or 'phenotype' could be used, but their implications are more limited."


    Modes of Reproduction

    It is common for many groups that the decision on the species level is tied to reproductive mode as well as morphology (here morphology can be viewed in a broad sense). Before we proceed some definitions are necessary.




    Asexual; produced asexually




    Species without sexual reproduction as in parthenogenetic aneuploids (having fewer or more chromosomes than an exact multiple of the haploid number)




    Reproduction without intervention of sexual organs; development of a sporophyte from a gametophyte without fertilization.




    An individual produced or reproducing by apomixis; a biotype resulting from apogamy and vegetative propagation.




    Reproduction (as in apogamy or parthenogenesis) involving specialized generative tissues but not dependent upon fertilization.




    Production of diploid gametophytes directly from diploid cells of the sporophytes without intervention of spore formation (certain ferns and mosses)




    Individual with no apparent sex.




    A group of individuals propagated by mitosis from a single ancestor; an apomict strain.




    Relating to or being reproduction that does not involve genetic recombination of genetic material from different individuals but does involve meiosis and the formation of a zygote by fertilization as in sexual reproduction (observed in some fungi with multiple nuclei).




    Reproduction by development of an unfertilized, usually female, gamete; reproduction without fertilization by a male element (usually among lower plants, invertebrates and very few vertebrate groups)




    Plants or animals that develop from a seed or ovum without fertilization by pollen or sperm.

    Asexual species.

    Functional asexual species show not sexual differences between members of populations. Some of these species are actually parasexual rather than asexual. Asexual species may reproduce by a variety of means depending upon the group. For these types of species a decision on the species level involves a decision by an investigator on the distinctiveness of the clonal population. Usually, when a clone is diagnosable from others it is considered a distinct species; Wiley (1981) does not advocate diagnosing clones as species if they are only statistically distinct.


    Agamospermic and parthenogenic species.

    These species are unisexual and are usually are derived via hybridization between sexually reproducing species. Agamospecies frequently have apomictic species complexes composed of one bisexual species and several apomictic species. The apomicts may be distinguishable on the basis of molecular or cytological data but are morphologically indistinguishable. In plants, it is frequently observed that an asexual species is the sister group of an agamospermic species. Both of these may be found sympatrically with a closely related (and ancestral) bisexual species. Both agamospermic and parthenogenetic populations represent independent evolutionary lineages and are thus evolutionary species.


    Gynogenetic species.

    These are species that are usually unisexual and usually all female. If males are present they are usually infertile. Sperm is necessary for the initiation of cleavage. However, since males are rare or may not be present of a species the females court males of other species and will mate with them. The sperm from these other species is not incorporated into the nucleus and no genetic introgression occurs.


    Bisexual species.

    These species reproduce via the fusion of egg and sperm. Many bisexual species do have complex life histories that include both sexual and asexual phases to their life histories (e.g., many plants). In some cases the life cycles of species are not sufficiently well known to identify the larval, juvenile, or adult stages as part of the same life cycle. In these situations commonly these different stages to be placed into different taxa. This is common among many invertebrate groups, some parasites, and is know to have occurred with fishes (Elopiformes: Leptocephalus for leptocephalus larvae; Ammocetes for ammocete life stage of lampreys; different parrotfish species).

    Among bisexual species were their life cycles are sufficiently well known most decisions at the species level can be divided into two distinct areas that are discussed below:


    1) decisions concerning populations that are sympatric

    2) decisions concerning populations that are allopatric.

    It the instances below, it is assumed that a researcher has observed two or more different epiphenotypes among the populations that have been examined and has identified the variation as heritable and of taxonomic importance. It is also assumed that this researcher is attempting to address the question as to how many species may exist given the available data. If only one epiphenotype exists, and the researcher possesses no evidence for the existence of a cryptic species, then there would be no need for a discussion.


    I. Sympatric Occurrences


    A. Complete or more-or-less complete sympatry

    1. Specimens from various localities are not distinguishable as distinct epiphenotypes although they may fall into integrated size classes.

    Here, it appears that intrademic variation equals or exceeds interdemic variation. In this situation probably only one species is present.


    2. Certain epiphenotypes predominate in part of the range while others predominate in other parts of the range.

    a. Further sampling may be needed to identify epiphenotypes in intervening geographic areas to demonstrate the existence of distinct epiphenotypes. IF further sampling shows that the epiphenotypes are connected by a continuum of variation in the diagnostic characters that cannot be ascribed to hybridization AND if the distinctive epiphenotypes are closest relatives then probably only one species is present with geographic variation.

    b. If there is a continuum of variation between two non-sister species then one may conclude that hybridization or intergradation has occurred in the past or is currently in process. These two epiphenotypes were distinct evolutionary lineages at one time and may still be or they may be merging into a separate, distinct lineage with a new identity (= compilospecies)

    c. If it is shown that the epiphenotypes are allopatric then consult following section on allopatric situations.

    d. If the epiphenotypes are allotopic or syntopic and do not form a continuum of variation then they would represent distinct species.


    3. Epiphenotypes occur syntopically but each comprises a separate sex in mature individuals. A single sexually dimorphic species is present. Frequently, the immature specimens will be similar to one or the other sex.


    4. Epiphenotypes are syntopic, one composed of sexually mature males and females, the other(s) composed of sexually immature specimens. A single species is present. Growth patterns are frequently distinct because of discrete life history stages.


    5. Mature specimens fall into two epiphenotypes, each represented by both sexes, immature specimens fall into one or more epiphenotypes. Two sexually monomorphic species are present. In cases where the immature specimens fall into one epiphenotype, they may resemble one of the two epiphenotypes present in adults. If so, that mature epiphenotype is likely a plesiomorphic condition.


    6. Epiphenotypes are syntopic, one composed of diploid and one of haploid individuals. Likely a single species showing alternation of generations is present. One must, of course, match the haploid with the diploid where more than one species showing this life history is present in the same locality.


    7. Epiphenotypes are syntopic or allotopic, reproductive modes differ. This example covers apomictic and asexual complexes. See the earlier discussion on asexual species.


    8. Epiphenotypes are syntopic or allotopic, basic chromosome number differs. If the epiphenotypes form a natural group (a clade), the epiphenotypes are probably separate species. However, some species show chromosome polymorphisms and some polyploids occur as natural polymorphisms in a single species, so care should be exercised.


    B. Partial Sympatry

    1. Each geographic population is a distinct epiphenotype, no intermediate epiphenotypes are found in the area of sympatry. The two epiphenotypes are distinct species. Sympatry with. no evidence of interbreeding is prima facie evidence for separate species. The situation is somewhat complicated by sexual dimorphism where the females are very similar. In such cases, the female epiphenotype may be considered a plesiomorphic similarity and should not influence the decision.


    2. Each geographic population is a distinct epiphenotype in allopatry, sporadic hybridization or local (limited) introgression in part(s) of the area of sympatry. This can be a very complicated system to interpret. However, given the above information, Wiley (1981) concludes that the populations are separate species. Localized hybridization or introgression occurs regularly between species in some groups. Species that do not hybridize or introgress under normal circumstances may do so in disturbed habitats. Either may also occur under special ecological circumstances.


    3. Each population is a distinct epiphenotype in allopatry, but there is complete introgression throughout the zone of sympatry. Introgression is an interesting phenomena. Whether two introgressing populations warrant species status depends on (1) the phylogenetic relationships of the populations and (2) the width of the zone or intergradation. Three cases will be discussed:

    a. lntrogression occurs over a wide geographic area and the populations are nearest relatives

    Wiley (1981) considers such populations part of a single species showing local adaptation and, consequently, introgression between the locally adapted epiphenotypes. At the same time, it is quite possible that parapatric speciation is occurring or that the zone of introgression is becoming narrower. However, given the inability to diagnose two forms without resorting to geographic data and the inability to diagnose a significant geographic population (the introgressed individuals), there is no great advantage to naming the two "parental" populations as distinct species.

    Thus, such a pattern of variation could either be a case of parapatric speciation or alloparapatric speciation followed by complete fusion of the earlier descendants or simply divergent populations in a species with a wide geographic area of introgression.



    b. Introgression occurs between nearest relatives only in a narrow sympatric zone, parental epiphenotypes predominate outside the zone.

    The investigator is justified in naming the parental epiphenotypes as species. This situation is evidence for species-distinct characters (species identity) being selected outside the zone of sympatry. It is evidence that both epiphenotypes show independent species cohesion. Whether characters that do not contribute to species cohesion are introgressed is not relevant. Wiley (1981) notes that while no examples exist in the literature it is possible for characters that do not contribute to species cohesion to introgress widely without effecting the taxonomic conclusion that the two taxa are distinct species. When certain aspects of a parental epiphenotype determines assortative mating, it is quite possible for introgressed individuals possessing these features to mate back only with one parental type (A) carrying fully one-half of the nonessential genes of the second species (B) into the first species (A). Since these genes do not affect the cohesion of species A, they can introgress widely without altering a species' identity or cohesion.


    c. Introgression occurs over either a wide or narrow geographic area, the populations are not closest relatives.

    These instances of biodiversity necessitate more study than is usually accorded them. Unfortunately, in most cases sister-group relationships of the involved taxa are unknown; however, most often it is assumed that because of the introgression they must be close relatives. The tacit assumption is that interbreeding populations must be the same species, or at the most, closest relatives. However, Rosen (1979) has made the point that the ability to interbreed may be a plesiomorphic character. Thus, the ability to interbreed cannot be considered prima facie evidence for a sister group relationship. In each phylogenetic hypothesis shown above, the investigator is faced with one of two choices. First, the investigator can consider A1, A2, B, and C as members of a single species. Second, the investigator can consider each a separate species. Given that there is evidence provided for distinct species status of taxa B and C, then A1 and A2 are species regardless of their introgression unless it can be demonstrated that A1 and A2 are two populations of the ancestral species that gave rise to B and C. Such situations are possible given a complex biogeographic history involving peripheral isolation (see Figure below). If such is the case, similar tree topologies would be expected in at least some other groups inhabiting the same biota.


    Bisexual Species - Allopatric Occurrences

    Because sister species are most frequently allopatric, critical analysis of species status must be conducted without the aid of an interbreeding criterion. Laboratory studies may be helpful for some questions or in some situations if the organisms are suitable for such analysis. However, the vast majority of decisions must be made without the benefit of laboratory studies. This may be due to the interest of the systematist, but usually it is the nature of the systematic material one has to work with. The number of different situations involving allopatric populations are many. Wiley (1981) has organized them into three general cases.


    1. A single epiphenotype (or two if sexually dimorphic) found in two or more disjunct areas.

    A single species is probably present. The inability of an investigator to diagnose the disjunct populations is evidence that they are the same species. Variation within each disjunct population would be expected to equal or exceed variation between the populations. Consistent differences should be on a statistical level only. Multivariate analyses would not be expected to cluster individuals into consistent clusters corresponding to the a priori designated disjunct populations.


    2. Two disjunct populations are obviously differentiated but no one character or combination of characters can be used to diagnose them as separate entities.

    This type of taxonomic situation may involve one or two species and further study is warranted. Further study may be carried out at several levels.

    a. Look for less obvious but diagnostic characters (in many fishes subtle but constant pigment patterns. have frequently been used).

    b. Use of canonical variates analysis or other multivariate methods.

    If individuals are consistently classified correctly into a priori designated groups, the hypothesis that two species are involved is given some support. c. Use canonical variates analysis with local deme samples and not the two suspected species being designated as a priori groups.

    This type of analysis is illustrated to the right. Here there are two suspected species, A1 and A2 (see top figure). Rather than lump population samples 1-6 into A1 and 7-12 into A2 and perform a discriminate analysis (or other multivariate analysis), the investigator conducts a canonical variates analysis with 1 to 12 as the a priori groups. If 1-6 separate into one cluster (the original A1) and 7-12 into another (the original A2), then the hypothesis that A1 and A2 are separate species is given support (see middle figure). If the samples are mixed (see lower figure), then the hypothesis of two species is not supported. If these organisms are suitable for lab study, physiological, breeding, and behavioral characteristics can also be studied. Finally, one may also examine molecular characters to assist in delineating possible boundaries of these taxa.

    3. Both populations are diagnosable without recourse to geographic data.

    This situation indicates two species unless the investigator can ascribe the observed differences to ecophenotypic variation. Given that ecophenotypes are not involved, consistent differences between two allopatric populations is indicative of independent evolution of one or both from the original, ancestral epiphenotype. This is reinforced if the populations are continuous but show no signs of intergradation.


    Biogeographic Data and Allopatric Decisions

    Wiley (1981) discusses the use of biogeographic data, replicated patterns of speciation in an area, and evidence for diagnosability in making decisions as to the limits of species.

    It is clear that there elements of our biotas that frequently display replicated speciation patterns, with sister groups in one monophyletic group having species boundaries that correlate with those of unrelated groups (discussed in previous lecture). Many argue that the most plausible explanation patterns of this nature is that a common speciation event effected these diverse clades with diverse dispersal capabilities (also discussed in great detail by Croizat, 1964, and many others). Thus, an ichthyologist can benefit from knowledge about the speciation patterns of aquatic insects and the aquatic entomologist may benefit from knowledge of the speciation patterns of fishes.

    Of course, it should be clear to you that population structure, species cohesion, and vagility can vary from one group of organisms to another. Vagile organisms or species with constrained developmental homeostasis may not be affected at all by the same isolation event that resulted in a replicated pattern of speciation in other groups of organisms. However, where replicated patterns of speciation (biogeographic patterns) are evident, they may be regarded as clues in helping to investigate species questions in other groups of organisms.

    Wiley (1981) uses the following example. Given that two taxa can be distinguished or diagnosed we might look at their phylogenetic relationships to other species and the potential biogeographic relationships to other groups of organisms sharing the same range. If in several groups of organisms occupying the same geographic area there are recognized, valid species that identical biogeographic pattern this may lend some support to the hypothesis that the two species in question are also valid taxa.


    Experimental Data and Species Decisions

    Some have argued in the past that laboratory breeding studies are necessary to demonstrate the validity of species that occur in allopatry. Hence, speciation is not complete until it can be demonstrated that reproductive isolating mechanisms have evolved to prevent the loss of identity and cohesion.

    It is possible to strengthen a decision regarding a species' validity based on morphology, molecular data, biogeographic information, etc. by performing experiments in the laboratory or garden. If members of two presumptive species are successfully crossed and the progeny are inviable, sterile, or largely sterile then the hypothesis of species status is reinforced or corroborated. However, some animals are hard to breed in the laboratory and a failure to mate successfully may be due to conditions (or lack of) in the laboratory or to species distinctions. Behavioral information may resolve the situation if it can be shown that mating behavior differs in the two presumed species so that premating isolating mechanisms are working (frequently the case in some taxa, best demonstrated in species of Drosophila).

    Laboratory studies of this nature must be carefully interpreted. Often isolating mechanisms break down in a laboratory setting when a female of one species is offered only the male of another species. Furthermore, many perfectly "good" species may freely interbreed in the laboratory or garden but not in nature because they exhibit ecological differences (perhaps they are allotopic) or, in the case of plants, they are pollinated by different animals. Thus, successful crossing between taxa is subject to many interpretations, only one of which is the conclusion that the two populations belong to the same species.




    Mensural Data - measurement data.

    Data Type:

    Shape, Proportional comparisons

    Data Analysis:

    Various univariate and multivariate methods. Univariate Comparisons: ANOVA, ANCOVA, Multivariate Comparisons: ANOVA, ANCOVA, Principal component analysis, discrimiNate function analysis.

    Data Presentation:

    Univariate Comparisons: bivariate plots, dice diagrams
    Multivariate Comparisons: two axis plots, three axis plots

    Meristic Data - count data

    Data Type: Numbers of elements (fin rays, scales, feathers, segments)

    Data Analysis:

    Various univariate and multivariate methods. Univariate Comparisons: ANOVA, ANCOVA, Chi-Square, t-test Multivariate Comparisons: ANOVA, ANCOVA, Principal component analysis, discriminate function analysis.

    Data Presentation:

    Univariate Comparisons: tables, bivariate plots, dice diagrams
    Multivariate Comparisons: two axis plots, three axis plots


    Coloration Data

    Data Type: Colors, Color patterns

    Data Analysis: Qualitative comparisons, standard codes for colors

    Data Presentation: Usually color figures.



    Data Type: Allozymes, Isozymes, Haplotypes, Sequences, Rapids, Microsatilites

    Data Analysis:

    F-statistics, Hardy-Weinburg Statistics, Heterogeneity Chi-Square, genetic distances (similarities),

    Data Presentation:

    Tables of variable genotypes, frequency tables, tables of F-statistics, phylogenetic analyses, cluster analyses, genetic distances