Population Genetics for Frog Hobbyists

Brent L. Brock M.Sc.
Ecologist/Information Manager
Konsa Prairie Biological Station
Kansas State University

One frequent topic in the frog hobby is whether the latest morph of species X is legitimate. The question rarely has a definitive answer because the term "morph" lacks a clear definition. Taken literally a morph is simply an animal that conforms to some predefined morphological conformation. Questions of legitimacy arise because within the hobby labelling an animal as a specific morph generally implies that the morphological traits used as the basis for the label will "breed true". In other words, the "morph" is some genetic subgroup of the species. Problems arise because the information needed (principally locality data) to infer a genetic sub grouping is almost always lacking for captive animals. Additionally, there is an economic incentive to "discover" new morphs since new or rare morphs typically command premium prices. Finally, new morphs can be created through hybridisation or selective breeding which are practices condemned by a significant number of people in the hobby. Therefore much of the debate over legitimacy of morphs is driven by a desire among hobbyists to protect their collections from breeding stock that was derived through artificial manipulations. The purpose of this article is to illustrate graphically how genes commonly flow among populations and highlight mechanisms and outcomes relevant to frog hobbyists.

The discussion that follows will build on the most simple population model to illustrate what morphs are and how they are derived in nature. Hopefully the reader will gain insight into the conundrum of assessing the validity of morphs in the hobby. The following models and arguments depend on some basic assumptions:

Conventions
To illustrate these models I will use standardised diagrams. In each diagram:

Frogs tend to be more closely related to frogs living close by than to distant frogs. To my knowledge frogs in the wild have not discovered the Internet dating services that we provide our captive specimens. Instead they rely on the old method of finding mates at the local pub. Therefore a frog in the wild is much more likely to mate with "the girl next door" than one several kilometres away.

Habitat is not uniformly distributed throughout a species' range. For any species there are usually areas of preferred, marginal, and unsuitable habitat within the range of the species. The patterns of habitat quality can create bottlenecks or barriers to gene flow. Genes tend to flow more easily within patches of high quality habitat than between them.

Hobbyists want to maintain captive populations genetically similar to wild populations. This one is more of a warning than an assumption. Obviously not all hobbyists have the same goals and the information here should be as useful to someone interested in selective breeding as to some one maintaining "wild type" frogs. However, we all travel with biases and this is one of mine so the discussion will be based on this assumption. Try not to be offended if this assumption does not apply to you.

An orange polygon with solid black boundary represents a species range. Simple models with only one population in the species range will not have an orange polygon because the population boundary and species range are the same.

A baby blue circle with solid or dashed boundary represents a population of animals. The dashed boundary symbolised a permeable boundary that individual animals can pass through (genetic exchange). A solid boundary indicates a species with a single population so the species range and population boundary are the same.

Arrows between populations represent gene exchange. Thick arrows indicate frequent gene exchange while thin arrows indicate reduced frequency of exchange.

Coloured circles within a population indicate individual animals with differences in colour denoting morphological differences.

For simplicity I will restrict this article to population genetics within a single species. However, these same principles can apply to gene flow between closely related species. In the diagrams that follow, you could easily substitute "species" for the population circles and substitute "populations" for the individual animal circles. Thinking about the models in this way may shed some light on why species designations change and are so hotly debated among taxonomists. Through out the discussion I define a population as a collection of animals with some isolating mechanism that restricts the exchange of genes between one population and another. I will focus primarily on physical barriers (e.g. mountains, deserts, or water barriers) and distance as isolating mechanisms. I will not include discussions of behavioural or other genetically driven isolating mechanisms that can create subpopulations (populations within populations). What is presented is al ready complicated enough!

Single Population Models

 

Scenario 1 shows the most basic model for a species. It consists of a single population with morphologically similar individuals in the population. However, this situation rarely occurs among sexually reproducing animals because sexual reproduction continuously mixes genes into new combinations. The result is discernable morphological variation. Scenarios 2 and 3 show simple representations of morphological variation. The population in scenario 2 has 3 distinct morphs randomly distributed. However the degree of relatedness between individuals is independent of the morph. A red frog could be just as related to a yellow one as it could to another red. Scenario 3 depicts another form of morphological variation often exhibited at large spatial scales. This form is sometimes referred to as clinal variation. Individuals at opposite ends of the range can be morphology distinct with a continuous gradation of inter grades in between. This is the result of assumption number 1 that animals are more related to those close by than to distant members of the species or population. Here, blue frogs tend to breed with blue and red with red simply based in locational distance and probability. However, there are no barriers to gene transfer across the range so a gradient of morphology develops.

Assumption 2 states that habitat is not uniformly distributed throughout a species range. Almost invariably a species range will contain patches of varying quality habitat. Within the range there will be areas of unsuitable habitat that are too hot or cold, wet or dry, etc. for the species to exist. Other areas may be marginally suitable. For example the area may be lacking in tadpole deposition sites but otherwise suit able for supporting a species of frog. These areas may become population "sinks" where animals can disperse into and survive, but they cannot reproduce in sufficient numbers to sustain the population without a constant influx of animals from higher quality ("source") habitats. Small patches of unsuitable habitat rarely pose more than a very localised barrier to genetic exchange but large patches like rivers, extensive clearings, or mountain ranges tend to genetically isolate populations on either side of the barrier. These barriers may be impermeable where no gene exchange of any consequence crosses the barrier, or semi-permeable where genes can cross but at a reduced frequency. The latter might be barriers like rivers where the occasional frog crosses (often unintentionally), or marginal habitat might bridge the gap between populations and the sink habitat gets colonise by animals from source populations on either side.  These semi permeable barriers create bottlenecks in genetic exchange and allow animals on either side to evolve different morphologies or local adaptations.

 

Multi-population models
Scenarios 4-6 show 3 hypothetical examples of multi population gene exchange. In scenario 4 there are 3 discrete populations but all of the frogs are morphologically similar. However, the rate of genetic exchange between the populations varies. Exchange between populations A and B is much greater than between A and C. Therefore we would expect greater genotypic differences between A and C than between A and B. Just because all of the frogs among these populations look similar does not mean they are. Each population may have genetically fixed adaptations to local conditions. For example, perhaps population C lives in a much wetter habitat and does not have well developed egg wetting behaviour compared with population A from a dryer region. This could have consequences for the hobbyist because frogs from population C may not breed well in a setup that has been dialled in successfully for population A. The hobbyist may be perplexed as to why their frogs won't breed in a setup that has proven successful for identical looking frogs before. Scenario 5 depicts basically the same scenario as number 4. The difference is that each population contains a number of morphs but all morphs are represented in all populations. The rates of genetic exchange are the same between the two scenarios so the same potential consequences apply. Simply choosing only yellow frogs from among the 3 populations does not insure that you are pairing like frog together. Additionally, pairing based on morphology may rob the collection of much of the charming variability found among the population. The breeder striving to maintain wild type frogs would be better off choosing frogs of mixed morphology from within a population. Finally there is scenario 6 that is the ideal situation for determining morphs. This scenario depicts 3 well-isolated and distinct populations with each population representing a different morph. There is little ambiguity here because the genetic population and morphology are matched to produce frogs easily distinguishable based on appearance. Of course often the difference between morphs are rather subtle even when strong isolating mechanisms are at work.

These scenarios are not comprehensive. What I have depicted is only a smattering of the more simplistic hypothetical possibilities. Regardless, the scenarios illustrate the ambiguity of determining a "morph". Let's briefly review the morphs depicted in these scenarios. Scenario 1 contains a single morph in a single genetic population. Scenario 2 depicts 3 morphs but only one genetic population. Scenario 3 contains at least 2 morphs blue and red. There could be even more morphs depending on how one wants to slice up the intergrades or how valuable they want to make their frogs. There are no clear boundaries delineating where one morph ends and another begins so the legitimacy of intergrades will likely be contested. Scenario 4 is back to a single morph again but this time there are 3 genetic populations with varying degrees of relatedness. The frogs look alike but are they really? Scenario 5 extends this a step further and introduces 3 morphs but distributed within each of 3 populations of varying related ness. As with scenario 4, yellow frogs from each population may look alike but are they really? Finally scenario 6 brings us to morph Nirvana complete reconciliation of morphology and genetic populations. If only all morphs could be this clear-cut.

Hopefully this sheds a little light on why morph designations are so confusing and subject to fraudulent representation. However, this discussion begs the question of whether anything can be done to clarify things. For frogs already in the hobby, the answer is, probably not. For frogs yet to be imported, yes locality data. If hobbyists began to demand locality data about the frogs that they import or purchase, much of the ambiguity would go away. A morph could be named for the locality of collection rather than its appearance. That way the natural variability characteristic of a local population would move along with the morph. No longer would one have to choose suitable breeding stock based on the minutia of morphology as if they were trying to breed show dogs. Two frogs from the same location could be bred without repercussion despite whether one frog had a black spot on its butt and the other did not. The result would be frogs that retain their local distinction and flavour without losing their variable charm.

Hit Counter