By Greg Lutz*
More often than not, this is a reasonable concern. Most aquaculture operations are already working with stocks that are at least partially “inbred.” Most of these operations are still quite competitive and profitable, however, because the real danger from inbreeding occurs if it is left to follow a random course. Controlled inbreeding, of course, has a name we all recognize: SELECTION. Sounds familiar (pun intended)? This type of inbreeding, of course, has been proven to be a very effective approach to genetic improvement.
Whenever someone boasts about fish, or shrimp, or mollusks that have undergone rigorous selection for some trait or another, they are talking about comparatively inbred animals. Most established varieties of livestock, ornamental plants, and aquatic species in the aquarium trade are the product of intense inbreeding, and yet these organisms persist and are cultured profitably. Inbreeding can be defined simply as the mating of two individuals whose degree of relationship is higher than the expected, or average, value would be if members of the population were allowed to mate randomly.
By way of review, let’s remember that chromosomes are paired within each cell of an organism, with the two chromosomes in a pair having the same regions coding for the gene products that affect particular traits. Each region on a chromosome, with its own particular genetic products, is referred to as a “locus.” Within an organism, there can be no more than two forms of a gene (“alleles”) for any given locus, because there are only two (paired) chromosomes that include that particular locus. If both forms of the gene are identical, the organism is “homozygous” at that locus. All of its offspring will inherit that particular allele, because that is the only one it possesses. If two different forms of the gene are present at the locus – with each chromosome producing a different gene product, the organism is “heterozygous” at that locus. On average, half of its offspring will inherit one of the alleles it carries and half will inherit the other allele. Simple stuff, really.
A practical way to look at inbreeding is by determining the probability that both alleles (forms of a gene) at any given locus (point on a chromosome where that gene exists) are identical by descent from a common ancestor. There is a 50% chance of two offspring from the same individual inheriting the same allele at a given locus, and a 50% chance for every generation thereafter that that particular allele will be passed on from parent to offspring (because an individual only carries two alleles for the gene in question – one on either chromosome of the pair where we find that particular locus).
Now, people often forget that most genes in most organisms only have a few common forms, so an organism can have identical alleles at many loci without necessarily being “inbred” at all. Many organisms have only one form of a gene at most of their loci. This allows populations and species to be differentiated in many biological and ecological studies. It also allows for performance to be improved when animals and plants are crossbred or hybridized.
So… as you scratch your head you are still wondering why inbreeding is a bad thing. It is intrinsically linked to the concept of dominance. As we have just seen, for any given “gene” - organisms have two forms of that gene (alleles) at the locus in question. When dominance is present, one form of that gene can partially or completely mask the expression of the other form. In traits associated with fitness, inbreeding depression is thought to result from ‘directional dominance’ at the majority of loci affecting the trait. Directional dominance in this sense means dominance in the direction of increasing (improving) fitness. So, the “better” alleles tend to be dominant over the “inferior” alleles. Theoretically, this is an oversimplification, but it helps in a discussion of inbreeding.
Breeding between more closely-related individuals tends to increase homozygosity (the presence of two identical gene forms) at all loci. So, as closed populations reproduce over a number of generations all alleles tend toward becoming fixed, one way or the other. The term fixed simply means that over time one gene form or another becomes more and more prevalent until the other gene forms simply disappear from the population. The larger the population, the longer this may take.
As this process runs its course, many of those inferior gene forms, whose expression would normally be masked, become increasingly common. Just by the fact that mating more closely related individuals increases the number of animals who can only pass on one particular gene form… be it superior OR inferior. As a consequence inbreeding results in a reduction of the mean values of traits associated with fitness (specifically in the direction of the less dominant alleles) when the effects of all the genes impacting the traits are considered together.
Inbreeding, by its very nature, is of greater concern in small populations, where random “sampling” in the formation of genotypes from an available pool of gametes can result in the eventual exclusion of certain alleles and the fixation of others. The smaller the effective population number, the greater the incremental increase in inbreeding per generation, and once inbreeding levels have accumulated within a population, they cannot be reduced simply by increasing the population number. At that point, it’s not a question of the number of animals one has to work with, but rather the number of alleles.
At the very least, if inbreeding is unavoidable due to limited facilities it can be offset somewhat through rigorous selection for fitness-related traits. Additionally, a breeding population can be divided into two or more sub-populations, each under selection for production- and fitness-related traits. These lines can then be crossed regularly to produce fingerlings for growout purposes, in theory reducing the overall inbreeding load in the commercial stock. If individual animals can be marked and tracked, pedigrees can be used to mate the least-related individuals within a population, but this tends to simply delay the inevitable in many situations. And, it often results in the use of inferior breeding stock simply to minimize inbreeding at the expense of selection gains.
How much inbreeding is too much? How can it be measured? Inbreeding is a relative concern (no pun intended this time). It must be gauged against some arbitrarily-identified base population. While many quantitative genetics texts address the calculation of inbreeding coefficients for individual breeding animals, this approach often has little use in commercial production of aquatic species. An inbreeding coefficient serves to describe homozygosity within an individual, but only that homozygosity resulting from inheritance of alleles that are alike by descent: alleles originating from a common ancestor.
So, the calculation of an inbreeding coefficient must also take into account the level of inbreeding in any common ancestor(s), but pedigrees are often impossible to trace back far enough to allow for much precision. The pedigree information required for individual fish, crustaceans or shellfish is usually unavailable or entirely impractical to generate. For most purposes, the accumulation of additional inbreeding can only be tracked over time for each captive generation of the breeding population in question. In most cases, this can be estimated as:
where Ne is the effective number of breeding individuals.
And Ne is
where Nm is the number of breeding males and Nf is the number of breeding females.
However, many individuals may not actually contribute much in terms of breeding due to behavioral hierarchies or biological factors.
If the base population for a commercial hatchery was originally 80% heterozygotic, as might be the case if two distinctly different strains were crossed and then the offspring used as the initial breeding stock, and an inbreeding coefficient of 0.30 has accumulated, then (0.30)*(80%), or 24%, of the original heterozygotic loci can be expected to have become homozygous due to inbreeding. The resultant level of heterozygosity is now 80%-24%, or 56%.
When available, however, pedigree information allows a much better estimation of inbreeding levels than simply tracking population numbers from generation to generation. Pante, Gjerde and McMillan reported on inbreeding levels in rainbow trout (Oncorhynchus mykiss) from three isolated populations. Inbreeding levels were calculated based on pedigree information (which in this case was available), and on effective population size. After the initial generations, levels of inbreeding estimated from population numbers were lower than those calculated using pedigrees. For the three populations, average rates of inbreeding calculated from population sizes were 0.99%, 0.90%, and 0.72%. Corresponding average rates calculated from pedigree information were 2.00%, 0.53%, and 1.38%, respectively. This is because all individuals do not contribute equally – some males (and often some females) produce more than their “fair” share of the next generation.
These inbreeding values were considered acceptable, based on data presented by Meuwissen and Woolliams (1994), which suggested inbreeding levels of 0.2% to 2.0% should not cause any loss of fitness. However, in a companion study, the authors determined that the actual inbreeding effect might vary between populations due to the distributions of inbreeding coefficients within each generation. That is, some individuals might be much more highly inbred than others. Taking this into account with statistical models that included sires and dams, as well as additive and dominance genetic effects, the authors determined that inbreeding effects might actually be somewhat higher. For the three populations in question, inbreeding effect on body weight at harvest, calculated from the more complex statistical models, was determined to range from -1.6% to -5.0%. Nonetheless, these impacts were not considered significantly serious to impact the selective breeding programs being carried out with these fish.
Again, one strategy available to producers with limited facilities involves maintaining separate lines, subjected to intense selection, and crossing them to produce juveniles for growout. In a study of inbreeding in Ostrea edulis, the European oyster, Naciri-Graven examined three populations that had previously been selected (read: INBRED) for resistance to the protozoan parasite Bonamia ostreae. In addition to these three populations, a population comprised of their crossbreds was studied, as well as a control population. Each population was composed of full-sib families derived from individual oysters that had been genotyped using microsatellite markers. In this way, the authors knew not only the parentage of specific oysters, but also the degree of relatedness of the parents. Inbreeding depression could be evaluated by comparing the relatedness of parents with the growth performance of their offspring. Growth was monitored for 10 months. At the end of the grow-out period, the crossbred population had the highest growth rate and, of course, the lowest level of inbreeding. Within the selected populations, although all exhibited high levels of inbreeding, the rankings in both growth rate and inbreeding levels were similar. So, crossbreeding produced the best growth, but inbreeding still resulted in good growth… once again, this phenomenon reflects prior selection. The slowest growing population in the study was the control group.
Similar results were reported for Pacific oysters, Crassostrea gigas, by Pace and Manahan. They examined growth rates and feeding rates for crossbred and inbred C. gigas larvae. In four separate experiments, each involving 2 or more inbred lines, crossbreds exhibited superior performance, not only in growth but in size-specific feeding activity. At any given size, crossbred larvae were ingesting more algae than their inbred counterparts.
C. Greg Lutz, has a PhD in Wildlife and Fisheries Science from the Louisiana State University. His interests include recirculating system technology and population dynamics, quantitative genetics and multivariate analyses and the use of web based technology for result-demonstration methods.