Using genetics to estimate spawning stock size for sustainable fisheries

After a recent meeting about innovative ways to use genetics in the field of wild fisheries management, a colleague said as he left my office, "Is there nothing that genetics will not be able to tell us?"

The complete sequences of animal and plant genome are providing waterfalls of raw data about how many genes and how much DNA an organism needs to function. But the raw sequence data needs considerable analysis effort before its usable. For example, even knowing the complete sequence of the host (human), the vector (mosquito, Anopheles spp.) and the disease (Plasmodium falciparum) will not quickly rid the world of malaria.

Not all advances in the field of genetics are coming from sequence genomics. "Genomic epidemiology" or population genetics is being used more and more in wild fisheries management. Nothing is more frustrating than trying to manage the harvest of a population of wild animals that you cannot stand on top of a mountain top and see, or observe from a hide in the bush. You only see the fish dead in large numbers on the wharf or processing factory. Harvesting of wild fisheries is the fifth-largest earner of export dollars for the Queensland economy, and supports many regional communities in areas with little alternative employment opportunities. So in desperation (almost!), fisheries managers have turned to an unconventional way of learning about the species they need to regulate.

The scientific field of population genetics is particularly good at uncovering population structure where none is apparent. For example, how do closely related fish species on the same reef recognize the right mate? Two small, brightly coloured fish may look the same to us, but they know they are not the same species and biologists often need to use population genetics to find out. Sometimes, the genetic differences between animal populations can be related to events in the earlier history of their habitats. Fish in streams that once shared a common mouth will be more similar to each other than fish in close-by streams that flow in the opposite direction down the other side of the range.

However, beyond this accepted application of population genetics, biologists are skeptical that population genetics can also provide estimates of the number of individuals in a population. And when the populations to be counted are naturally occurring, and abundant fisheries species, such as prawns, the response is incredulity. Yet, the groundwork for genetic counting was laid in the field of population genetics in the 1930s by two learned mathematicians, Sir Ronald Fisher (UK), and Professor Sewall Wright (US). They were interested in the most basic of evolutionary forces; genetic drift, and independently discovered that it could be elegantly described with equations if a series of simplifying assumptions were made about the population. Nowadays, the effect of breaking these assumptions on their theory has been quantified.

Genetic drift has nothing to do with the driftwood that is washed up on a beach. Rather, it is the overall change in genotypes that may occur from one generation to the next. Here, a genotype is a summary of the genetic make-up of an individual; for example the genetic traits that determine size, shape and colour, and it is inherited equally from mother and father. Each individual in a wild population generally has a different genotype and the population may consist of a range of animals from dark blue small fish to light blue large fish. If each of these fish finds a mate and all of the pairs have two offspring, then the offspring population generally is similar in size and colour to the parent population. But, it's often difficult to find a mate, to find a place to leave your eggs, to have eggs that hatch and not get eaten, and to have baby fish that find enough food to grow into adulthood. Consequently, only some of those mated pairs of fish will have offspring, and it's quite possible that by chance the light blue, large fish will be more successful than some other type. So the offspring will generally be different, and have different genotypes compared to the parent generation. This is genetic drift; it occurs by chance and is a powerful evolutionary force.

Fisher and Wright made the connection between the magnitude of genetic drift; how much a population changes from generation to generation; and the number of parents that were successful in leaving offspring. A large number of successful parents means a small amount of drift and visa versa. Hence, a novel application of population genetics to "counting" fish: measure how much drift is occurring and you have a measure, albeit indirect, of the number of individuals in a population that breed successfully.

This relationship has been known for at least 60 years, so why is it only now being taken up by scientists? In part, because developments in genotyping have only now made it possible to rapidly and cheaply genotype large numbers of individuals from a population. But also because it has just been realized that "genetic counting" can provide a unique and critically important measurement for fisheries management.

To predict the future size of a fisheries population, mathematicians make models that describe the growth of the population based on past measurements. An important data source for these models is the number of spawning fish that successfully leave offspring, called recruits. For those who deal with human populations, or captive domestic animals, this is a difficult concept. But fisheries biologists, and other scientists dealing with invertebrate populations, know that even though a female prawn lays 500,000 eggs her chances of having even one offspring in the next generation is low. Why, indeed, would she bother to produce that many eggs if all of them were going to grow up? So, high fecundity is balanced against severe environmental mortality; predation, starvation etc. However, because of high fecundity the potential exists for a relatively small number of spawners to be the parents of the entire next (offspring = recruit) population. This is what is thought to occur with many fisheries populations. Consequently, genetic drift is large and measurable in fisheries species and provides an independent way of estimating how many spawners are successful.

Genetic "counts" of successful spawners is the "holy grail" for fisheries population modellers. Better population models have the potential to provide more accurate information about fish population growth or decline that can be used by fisheries managers for better resource management. Genetic counts are estimates of spawner and recruit abundance in one package. They represent only those spawners in the breeding frenzy that will be successful, but equally are also the number of recruits produced. Further computer simulations are planned that will test the idea that "genetic counts" are the glue that connects the spawner-recruit relationship in fisheries modelling.

Innovation is often the judicious combination of the old and new in a novel context. In this case, resource sustainability concerns of fisheries managers provided the motivation to blow the dust off some old volumes about mathematical population genetics. Add a twist of the latest methods in genotyping, and "genetic counting" has been brought to life in modern fisheries science.