Genetic methods hold promise for estimating dispersal patterns, though it is important to remember that genetic methods only measure effective dispersal, and not dispersal ofindividuals that did not successfully breed in the new population. In addition, for organisms with motile gametes, genetic patterns will likely reflect the movement ofgametes among populations as well as the movement of diploid individuals. Most genetic methods involve collecting DNA from immature or mature individuals, then analyzing the DNA to try and identify the origin of a particular individual. In addition, if a dispersed individual has the same genetic signature as the individuals in the new population, the dispersal event will be undetected; this becomes less likely when highly variable markers are used. Another possible drawback to genetic data is the potential for unsampled source populations to contribute to apparent gene flow estimates between two sampled populations. Also, it is difficult to generate a detailed dispersal curve using solely genetic data, especially at local distances, due to the large amounts of data that would have to be collected and the heavy expense involved. Nevertheless, genetic data enable estimation of many parameters of interest, such as historical amounts of gene flow, effective dispersal rates among differentiated populations, and dispersal rates over long distances. Genetic data are also often (but not always) easier and faster to collect than detailed demographic data. These advantages can outweigh the potential difficulties with genetic data, depending on the parameters ofinterest.
The first genetic estimates of dispersal (see Evolutionary Algorithms were derived from Sewell-Wright's equation
e nFst j where Ne is effective population size (an estimate of the number of individuals in an idealized population that would show the same patterns of genetic diversity or levels ofinbreeding), m is migration (dispersal) rate, and FsT is a measure of population structure. This equation can be used to estimate dispersal among populations using DNA sequence data, DNA markers of variable length, or allozyme (protein) markers. However, more recently the use of this equation to estimate dispersal rates has come into question based both on unlikely assumptions made when calculating Fst, and on the applicability of the above equation to natural populations as opposed to idealized populations. Some of the assumptions made include equal and constant population sizes, and equilibrium between gene flow and genetic drift (stochastic variation in genetic frequencies over time). In addition, FST-based methods do not distinguish between historical and contemporary gene flow. In some cases, however, as in well-established, large populations, FST-based methods may be sufficient for estimates of dispersal rates.
Parentage analyses and assignment methods are both techniques that use genetic marker data to estimate dispersal (see Behavioral and Ecological Genetics). They both assume that source populations are discrete and that there is no genetic linkage disequilibrium (nonrandom associations of genetic marker variants due to inbreeding or chromosomal proximity) among the different markers used. Parentage analyses are based on multilocus genotypes, and can generate data about local animal movement and both seed and pollen dispersal. However, parentage analyses require extensive sampling to ensure that all possible parents in the source area have been included in the study. If a parent present in source populations is not sampled, dispersal from an unsampled (possibly quite far away) population may be inferred, potentially altering dispersal estimates. Depending on the situation, parentage analysis can be expensive enough to outweigh the potential advantages of genetic analyses over demographic studies.
Assignment methods, on the other hand, use allele frequencies, or frequencies of different variants of genetic markers, to predict from which source a particular individual came. This means that exhaustive sampling of source populations is unnecessary. It is still desirable to have representatives from all possible sources. In order to distinguish sex-biased dispersal, sex-specific markers must be used. Assignment methods assume discrete source populations and no linkage disequilibrium, and often assume equilibrium of populations. However, methods that enable dispersal estimates when populations are not in equilibrium are being developed.
A method that has come into use recently for many questions in genetics involves Bayesian analyses (see Bayesian Networks). Bayesian methods can be used in nonequilibrium situations, such as during range expansions, with high levels of inbreeding, or with unequal population sizes. Information known about the populations in question is entered in the analysis in the form of prior probability distributions. This information, commonly known as a prior, is basically a guess about how the populations might act based on data already available, such as experimental data. If no information is already known about the populations, an uninformative prior can be used. Then the genetic data is used, in conjunction with the prior, to calculate posterior probabilities of the data using a maximum likelihood algorithm, given the parameters currently in place in the maximum likelihood model. Markov chain Monte Carlo resampling is then used to explore parameter space and find values that optimize the fit ofthe parameters in the model to the data. The accuracy of detecting dispersal events with these methods is still being explored. Factors affecting accuracy of dispersal detection include the level of genetic diversity in the populations, amount of dispersal occurring among populations, how many genetic markers are used, and the level of variability in the markers themselves. Depending on the population structure in the system in question, these methods may be equally viable both for detection of local dispersal and longdistance dispersal events.
As mentioned in the beginning of this article, dispersal in the context of population dynamics usually refers to the dispersal of diploid individuals. However, especially when genetic methods are used, movement of haploid gametes can affect dispersal estimates. Male and female gametic dispersal patterns often differ, and if not accounted for, can skew dispersal estimates from genetic data. Nuclear genetic markers come from both paternal and maternal sources, due to the combination of nuclear genetic material during fertilization. If gametic gene flow is significantly different than diploid dispersal patterns, such as with pollen in many plants, care must be taken not to confuse gametic patterns of gene flow with movements of diploid individuals. In plants, for example, the male pollen often travels farther than seeds, especially if the species in question is wind-pollinated. In order to distinguish the dispersal of diploid seeds from haploid pollen, sex-specific genetic markers must be used. In plants, the chloroplast genome is generally (but not always) maternally inherited, and comparing patterns of genetic differentiation between the chloroplast and the nucleus can be used to estimate pollen versus seed dispersal patterns. If only seed dispersal is of interest, it may be sufficient to focus on chloroplast genetic markers, if the chloroplast is indeed maternally inherited in the species in question. Similarly, markers based on mitochondrial DNA will only show dispersal patterns of the female in animals. If dispersal patterns of both males and females are desired, both nuclear and mitochondrial markers must be used.
Because different analysis techniques have different strengths, weaknesses, and assumptions, it is important to consider what the goals of a given study are, and what methods are best suited to the questions at hand. Genetic methods hold promise for dispersal estimations, as data can be gathered relatively quickly and with less cost than demographic data. However, depending on the species and/or populations under consideration, estimates may require more genetic data than are currently available to optimize the parameters of a genetic analysis technique. If this is the case, then the cost of genetic analyses could equal or exceed the costs of demographic data collection. In addition, some data must be gathered through observation, such as detailed movement patterns among sites or populations and breeding success rates. Other data can best be estimated using genetic data, such as historical patterns of gene flow or long-distance dispersal events, depending on the organism in question. As techniques for measuring dispersal and its consequences improve, we will be better able to predict the survival, extinction, range expansion, and range contraction of populations and species. These predictions will improve estimates of the effects of habitat destruction and fragmentation on populations, a growing concern at multiple scales worldwide. In addition, accurate predictions of range expansion and contraction rates based on models of global warming will enable us to prepare for the possible effects of a rapidly changing climate on both marine and terrestrial species.
Was this article helpful?
You Might Start Missing Your Termites After Kickin'em Out. After All, They Have Been Your Roommates For Quite A While. Enraged With How The Termites Have Eaten Up Your Antique Furniture? Can't Wait To Have Them Exterminated Completely From The Face Of The Earth? Fret Not. We Will Tell You How To Get Rid Of Them From Your House At Least. If Not From The Face The Earth.