Comparison of different genetic distances to test isolation by distance between populations

Abstract

Studying isolation by distance can provide useful demographic information. To analyze isolation by distance from molecular data, one can use some kind of genetic distance or coalescent simulations. Molecular markers can often display technical caveats, such as PCR-based amplification failures (null alleles, allelic dropouts). These problems can alter population parameter inferences that can be extracted from molecular data. In this simulation study, we analyze the behavior of different genetic distances in Island (null hypothesis) and stepping stone models displaying varying neighborhood sizes. Impact of null alleles of increasing frequency is also studied. In stepping stone models without null alleles, the best statistic to detect isolation by distance in most situations is the chord distance D. Nevertheless, for markers with genetic diversities H<0.4-0.5, all statistics tend to display the same statistical power. Marginal sub-populations behave as smaller neighborhoods. Metapopulations composed of small sub-population numbers thus display smaller neighborhood sizes. When null alleles are introduced, the power of detection of isolation by distance is significantly reduced and D remains the most powerful genetic distance. We also show that the proportion of null allelic states interact with the slope of the regression of F/(1-F) as a function of geographic distance. This can have important consequences on inferences that can be made from such data. Nevertheless, Chapuis and Estoup's FreeNA correction for null alleles provides very good results in most situations. We finally use our conclusions for reanalyzing and reinterpreting some published data sets.

Detecting Wahlund effects together with amplification problems: Cryptic species, null alleles and short allele dominance in Glossina pallidipes populations from Tanzania

“…It is also noteworthy that removing the locus with null alleles (i.e. PAL26) did not significantly alter the slope, as expected with FreeNA corrected values when the frequency of null alleles stays below or equal to 20% (Séré et al, ), as it is the case here. Consequently, for the following analyses, we kept the data with 6 loci.…”

“…In some instances, panmixia agreement was measured and tested with Genepop . For small sample sizes and substantial amount of missing genotypes, unreliable results at some loci can indeed be observed with Fstat (Séré, Thévenon, Belem, & Meeûs, ). Tests were undertaken unilaterally in both directions and the bilateral p ‐value was obtained as p bilateral = p mini + 1 − p maxi , where p mini and p maxi were the smallest and biggest unilateral (in opposite directions) p ‐values observed respectively.…”

“…If D e is known, a rough proxy of dispersal ( δ ), in order of magnitude, can be computed as (e.g. Séré et al, ). The significance of the regression is typically tested with a Mantel test (Mantel, ) between the geographic distances and the genetic distance.…”

“…If such relationship exists, neighborhood size is Nb = 4πD e σ 2 = 1/b and immigration is N e m = 1/(2πb), where D e is the effective population density and σ 2 is the average of squared axial distances between adults and their parents (Rousset, 1997;Watts et al, 2007). If D e is known, a rough proxy of dispersal (δ), in order of magnitude, can be computed as δ≈2 Séré et al, 2017). The significance of the regression is typically tested with a Mantel test (Mantel, 1967) between the geographic distances and the genetic distance.…”

Deceptive combined effects of short allele dominance and stuttering: an example withIxodes scapularis, the main vector of Lyme disease in the U.S.A.

“…These problems, like all others associated with amplification, are locus specific (Van Oosterhout et al, 2004;De Meeûs et al, 2007;De Meeûs, 2018a) and thus lead to locus specific variation (namely, an increase) of F IS . A less well known, though well documented (Chapuis and Estoup, 2007;Séré et al, 2017b;Manangwa et al, 2019) effect of such amplification problems consists of an increase of Wright's F ST (Wright, 1965) that is commonly used to measure the degree of genetic differentiation between subpopulations.…”