TY - JOUR
T1 - Disentangling genetic structure for genetic monitoring of complex populations
AU - Milligan, Brook G.
AU - Archer, Frederick I.
AU - Ferchaud, Anne Laure
AU - Hand, Brian K.
AU - Kierepka, Elizabeth M.
AU - Waples, Robin S.
N1 - Publisher Copyright:
© 2018 The Authors. Evolutionary Applications published by John Wiley & Sons Ltd
PY - 2018/8
Y1 - 2018/8
N2 - Genetic monitoring estimates temporal changes in population parameters from molecular marker information. Most populations are complex in structure and change through time by expanding or contracting their geographic range, becoming fragmented or coalescing, or increasing or decreasing density. Traditional approaches to genetic monitoring rely on quantifying temporal shifts of specific population metrics—heterozygosity, numbers of alleles, effective population size—or measures of geographic differentiation such as FST. However, the accuracy and precision of the results can be heavily influenced by the type of genetic marker used and how closely they adhere to analytical assumptions. Care must be taken to ensure that inferences reflect actual population processes rather than changing molecular techniques or incorrect assumptions of an underlying model of population structure. In many species of conservation concern, true population structure is unknown, or structure might shift over time. In these cases, metrics based on inappropriate assumptions of population structure may not provide quality information regarding the monitored population. Thus, we need an inference model that decouples the complex elements that define population structure from estimation of population parameters of interest and reveals, rather than assumes, fine details of population structure. Encompassing a broad range of possible population structures would enable comparable inferences across biological systems, even in the face of range expansion or contraction, fragmentation, or changes in density. Currently, the best candidate is the spatial Λ-Fleming-Viot (SLFV) model, a spatially explicit individually based coalescent model that allows independent inference of two of the most important elements of population structure: local population density and local dispersal. We support increased use of the SLFV model for genetic monitoring by highlighting its benefits over traditional approaches. We also discuss necessary future directions for model development to support large genomic datasets informing real-world management and conservation issues.
AB - Genetic monitoring estimates temporal changes in population parameters from molecular marker information. Most populations are complex in structure and change through time by expanding or contracting their geographic range, becoming fragmented or coalescing, or increasing or decreasing density. Traditional approaches to genetic monitoring rely on quantifying temporal shifts of specific population metrics—heterozygosity, numbers of alleles, effective population size—or measures of geographic differentiation such as FST. However, the accuracy and precision of the results can be heavily influenced by the type of genetic marker used and how closely they adhere to analytical assumptions. Care must be taken to ensure that inferences reflect actual population processes rather than changing molecular techniques or incorrect assumptions of an underlying model of population structure. In many species of conservation concern, true population structure is unknown, or structure might shift over time. In these cases, metrics based on inappropriate assumptions of population structure may not provide quality information regarding the monitored population. Thus, we need an inference model that decouples the complex elements that define population structure from estimation of population parameters of interest and reveals, rather than assumes, fine details of population structure. Encompassing a broad range of possible population structures would enable comparable inferences across biological systems, even in the face of range expansion or contraction, fragmentation, or changes in density. Currently, the best candidate is the spatial Λ-Fleming-Viot (SLFV) model, a spatially explicit individually based coalescent model that allows independent inference of two of the most important elements of population structure: local population density and local dispersal. We support increased use of the SLFV model for genetic monitoring by highlighting its benefits over traditional approaches. We also discuss necessary future directions for model development to support large genomic datasets informing real-world management and conservation issues.
KW - density
KW - dispersal
KW - genetic monitoring
KW - isolation by distance
KW - multiple merger coalescent
KW - population structure
KW - spatial Λ-Fleming-Viot model
KW - Λ-coalescent
UR - http://www.scopus.com/inward/record.url?scp=85044233124&partnerID=8YFLogxK
U2 - 10.1111/eva.12622
DO - 10.1111/eva.12622
M3 - Article
AN - SCOPUS:85044233124
SN - 1752-4563
VL - 11
SP - 1149
EP - 1161
JO - Evolutionary Applications
JF - Evolutionary Applications
IS - 7
ER -