Sensitivity analysis of effective population size to demographic parameters in house sparrow populations

Abstract

In population management, the e ective population size, N e , can be viewed in tandem with actual population size, N , as the main factors determining a population's long-term viability and sustainability. N e is the number of individuals in an observed population that would lose genetic variation at the same rate as an ideal population. Understanding which demographic factors that a ect N e /N , will make resource allocation and decision making more e ective, either if the management goal is to maximise, maintain or minimize N e /N . The goal of this thesis was to calculate the demographic parameters that determine N e /N , following the method of Engen et al. (2010), and then determine which of these parameters N e /N is most sensitive to. In other words, determine which parameters that contribute most to the total variation in N e /N . This was done, using data on 13 Norwegian populations of house sparrows (Passer domesticus), including more than 4000 individuals, and spanning up to 20 years. To find which of the demographic parameters (demographic variance, generation time, stable age distribution, reproductive values, individual fecundity and survival) that a ect N e /N most, sensitivity analyses were carried out. Using the global variance-based Sobol' method, it was found that demographic variance, especially of older individuals, was the parameter that N e /N was most sensitive to. Generation time was found to be less important than demographic variance, which includes all the other parameters. The demographic variance of a population is determined by fecundity and survival on the individual level. The individual reproductive values were found to be most sensitive to fecundity, followed by survival. In contrast, the stable sex-age distributions, and the sex-age specific reproductive values, were found to be of little importance. For population management purposes, the results from this study show that resources should be focused on the manipulation of demographic variance in older individuals, more specifically their fecundity and survival. Even though these results are from insular populations of house sparrows, they may also apply to fragmented populations of other species with similar life histories and demography. i SammendragInnen populasjonsforvaltning er e ektiv populasjonsstørrelse, N e , sammen med observert populasjonsstørrelse, N , hovedfaktorene som avgjør overlevelse og hvor baerekraftig en populasjon er på sikt. N e er antallet individer i en observert populasjon som ville miste genetisk variasjon med samme rate som i en ideell populasion. A forstå hvilken faktorer som er med påå påvirke N e /N , vil gjøre både ressursallokering og avgjørelser innen forvaltning mer e ektive. Dette gjelder både for mål omå maksimere, opprettholde eller minimalisere N e /N . Formålet med denne oppgaven varå beregne de demografiske parameterne som inngår i N e /N , vedå følge metoden i Engen et al. (2010), for deretterå finne hvilke av disse parameterne N e /N er mest sensitiv til. Det vil si at man ide...

Measures of effective population size in sea otters reveal special considerations for wide‐ranging species

“…Typically, this is a result of linkage disequilibrium caused by the admixture of parents of different age mating and creating a two‐locus Wahlund effect (Nei & Li, ; Waples et al., ). In contrast, the consistent individual differences in lifetime reproductive success of sea otters, which would decrease the true N e (Lee, Engen, & Sæther, ; Stubberud et al., ), are not accounted for in demographic estimates of N e resulting in a potential overestimate (Lee et al., ; Staedler, ; Stubberud et al., ; Tarjan, ). We suggest that the combined effects of isolation by distance over a large geographic area, the influence of overlapping generations, and consistent individual variation in reproductive success, explain the difference between genetic and demographic range‐wide estimates of N e .…”

“…The theory and techniques to address the majority of these considerations are well established. However, unlike more established genetic methods, algorithms for calculating current effective population size (N e ) are still being developed and refined (Do et al, 2014;Waples, Antao, & Luikart, 2014), have stringent model assumptions (Wang, 2016;Waples et al, 2014), and different methods can result in different estimates (Baalsrud et al, 2014;Menéndez, Álvarez, Fernandez, Menéndez-Arias, & Goyache, 2016;Stubberud et al, 2017). N e is usually much less than the census population size (N c ); however, the extremely low N e /N c ratios reported for some species (e.g., marine fishes) are controversial (Waples, 2016).…”

“…Furthermore, N e can be calculated using genetic or demographic data. To date, demographic and genetic N e estimates have rarely been calculated on the same population, and when they have, there were inconsistencies between demographic and genetic N e estimates (Baalsrud et al, 2014;Stubberud et al, 2017;Waples et al, 2011).…”

A genome‐wide linkage map for the house sparrow (Passer domesticus) provides insights into the evolutionary history of the avian genome

“…The relationship between background LD and average effective population sizes for each of the eight populations was tested using a linear model in R version 3.5.3 (R Core Team, ). As estimates of N e for seven of the islands (Aldra, Gjerøy, Hestmannøy, Indre Kvarøy, Nesøy, Leka and Vega) we used the estimates of local demographic variance N e in Stubberud et al () (Table ). Importantly, these N e estimates account for effects of overlapping generations and environmental stochasticity on the rate of genetic drift, in addition to variation in population size, and among‐individual variation in survival and reproduction within the local population (Engen, Lande, Sæther, & Gienapp, ).…”

Estimating demographic contributions to effective population size in an age‐structured wild population experiencing environmental and demographic stochasticity

“…This calculation is potentially very insightful but imposes challenging data demands, requiring individual‐level information on realized sex‐ and age‐specific reproductive success and survival alongside population‐level mean rates. To date, it has only been implemented in a Siberian jay ( Perisoreus infaustus ) population (as a methodological example considering three age classes, Engen et al., ) and a house sparrow ( Passer domesticus ) metapopulation (considering two age classes, Stubberud et al., ) where genetic variation and inbreeding rates may be influenced by immigration rather than solely local demography (e.g. Baalsrud et al., ).…”