An analytical investigation of the dynamics of inbreeding in multi-generation supportive breeding

Abstract

Supportive breeding is being increasingly usedas a measure to reduce the short-termprobability of extinction of populations withhighly reduced abundance relative to historicallevels. In this paper, we provide a conceptualframework and analytical tools to computechanges in inbreeding coefficient (F) in thecase of supportive breeding over any number ofgenerations. The dynamics of inbreedingcoefficients were investigated by means of asystem of recurrence equations. We focussed onquantifying the dynamics of F for specificcombinations of parameter values in terms ofthe effects of captive population census size,refreshment rate of breeders in captivity,scale of supplementation program, and migrationrate. We observed that supplementation did notalways result in substantial inbreedingincrement and several conditions loweredoverall inbreeding relative to controlsituations without supplementation. The censussize of captive populations was the single mostimportant controllable parameter determiningthe genetic consequences of supportivebreeding. While the proportion of captivebreeders brought into captivity from the wildbore a complex relationship to inbreedingcoefficient dynamics, the results indicatedthat managers should generally aim at highrefreshment rates (that is, large proportionsof their captive stock originating from thewild). This is especially important when asmall captive population is expected tocontribute large numbers of breeders to thesupplemented population. The analysis alsoshowed how supplemented populations connectedto a large metapopulation through gene flowrecover from the genetic risks of inbreedingdue to supportive breeding program more quicklythan isolated populations. The results of thisstudy join those of an increasing number ofinvestigations showing that supportive breedingdoes not always increase inbreeding, and mayeven decrease it in several circumstances.However, supportive breeding systems arecomplex, and results such as presented hereshould not be used in isolation, but inconsideration of other issues such as theconsequences on long-term fitness of wildindividuals.