Inbreeding depression in three generations of selfed families of silver birch ( Betula pendula ) Tongli Wang, Risto Hagqvist, and Peter M.A. Tigerstedt Abstract: Inbreeding depression in survival and growth vigor was investigated through three generations of selfing on the basis of a 13-year-old field trial in silver birch (Betula pendula Roth). Survival showed significant decline from outcrossed (S 0 ) to first-generation selfed trees (S 1 ), and some recovery from second- (S 2 ) to third-generation selfed trees (S 3 ), indicating the rapid purging of lethal, sublethal, or highly deleterious alleles by selfing. The continuous decline of growth vigor with increased inbreeding suggests a low efficiency in purging mildly deleterious alleles affecting quantitative traits and a low possibility of recovery in growth with further inbreeding. Linear effects of inbreeding on survival and growth vigor (taken as height, diameter at breast height, and stem volume) explained the major variation caused by inbreeding and indicates that the additively combined dominant effects at different loci is the major source of inbreeding depression in this species. A weak and significant nonlinear effect of inbreeding on height and stem volume was also observed, suggesting the existence of epistatic effects. Résumé : Les auteurs ont étudié, à partir des résultats d’un test au champ d’une durée de 13 ans, les effets de la dépression consanguine sur la survie et la vigueur de croissance du bouleau verruqueux (Betula pendula Roth) sur trois générations d’autofécondation. La survie affichait une baisse significative entre le croisement sans consanguinité (S 0 ) et la descendance de première génération issue d’autofécondation (S 1 ), pour ensuite se rétablir en partie de la deuxième (S 2 ) à la troisième génération (S 3 ) issues d’autofécondation. Ce résultat implique l’élimination rapide des allèles létaux, sublétaux ou hautement délétères suite à l’autofécondation. La diminution continue de la vigueur de croissance avec l’augmentation de la consanguinité suggère une faible efficacité à éliminer les allèles légèrement délétères qui affectent les caractères quantitatifs, et une faible capacité de rétablissement de la vigueur de croissance si la consanguinité devait continuer. Les effets linéaires de la consanguinité sur la survie et la vigueur de croissance (exprimée par la hauteur, le diamètre à hauteur de poitrine et le volume de la tige) expliquaient la majeure partie de la variation causée par la consanguinité. Ce résultat indique que la source majeure de dépression consanguine chez cette espèce est constituée d’effets de dominance à différents loci, qui sont combinés de façon additive. Les auteurs ont également observé un effet non linéaire faible mais significatif de la consanguinité sur la hauteur et le volume de la tige, suggérant l’existence d’un effet d’épistasie. [Traduit par la rédaction] Wang et al. 668 Introduction Inbreeding depression (ID) is a common phenomenon in allogamous plant species. The interest of forest geneticists and tree breeders in ID is based on several objectives. A ma- jor question has been concerned with the extent and negative effects of self-pollination and other levels of inbreeding in seed orchards, particularly in advanced generations of breed- ing populations, in which the size of the population is get- ting smaller. A number of studies concerning such objectives have been conducted in major forest trees, such as in the co- nifers Pinus resinosa Ait. (Fowler 1965), Picea abies (L.) Karst. (Koski 1973), Pseudotsuga menziesii (Mirb.) Franco, Pinus ponderosa Dougl. ex Laws. (Sorensen and Miles 1982; Woods and Heaman 1989), Pinus radiata D. Don (Wilcox 1983), Pinus taeda L. (Sniezko and Zobel 1988), Pinus sylvestris L. (Koski 1973; Lundkvist et al. 1987), and Pinus elliottii Engelm.(Matheson et al. 1995); and in the broad-leaved trees Eucalyptus regnans F. Mull. (Griffin and Cotterill 1988; Moran et al. 1989) and Eucalyptus globulus Labill. (Hardner and Potts 1994). Theoretically, inbreeding increases among-family varia- tion (Robertson 1952; Falconer 1989). It thus uncovers ge- netic variability concealed in heterozygotes and brings it into the open where it can be acted upon by selection (Allard 1960). Inbreeding and crossing between inbred lines for pro- ducing high-yielding varieties has been greatly successful in crops, notably in maize (Zea mays L.). The possibility of us- ing inbred material in forest tree improvement has been in- vestigated and suggested by some forest geneticists and breeders. Wilcox (1983), Griffin and Cotterill (1988), and Hardner and Potts (1994) have suggested testing selfed lines for identification of superior parents, whereas Bingham (1973), Andersson et al. (1974), Lindgren (1975), Rudolph (1981), Sniezko and Zobel (1988), and Wang et al. (1996) have been interested in crossing inbred lines for producing Can. J. For. Res. 29: 662–668 (1999) © 1999 NRC Canada 662 Received September 18, 1998. Accepted March 2, 1999. T. Wang 1 and P.M.A. Tigerstedt. Department of Plant Biology, Box 27, FIN-00014 University of Helsinki, Finland. e-mail: peter.tigerstedt@helsinki.fi R. Hagqvist. Foundation for Forest Tree Breeding, Haapastensyrjä, Karkkilantie 193, FIN-12600 Läyliäinen, Finland. e-mail: Risto.Hagqvist@mjs.fi 1 Author to whom all correspondence should be addressed. Present address: Faculty of Forestry, 3609-2424 Main Mall, Vancouver, BC V6T 1W5, Canada. e-mail: tlwang@unixg.ubc.ca