Ecological and biological information improves inferred paternity in a white spruce breeding orchard

Abstract

Biological, ecological, and genetic marker information was used to predict paternal (nF= 104) siring success for offspring (nO= 522) sampled over two years from two mother clones. Distance alone was predictive of siring success, whereas fecundity and a provenance indicator variable captured additional, but not all, remaining variation. Using additional nongenetic measures to predict siring success increased individual probabilities of paternity over a genetic-only model. Reproductive success of males was highly skewed, and not all successful males were consistently successful over years. Overall rate of selfing was 14% in the surviving (56%–63%) seedlings. The estimated number of (unsampled) sires outside of the seed orchard was highly variable, resulting in unassigned seed orchard fathers for 6% of the sampled progeny. Some benefits and limitations of using full-likelihood paternity analyses are discussed.