Outbreeding depression and breeding system evolution in small, remnant populations of Primula vulgaris: consequences for genetic rescue

Abstract

Many species suffer from anthropogenic habitat fragmentation. The resulting small and isolated populations are more prone to extinction due to, amongst others, genetic erosion, inbreeding depression and Allee-effects. Genetic rescue can help mitigate such problems, but might result in outbreeding depression. We evaluated offspring fitness after selfing and outcrossing within and among three very small and isolated remnant populations of the heterostylous plant Primula vulgaris. We used greenhouse-grown offspring from these populations to test several fitness components. One population was fixed for the pin-morph, and was outcrossed with another population in the field to obtain seeds. Genetic diversity of parent and offspring populations was studied using microsatellites. Morph and population-specific heterosis, inbreeding and outbreeding depression were observed for fruit and seed set, seed weight and cumulative fitness. Highest fitness was observed in the field-outcrossed F1-population, which also showed outbreeding depression following subsequent between-population (back)crossing. Despite outbreeding depression, fitness was still relatively high. Inbreeding coefficients indicated that the offspring were more inbred than their parent populations. Offspring heterozygosity and inbreeding coefficients correlated with observed fitness. One population is evolving homostyly, showing a thrum morph with an elongated style and high autonomous fruit and seed set. This has important implications for conservation strategies such as genetic rescue, as the mating system will be altered by the introduction of homostyles.

Keywords: Fitness; Heterosis; Heterostyly; Homostyly; Inbreeding depression; Restoration biology.

Fig. 1

Outline of the province of Drenthe with an insert map of the Netherlands showing the location of this province in black. In the province outline, dots indicate the locations of all historically known P. vulgaris populations. Filled dots indicate the three extant study populations, with their abbreviated names, empty dots indicate populations that went extinct between 1950 and 2000 (n = 9). Distances between the study populations in kilometers are also shown

Fig. 2

Schematic overview of the pollination experiment from the natural populations (a, field) to the F1 offspring (b, common garden) to the selected experimental plants (c, greenhouse). Curved arrows (1, 2, 5–9) represent pollen flow while straight arrows represent seed (3) and plant (4) transfer. Note that (1) except selfing (5), only legitimate pollinations were performed (reciprocally) and (2) pollinations 5 and 6 are repeated over all experimental plants (not shown for clarity)

Fig. 3

Mean cumulative fitness (fruit set × seeds/seed set × mean seed weight) for a pins (n = 27–32) and b thrums (n = 30–32). Error bars represent the standard error. MSP manual self-pollination, WF within-family, WP within-population, BP between-population. Note that the autonomous self-pollination ‘treatment’ is not included, as there was no data on its fruit set

Fig. 4

Mean cumulative fitness calculated over all within-population outcrossing treatments in the three offspring populations relative to the inbreeding coefficient, F _IS, determined from 12 microsatellite loci