Conclusions and Discussion - Genetic Diversity in Establishing Plant Populations: Founder Number and Geometry - page 337

such as pollen and offspring dispersal distances, rates of selfi ng (e.g.,
frequencies of distinct self-incompatibility alleles), age-specifi c mortality
and reproduction schedules. Earlier NEWGARDEN trials also demonstrated
that the pattern of introduction relative to the number of groups of
founders, spatial distribution of these groups, founder distance from
borders, and other factors interact with life history attributes in complex
ways. Contrary to some guideline references for re-establishment given in
the introduction, the concept that the more numerous and centralized the
founders the better may not be an optimal re-introduction strategy (and
see Ghazoul 2005). NEWGARDEN modeling with the goal of preserving
genetic diversity and promoting population growth will improve with the
accumulation of knowledge about the life history characteristics of the
target species. If, for the species of interest (or closely related species), one
does not have reliable information with regard to realistic values for the
various NEWGARDEN life history input variables, the best option might
be to introduce founders at varying densities and in different geometric
shapes approximating estimated input conditions of the NEWGARDEN
comparative trials yielding the best results.
Earlier, it was noted that geometric differences in founding events can
alter the potential adaptive landscapes between two otherwise identically
founded natural populations, and this holds for restoration populations
as well. The fact that the pattern of genetic diversity retention set by the
founding characteristics of a population persists into the future suggests that
the adaptive landscapes of such populations will differ through time as well.
Two such populations may, at some point, respond differently to the same
selection pressures, ultimately increasing differences in genetic diversity
between them. More rapid expansion of populations after a founding event
not only promotes the retention of allelic diversity, it promotes, at least to
some small degree, an increase of phenotypic variance (more permutations
of combinations of distinct alleles) for selection to act on. Such rapid growth
increases the likelihood that selection will be acting on individuals with a
greater range of fi tness values for various traits, and thus partly decreases
the likelihood that unique alleles may be lost because of selection against
idiosyncratic allele combinations of lower fi tness. NEWGARDEN analyses
of the effects of founder geometry can assist in lessening, at the least in
some small measure, idiosyncratic responses of this type in restoration
populations by providing information on introduction patterns that may
preserve more genetic diversity and/or promote higher population growth
rates.
Comparative NEWGARDEN trials can also be used to model what has
actually happened in an expanding restoration population, “recreating”
approximate patterns of establishment that have occurred, and also then
providing approximations of the genetic diversity that has been retained.