Premise of the study: Living collections maintained for generations are at risk of diversity loss, inbreeding, and adaptation to cultivation. To address these concerns the zoo community uses pedigrees to track individuals and implement crosses that maximize founder contributions and minimize inbreeding. Using a pedigree management approach in an exceptional plant, we demonstrate how conducting such strategic crosses can minimize genetic issues that have arisen under current practices.

Methods: We performed crosses between these collections and compared the fitness of progeny, including plant performance and reproductive health. We genotyped the progeny and paternal accessions to measure changes in diversity and relatedness within and between accessions.

Results: The mean relatedness among individuals of an accession, suggests they are full siblings. As a result there was high inbreeding and low diversity within an accession, although less so among accessions. Progeny from the wider crosses had increased genetic diversity, while selfed accessions were smaller and less fertile.

Discussion: Institutions which hold exceptional species should consider how diversity is maintained within their collections. Implementing a pedigree-based approach to managing reproduction of ex situ plants will slow the inevitable loss of genetic diversity and in turn, result in healthier collections.

Genomic DNA was extracted from parental accessions using the 2× CTAB method (Doyle and Doyle, 1987) and the progeny of the aforementioned crosses using DNeasy Plant Mini Kit (Cat 69104, Qiagen, Germantown, MD, USA). For the parental accession plants, we extracted DNA from the maternal accession (CBG), and for three of the four paternal accessions (SDZ (Accession 2001-0273-022), NTBG (Accession 990833.3) and USBG1 (Accession 2014-0070)), however as the fourth accession (USBG2) died before we could extract DNA, we used progeny derived (Accession 2019-000) from a self of that plant to estimate paternal contribution. All samples were genotyped using markers and protocols described in Fant et al. (2019) which had already been tested for Hardy-weinberg and presence of null alleles. For this study we used a total of 11 primers; eight primers (B05, B08, B43, B44, B46, B47, B51, B57) were designed for B. insignis (Fant et al., 2019) and three primers (L23, L33, L34) were designed for Lobelia villosa (Rock) H. St. John & Hosaka. The eight B. insignis primers were visualized using pre-labeled forward primers with either WellRed Black (D2), Green (D3) or Blue (D4) fluorescent dye (Sigma-Proligo, St. Louis, Missouri, USA), while for Lobelia villosa markers, the forward primer was modified at the 5’ end (5’-CACGACGTTGTAAAACGAC-3’) so they could be labelled separately (Schuelke, 2000). All products were analyzed and scored using a CEQ 8000 Genetic Analysis System V9.0 (Beckman Coulter, Brea, California, USA). Given that B. insignis is a paleotetraploid (Lammers, 1988), four of the primer pairs (B44, B47, B51 and L23) produced more than two bands. As the peaks were separated from other alleles by large range (20-30bp), and segregated independently (Fant et al., 2019), they were scored as independent loci.

Ready to be used in genalex.