Testing the effect of individual scent compounds on pollinator attraction in nature using quasi-isogenic Capsella lines.

Abstract

Floral scent, usually consisting of multiple compounds, is a complex trait, and its role in pollinator attraction has received increasing attention. However, disentangling the effect of individual floral scent compounds is difficult due to the complexity of isolating the effect of single compounds by traditional methods. Using available quasi-isogenic lines (qILs) that were generated as part of the original mapping of the floral scent volatile-related loci CNL1 (benzaldehyde) and TPS2 (β-ocimene) in Capsella, we generated four genotypes that should only differ in these two compounds. Plants of the four genotypes were introduced into a common garden outside the natural range of C. rubella or C. grandiflora, with individuals of a self-compatible C. grandiflora line as pollen donors, whose different genetic background facilitates the detection of outcrossing events. Visitors to flowers of all five genotypes were compared, and the seeds set during the common-garden period were collected for high-throughput amplicon-based sequencing to estimate their outcrossing rates. Benzaldehyde and β-ocimene emissions were detected in the floral scent of corresponding genotypes. While some pollinator groups showed specific visitation preferences depending on scent compounds, the outcrossing rates in seeds did not vary among the four scent-manipulated genotypes. The scent-manipulated Capsella materials constructed using qILs provide a powerful system to study the ecological effects of individual floral scent compounds under largely natural environments. In Capsella, individual benzaldehyde and β-ocimene emission may act as attractants for different types of pollinators.