Classification Importance of Seed Morphology and Insights on Large-Scale Climate-Driven Strophiole Size Changes in the Iberian Endemic Chasmophytic Genus <i>Petrocoptis</i> (Caryophyllaceae)

Abstract

Recruitment poses significant challenges for narrow endemic plant species inhabiting extreme environments like vertical cliffs. Investigating seed traits in these plants is crucial for understanding the adaptive properties of chasmophytes. Focusing on the Iberian endemic genus Petrocoptis A. Braun ex Endl., a strophiole-bearing Caryophyllaceae, this study explored the relationships between seed traits and climatic variables, aiming to shed light on the strophiole’s biological role and assess its classificatory power. We analysed 2773 seeds (557 individuals) from 84 populations spanning the genus’ entire distribution range. Employing cluster and machine learning algorithms, we delineated well-defined morphogroups based on seed traits and evaluated their recognizability. Linear mixed-effects models were utilized to investigate the relationship between climate predictors and strophiole area, seed area and the ratio between both. The combination of seed morphometric traits allows the division of the genus into three well-defined morphogroups. The subsequent validation of the algorithm allowed 87% of the seeds to be correctly classified. Part of the intra- and interpopulation variability found in strophiole raw and relative size could be explained by average annual rainfall and average annual maximum temperature. Strophiole size in Petrocoptis could have been potentially driven by adaptation to local climates through the investment of more resources in the production of bigger strophioles to increase the hydration ability of the seed in dry and warm climates. This reinforces the idea of the strophiole being involved in seed water uptake and germination regulation in Petrocoptis. Similar relationships have not been previously reported for strophioles or other analogous structures in Angiosperms.

Keywords

• chasmophyte
• cliff environment
• Iberian Peninsula
• LMEs
• local adaptation
• machine learning