Fine-tuning functional syndromes for stressful environments: lessons on survival from the South African resurrection plant Myrothamnus flabellifolia

Resilience to abiotic stress is associated with a suite of functional traits related to defense and longevity. Stress tolerant plants are generally slow growing with extended leave lifespans and reduced allocation to reproduction. Resurrection plants are ideal systems to test for trade-offs associated with stress tolerance due to their extreme resiliency. While, growth defense trade-offs are well-characterized, few studies have tested for natural variation associated with tolerating the harshest environments. Here, we surveyed a suite of functional traits related to stress tolerance, leaf economics, and reproductive allocation in natural populations of the South African resurrection plant Myrothamnus flabellifolia. We selected three distinct field sites in South Africa ranging from mesic to xeric. Despite considerable environmental variation across the study area, M. flabellifolia plants were extremely and similarly stress tolerant at all sites. However, we detected notable variation in other life history and morphological traits. Plants in more mesic sites were larger, faster growing, and had more inflorescences. In contrast, plants from the most xeric sites appeared to invest more in persistence and defense, with lower growth rates and less reproductive allocation. Together, this suggests that desiccation tolerance is a binary trait in M. flabellifolia with little natural variation, but that other phenotypes are more labile. The trait syndromes exhibited by plants at the different study sites align with general expectations about growth defense tradeoffs associated with the colonization of extreme environments. We show that plants from the least stressful sites are more reproductive and faster growing, whereas plants from the most stressful sites were slower growing and less reproductive. These findings suggest that M. flabellifolia plants are finely tuned to their environment.