Effects of source and exposure environment on vegetative and reproductive traits
Vegetative and reproductive traits responded differently to source and exposure environments in the greenhouse, as expected based on the fundamental relationship of each type of trait with overall fitness. Reproductive traits corrected for biomass showed stronger source effects than vegetative traits, and less variability across environmental treatments. According to evolutionary theory, traits with the strongest impact on fitness should show evolutionary conservation (Scheiner 1993, Stearns & Kawecki 1994, Sih 2004). In parallel, the demographic buffering theory predicts that the most influential processes in species life cycles should be maintained relatively constant around local optimal values, to reduce variation in population growth rates (Pfister 1998, Burns et al . 2010, Hilde et al . 2020). For short-lived plants like P. lanceolata , reproduction has been identified as the most influential fitness component (Silvertownet al . 1996, GarcĂ­a et al . 2008, Shefferson and Roach 2012), which may explain the smaller role of plasticity and the higher consistency in genetic differentiation found for biomass corrected reproductive traits.
Stronger genetic differentiation in reproductive investment seemed to be facilitated by a higher plasticity in vegetative traits, buffering short-term environmental perturbations (Scheiner 1993, Alpert & Simms 2002, Sih 2004). This phenomenon, known as fitness homeostasis, has been highlighted before as a mechanism for maintaining high individual performance across a range of environments (Sultan 1995, Richardset al . 2006). The adjustment of vegetative traits to environmental conditions was manifest in our greenhouse experiment in several ways, and is best exemplified by SLA patterns. SLA increased in the shade treatment to optimize light capture and decreased in dry conditions to reduce water loss through leaf surface, common plastic responses in herbaceous plants (Poorter et al . 2009, Dwyeret al . 2014). Remarkably, some effects of exposure treatments on SLA were opposed by source environment effects suggesting countergradient variation (sensu Conover & Schultz 1995), such as the positive effect of source Aridity combined with the negative effect of the dry treatment. This apparent contradiction possibly arises because water scarcity in populations from dry sites is compensated through selection for higher RSR and/or stomatal function.
The complex interplay between plasticity and genetic differentiation, and the trait-specific nature of environmental effects found in our study highlight the variety of strategies for plant response to local conditions (see also Albert et al . 2010b, Le Bagousse-Pinguetet al . 2015, Roybal & Butterfield 2019), but also the difficulty of assessing the mechanisms and drivers of trait variation. The trait patterns found in P. lanceolata , including countergradient variation, could be partly explained by the influence of additional drivers not considered in the analyses, such as nutrient availability or biotic interactions (Chevin & Lande 2015). Additionally, further research could be undertaken to disentangle genetic differentiation from unaccounted maternal environment effects.