COS 110-3 - Earlier flowering alters reproductive fitness of tundra plant species more than climate over decades of warming

Abstract: The Arctic is warming at an alarming rate, with estimates of 2-4x times the global average. Advancement in spring (i.e. flowering) phenology is well documented in rapidly warming tundra ecosystems with short growing seasons, as plants aim to ‘track’ optimal climate windows. While earlier flowering is often assumed beneficial, this is not always the case, as it may incur additional risks such as frost damage, increased herbivory and/or reduced pollinator synchrony. Furthermore, in previous studies that report a link between earlier flowering time and reproductive fitness (flower, fruit, seed production), it is often unclear to what degree the enhanced reproductive fitness is a direct response to warming temperatures and/or an indirect response to climate through shifting phenology. Finally, life history theory would predicts a unimodal relationship between flowering time and reproductive fitness, yet the prevalence of this in natural settings is uncertain. Here we used two long-term (30+ years total) field monitoring datasets on tagged individual plants from twelve tundra species in the Canadian Arctic as part of the International Tundra Experiment (ITEX). Phenology and reproductive fitness data were manually recorded via annual observations (repeat measurements) at Daring Lake, Northwest Territories (64.87,-111.58) and at Alexandra Fiord, Nunavut (78.83, -75.80). Observed plant species encompass a wide range of taxonomic diversity and plant functional types with circumpolar or circumboreal distributions. We ask: How does experimental and ambient warming influence the relationship between plant phenology and reproductive fitness (flower and fruit number) within tundra species? Do tundra plant species that advance their phenology with warming temperatures have altered (positive or negative) reproductive fitness outcomes? Do these outcomes reflect directional, stabilizing or divergent selection? We use Hierarchical Bayesian Structural Equation models to compare the direct and indirect effects of climate warming on phenology and reproductive fitness across species, sites and years. We find that warming, both experimental and ambient, drives earlier flowering across species, which leads to higher numbers of flowers and fruits produced, reflecting consistent, directional selection for earlier flowering phenology. Furthermore, this indirect effect of climate warming mediated through phenology was generally stronger (~2-3x magnitude) than the direct effect of climate on the number of flowers and fruits. This suggests that on average, the benefits of early flowering including increased development time and subsequent enhanced reproductive output may outweigh its risks. Overall, this work provides important insights into population-level consequences of phenological shifts in a warming Arctic over multi-decadal time scales.