Plant phenology predicts the stability of primary production in three ecosystems of the northern Chihuahuan Desert

Abstract The temporal stability of plant productivity affects species' access to resources, exposure to stressors and strength of interactions with other species in the community, including support to the food web. The magnitude of temporal stability depends on how a species allocates resources among tissues and across phenological stages, such as vegetative growth versus reproduction. Understanding how plant phenological traits correlate with the long‐term stability of plant biomass is particularly important in highly variable ecosystems, such as drylands. We evaluated whether phenological traits predict the temporal stability of plant species productivity by correlating 18 years of monthly phenology observations with biannual estimates of above‐ground plant biomass for 98 plant species from semi‐arid drylands. We then paired these phenological traits with potential climate drivers to identify abiotic contexts that favour specific phenological strategies among plant species. Phenological traits predicted the stability of plant species above‐ground biomass. Plant species with longer vegetative phenophases not only had more stable biomass production over time but also failed to fruit in a greater proportion of years, indicating a growth–reproduction trade‐off. Earlier leaf‐out dates, longer fruiting duration and longer time lags between leaf and fruit production also predicted greater temporal stability. Species with stability‐promoting traits began greening in drier conditions than their unstable counterparts and experienced unexpectedly greater exposure to climate stress, indicated by the wider range of temperatures and precipitation experienced during biologically active periods. Our results suggest that bet‐hedging strategies that spread resource acquisition and reproduction over long time periods help to stabilize plant species productivity in variable environments, such as drylands. Read the free Plain Language Summary for this article on the Journal blog.