Realistic time-lags and litter dynamics alter predictions of plant-soil feedback across generations

Plant-soil feedback is a critical process in natural plant communities. However, it remains unclear whether greenhouse-measured microbial effects manifest in natural systems with temporally separated growing seasons as classic experiments often overlook seasonal time lags and litter dynamics. We modified the classic two-phase experiment to study plant-soil feedback for three Californian annual plant species. Our response phase used soil inoculum obtained either immediately after plant conditioning, after a six-month dry period with the conditioning plant removed, or after a dry period with the litter of the conditioning plant. We characterized soil bacterial and fungal communities in different treatments and employed recent advancement in plant-soil feedback theory to predict plant coexistence. Temporal delays and the presence of litter caused distinct responses in the fungal and bacterial communities, resulting in divergent microbial compositions at the end of the response phases. The delayed response treatments also affected microbially mediated stabilization, fitness differences, and invasion growth rates differently across species pairs, influencing predictions of plant coexistence. Our study highlights that the interplay between seasonal delays and litter dynamics prevents the direct extrapolation of plant-soil feedback measurements across multiple seasons, emphasizing the necessity of considering natural history when predicting microbially mediated plant coexistence.