The Relationships Between Precipitation and Productivity Pulse Patterns in Rainfall-, Runoff-, and Flood-Driven Dryland Ecosystems - Israeli LTER Network Project

Background: Rainfall pulse patterns - defined by variability in the timing, intensity, frequency, and duration of precipitation events - play a critical role in shaping the functionality of all terrestrial ecosystems. These patterns regulate water availability, nutrient cycling, and primary productivity. Rainfall-dependent, runoff-dependent, and flood-dependent dryland ecosystems differ structurally and functionally due to their reliance on distinct hydrological processes triggered by rainfall pulses. These ecosystems serve as effective models for addressing a key scientific question: Do ecosystems with varying hydrological dependencies respond similarly to rainfall pulse patterns, or do they require distinct precipitation regimes? We hypothesize that ecosystems reliant on direct rainfall, runoff redistribution, or periodic flooding demand different rainfall pulse patterns - defined by specific intensity, duration, and seasonality characteristics - to sustain their functionality and productivity. Using the "Pulse-Reserve Paradigm" as a theoretical framework, we explored precipitation-productivity relationships across diverse water-limited ecosystems with distinct hydrological processes. Methods: The study was conducted at seven sites within the Israeli "Long-Term Ecological Research" (LTER) network, representing a climatic gradient from hyper-arid to dry sub-humid zones (40–700 mm annual precipitation). These sites encompass the three hydrological system types. Rainfall events were monitored from 2018 to 2024 using local meteorological stations, while primary productivity was assessed via Sentinel 2 satellite remote sensing of vegetation chlorophyll indices. Precipitation pulses were characterized by variables such as total yearly rainfall, rainfall event intensity, event duration, and seasonality. Productivity pulses, measured through time-series of spectral vegetation indices, included dimensions such as beginning, peak, and end of season. The relationships between precipitation and productivity pulse patterns were analyzed using machine learning models to identify the importance of rainfall pulse characteristics for each hydrological system type. Results: Rainfall-Dependent Systems : Productivity was positively correlated with higher total rainfall over the two preceding years, a later ending of the previous rainy season, an earlier start of the current rainy season, and higher average rainfall per event in the current year. These results highlight the critical role of rainfall timing and event intensity, where soil water reserves from previous years and substantial, early-season rain events drive the productivity. Rainfall-Runoff Systems : Productivity correlated negatively with higher total rainfall over the previous two years but positively with longer rainfall event durations and a later ending of the current rainy season. These findings suggest that extended rainfall events over a prolonged rainy season enhance runoff redistribution and soil moisture retention, while excessive cumulative rainfall may lead to resource losses from the system. Rainfall-Runoff-Flood Systems : Productivity was positively linked to longer time intervals between rainfall events, higher average rainfall per event, and higher total rainfall in both the current and previous years. These results indicate that infrequent but intense rainfall events are crucial for generating floods that redistribute water and nutrients and feed soil water reserves, supporting productivity in these systems. Discussion: The results reveal significant differences in the precipitation-productivity relationships across the hydrological system types, driven by variations in rainfall pulse characteristics. Rainfall- and flood-dependent systems benefit from intense rainfall events and cross-annual soil water reserves, whereas runoff- systems thrive under long moderate rain events extending to late season. These differences can be explained by the hydrological processes (e.g., surface runoff and flood dynamics) and vegetation structures (e.g., deep-rooted trees versus shallow-rooted shrubs and herbs) characteristic of each system. These features reflect the ecosystems' ability to maintain and utilize soil moisture reserves that are impacted by specific hydrological processes. Conclusion: Rainfall pulse patterns are fundamental to hydrological and biological productivity processes in dryland ecosystems. By shaping water availability, distribution, and movement, these patterns are critical determinants of ecosystem functionality and sustainability. Our findings align with the "Pulse-Reserve Paradigm" and "Source-Sink Dynamics" frameworks, providing valuable insights into the ecological dependencies of rainfall-, runoff-, and flood-driven ecosystems. Understanding these unique dependencies is essential for predicting and managing the impacts of climate change on rainfall regime in drylands specifically, and terrestrial ecosystems in general.