LB 24-208 - Widespread legacy effects in net primary production across US drylands.

Dryland functioning is strongly regulated by moisture availability and thus vulnerable to projected increases in climatic variability and extremes. However, dryland net primary productivity (NPP) shows far greater sensitivity to spatial than temporal variation in precipitation (PPT). This implies an incomplete understanding of the controls of temporal variation in NPP in drylands. Evidence suggest that previous-year conditions may produce lagged effects that amplify or constrain NPP responses to PPT, denoted as ‘legacy effects’, though the prevalence of these legacy effects across drylands remains unclear. Here, we asked two questions: 1) What is the prevalence of ‘legacy effects’ across western US drylands, and 2) how do these effects vary across climatic gradients? We hypothesized that legacy effects were a frequent control of temporal variation in current-year NPP, and that the strength of these effects were proportional to previous-year conditions but independent of spatial variation in climate. To this end, we built spatiotemporal models with different previous year “lag” effects and their interactions with spatial variations in climate using long term (30+ years) remotely sensed estimates of NPP and PPT. Understanding the patterns of legacy effects is important for agricultural practices and ecosystems management, particularly in light of increased climatic variability and extremes. We found widespread evidence for legacy effects across five different dryland ecoregions. Lag effects in our models explained a significant fraction of current-year NPP; however, previous-year NPP was a better predictor than previous-year PPT. Furthermore, even after accounting for current-year PPT, previous-year NPP was still relevant, improving predictions of current-year NPP across all ecoregions. The relationship between previous-year NPP and current-year NPP was consistently positive, regardless of ecoregion. This implies that a productive previous-year will tend to result in a productive current-year and vise-versa. This supports the hypothesis that legacy effects are proportional to the magnitude of previous-year conditions. Mechanistically, we suspect that biotic mechanisms are at play; if previous year NPP is high (or low), then an increase (or decrease) in leaf area and root production will enhance (or reduce) exploitation of resources for current-year NPP. Preliminary results further suggest that the strength of these lag effects in our models increases slightly with an increase in mean PPT, suggesting that the previously describe mechanisms are stronger at wetter sites. In total, our results suggest that previous-year NPP is a consistent control of current-year NPP in drylands at that the strength of this control increases with average moisture availability.