1. Understanding adaptive reservoir management strategies that balance ecological outcomes with other objectives necessitates properly articulated environmental objectives. Aside from flood pulse extent-related metrics, residual-based descriptors provide robust descriptions of fish assemblage structure and harvest.

2. We proposed a model framework based on spectral analysis of hydrologic variation and the Multivariate AutoRegressive State Space (MARSS) model to statistically quantify the effect sizes of hydrologic variation impacts on total fish catch and functional group (migration types) fish harvest and applied it to 17 years of fish harvest data from the Lower Mekong River Basin (LMB).

3. Our findings suggest that duration and timing of hydrologic anomalies matter as much as their magnitude. Anomaly droughts coupled with strong pulse can benefit species if timed correctly. Longitudinal migrators were more sensitive to anomalous floods and droughts than other migratory species. Fish catch projections using effect sizes derived from historical data revealed that a well-timed and protracted flood followed by a powerful flood pulse would be advantageous to the fishery, but a flood delay could negate such benefits.

4. Our results add to a growing body of research that suggests ecological flows can be engineered. For most dams, the rule curve describing reservoir releases and resulting downstream hydrograph is designed in an ecological vacuum in which the objectives are to maximize human services—power production, flood control or navigation. Our work demonstrates that hydrograph can be designed to manage aspects of functional biodiversity directly. Though the exact shape of our hydrograph may not apply to other engineered river systems, we suggest that the approach can be applied generally, and globally both to developed and developing river basins. Specifically, a functional biodiversity rule curve could be optimized as an additional objective function in a multi-objective optimization framework. Our methodology provides a quantitative method for deriving an ecological flow for this larger tradeoff analysis.

Data used in this analysis were from commercial fishing sampling activity and were provided by Mekong River Commission and the Inland Fisheries Research and Development Institute (IFReDI) of Cambodia. For detailed processing procedure, refer to the Methods section in the main article and Supplemental lnformation.

Data are uploaded in CSV format and can be opened with the open-source R free software environment as the article analysis was done.