Integration of life cycle and habitat conditions in modeling fish biomass in the floodplain of the Lower Mekong Basin

Migratory fishes utilize several distinct habitats during their life cycles, and thus the conservation of such species requires proper habitat management in both spatial and seasonal dimensions. However, their life cycle in the freshwater environment has not been well investigated, in particular in the tropical region. Therefore, our study aimed to elucidate the importance of the integration of life cycle and habitat conditions in modeling the biomass of migratory fish species. Focusing on the hydrological years 2000–2015, we collected the biomass of 74 migratory fish species and their seasonal habitat conditions in the floodplain of the Lower Mekong Basin, including Tonle Sap Lake. Then, generalized linear, generalized additive, and random forest models were applied to the collected data with and without life cycle information. As a result, the models and their performance indicate that the integration of the life cycle and seasonal habitat conditions in modeling the biomass is significantly important for migratory fishes. The results further showed that changing ecohydrological indices in the feeding and refuge seasons negatively affected the fish biomass. Those indices were hydrologic factors (the maximum water levels, the mean water levels, the start date of the flood, and the end date of the flood) and climate factors (water temperatures). Interestingly, the models indicated that the total area of inundated forests negatively influences the biomass of planktivorous fishes while it positively affects the biomass of piscivores. These results clearly showed the importance of a natural flow regime to ensure the fish life cycle and biomass production in the Lower Mekong Basin. The presented approach can be applied to model the biomass of migratory fishes in spatially connected and changing environments.