Migration dynamics simulation of migratory fish in rivers

Abstract

The decline of migratory fish stocks is a major problem facing the global aquatic ecosystem. Understanding the dynamics of fish migration and their entire life history is crucial for quantitatively assessing the impact of natural and anthropogenic barriers on migratory fish and to understand the mechanisms of their decline. However, this understanding is largely ambiguous due to limited field survey data in the absence of models. Although Huang and Wang (Curr. Biol. 28, 2018, 3640–3647) and Huang (iScience 19(9), 2019, 772–785) proposed a Migration Dynamics Model (MDM) to describe the upriver and downriver migration processes of the Chinese sturgeon, but a crucial intermediate link—the breeding aggregation process—is still missing, while the standard Crank-Nicolson scheme, a common computational method, of finite difference methods cannot adapt to the model parameters changing with time and space. Here we propose a six-point implicit scheme to address the complexity of model parameters. The parameters of migration speed and diffusion coefficient are decomposed into fish-related and current-related components for analyzing their effects on model parameters and estimating them. Additionally, we propose a parameterization method for the aggregation process of spawners during the breeding season by introducing a set of clustering functions. Next, we examine the complete migration processes of adult and juvenile Chinese sturgeons in the Yangtze River and elucidate the underlying mechanisms behind the fish migration dynamics, including their spatiotemporal distributions, aggregation behavior, and migration schedule. We discover that the Gezhouba Dam (GD) significantly impacted the breeding aggregation behavior of Chinese sturgeon during the breeding season. The GD altered the distribution of fish along their migration path and schedule. Prior to the construction of the dam, 99.9 % of the total Chinese sturgeon were able to reach the 240 km-long spawning ground. However, after the dam was built, only 36.7 % were able to reach the 30 km-long spawning ground below the GD on the day of mating. Finally, we discuss the aggregation phenomenon and its driving factors during Chinese sturgeon migration, and categorize it into three types, namely biotic-driven, abiotic-driven and dual-driven aggregation behaviors. We reveal the existence of aggregations for non-biological reasons, such as the slowdown of fish migration speed induced by natural and anthropogenic barriers and estuarine tidal effects.