Stocking alien carp leads to regime shifts in native fish populations: Evidence from long-term observation and ecological modeling of a Chinese reservoir

Alien species invasions are considered to be one of the multiple key drivers that trigger ecological regime shifts in ecosystem structure and function. Predicting population collapse and understanding the self-reinforced feedback mechanisms that erode the resilience of native species are two major challenges in invasion ecology. However, to date, empirical evidence of species invasion-induced regime shifts remains scarce, and the minimum required data to generate early warning signals (EWSs) before critical transitions remains unclear, despite its importance for ecosystem management. By combining 80-year (i.e., 1936–2016) fishery data from Lake Qiandao and a theoretical competition model, we provide evidence that relative to nutrient enrichment and climate change, stocking alien fish species of silver carp (Hypophthalmichthys molitrix) and bighead carp (Aristichthys nobtilis) is the most important driver leading to the collapse of local fish populations. Detectable EWSs indicated by the variance, autocorrelation and composite resilience indicators require at least ten years of data before an abrupt decline, demonstrating that a long-term ecological monitoring program is necessary for providing insight into resilience dynamics. A structural equation model (SEM) suggests that the combined effects of eutrophication, high water levels and carp stocking-induced water clarity decline may play an important role in the reduced fitness of local fish populations, which subsequently increase the carrying capacity of alien fish populations. Our results demonstrate that stocking alien carps result in competitive exclusion of local fish populations, especially in the current context of water eutrophication and global climate changes. Furthermore, quantifying the minimum required time series length with detectable EWS across various ecosystems should be further studied given that this information is critical for flexible management policies to take actions to avert the harmful transitions.

Abstract licence: CC BY-NC

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.