Aim: Species distribution models (SDMs) are an effective tool to explore the potential distribution of terrestrial, freshwater, and marine organisms; however, SDMs have been seldom used to model ichthyoplankton distributions and thus our understanding of how larval stages of fishes will respond to climate change is still limited. Here, we developed SDMs to explore potential impacts of climate change on habitat suitability of ichthyoplankton.

Location: Yangtze Estuary, China

Methods: Using long-term ichthyoplankton survey data and a large set of marine predictor variables, we developed ensemble SDMs for five abundant ichthyoplankton species in the Yangtze Estuary (Coilia mystus, Hypoatherina valenciennei, Larimichthys polyactis, Salanx ariakensis, and Chelidonichthys spinosus). Then, we projected their habitat suitability under present and future climate conditions.

Results: The ensemble SDMs had good predictive performance and were successful in estimating the known distributions of the five species. Model projections highlighted two contrasting patterns of response to future climates: while C. mystus will likely expand its range, the ranges of the other four species will likely contract and shift northward.

Main conclusions: According to our SDM projections, the five ichthyoplankton species that we tested in the Yangtze Estuary are likely to respond differently to future climate changes. These projected different responses seemingly reflect the differential functional attributes and life history strategies of these species. To the extent that climate change emerges as a critical driver of the future distribution of these species, our findings provide an important roadmap for designing future conservation strategies for ichthyoplankton in this region.

Ichthyoplankton samples were collected from 40 stations in the Yangtze Estuary between 1999 and 2013 (05/1999, 05/2001, 05/2004, 05/2007, 05/2009, 05/2010, 05/2011, 05/2012, and 05/2013). Presence records of five abundant ichthyoplankton species in this region were used to develope species distribution models.

To increase sample size, samples between 1999 and 2013 (05/1999, 05/2001, 05/2004, 05/2007, 05/2009, 05/2010, 05/2011, 05/2012, and 05/2013) were pooled together.