Understanding how marine organisms will be affected by global change is of primary importance to ensure ecosystem functioning and nature contributions to people. This study meets the call for addressing how life‐history traits mediate effects of ocean acidification on fish. We built a database of overall and trait‐mediated responses of teleost fish to future CO2 levels by searching the scientific literature. Using a meta‐analytical approach, we investigated the effects of projected CO2 levels by IPCC for 2050–2070 and 2100 on fish eco‐physiology and behavior from 320 contrasts on 42 species, stemming from polar to tropical regions. Moreover, since organisms may experience a mosaic of carbonate chemistry in coastal environments (e.g., in estuaries, upwelling zones and intertidal habitats), which may have higher pCO2 values than open ocean waters, we assessed responses from additional 103 contrasts on 21 fish species using pCO2 levels well above IPCC projections. Under mid‐century and end‐of‐century CO2 emission scenarios, we found multiple CO2‐dose‐dependent effects on calcification, resting metabolic rate, yolk, and behavioral performances, along with increased predation risk and decreased foraging, particularly for larvae. Importantly, many of the traits considered will not confer fish tolerance to elevated CO2 and far‐reaching ecological consequences on fish population replenishment and community structure will likely occur. Extreme CO2 levels well above IPCC projections showed effects on fish mortality and calcification, while growth, metabolism, and yolk were unaffected. CO2 exposures in short‐term experiments increased fish mortality, which in turn decreased in longer‐term exposures. Whatever the elevated CO2 levels considered, some key biological processes (e.g., reproduction, development, habitat choice) were critically understudied. Fish are an important resource for livelihoods in coastal communities and a key component for stability of marine ecosystems. Given the multiple trait‐mediated effects evidenced here, we stress the need to fill the knowledge gap on important eco‐physiological processes and to expand the number and duration of ocean acidification studies to multi‐generational, multiple stressor (e.g., warming, hypoxia, fishing), and species interactions experiments to better elucidate complex ecosystem‐level changes and how these changes might alter provisioning of ecosystem services.

Robust assessments of taxonomic and functional diversity are essential components of research programmes aimed at understanding current biodiversity patterns and forecasting trajectories of ecological changes. Yet, evaluating marine biodiversity along its dimensions is challenging and dependent on the power and accuracy of the available data collection methods. Here we combine three traditional survey methodologies (underwater visual census strip transects [UVCt], baited underwater videos [BUV] and small-scale fishery catches [SSFc]), and one novel molecular technique (environmental DNA metabarcoding [eDNA]-12S rRNA and cytochrome oxidase subunit 1 [COI]) to investigate their efficiency and complementarity in assessing fish diversity. We analysed 1,716 multimethod replicates at a basin scale to measure the taxonomic and functional diversity of Mediterranean fish assemblages. Taxonomic identities were investigated at species, genus and family levels. Functional identities were assessed using combinations of morphological, behavioural and trophic traits.We show that: (a) SSFc provided the higher taxonomic diversity estimates followed by eDNA, and then UVCt and BUV; (b) eDNA was the only method able to gather the whole spectrum of considered functional traits, showing the most functionally diversified and least redundant fish assemblages; and (c) the effectiveness of eDNA in describing functional structure reflected its lack of selectivity towards any considered functional trait. Our findings suggest that the reach of eDNA analysis stretches beyond taxon detection efficiency and provides new insights into the potential of metabarcoding in ecological studies.