Rising ocean temperatures pose a continuing threat to marine fish communities. As warming has far‐reaching impacts at multiple ecological levels, incorporating multimodal data is necessary for more accurately forecasting the responses of species and communities to the warming ocean. Range shifts, life‐history changes, and alterations of trophic dynamics are three important aspects of warming impacts, yet there has not been a formal integration of all three aspects in the same analysis. Here, we present a novel framework that integrates species distribution projections, life‐history changes, and food web dynamics to assess warming impacts on marine fish communities. We first introduce a simple yet effective way of incorporating thermal physiological data into the species distribution model without the need to empirically measure thermal performance curves. We then use the dynamic size‐spectrum model as the modeling backbone to incorporate data from species distributions and population‐level life history analyses. With this framework, we evaluate how individual species are affected under two warming scenarios (RCP4.5 and RCP8.5). We also simulate large‐scale top‐down and bottom‐up perturbations to examine community resilience under rising temperatures. We find that warming generally reduces species biomass and shifts species size spectra towards larger individuals, even though the maximum size tends to decrease under warming. However, the exact responses to rising temperatures differ among species and do not exhibit strong correlations with species size and the pace of life history. More severe warming also renders the focal community more vulnerable to top‐down perturbations, even though the community remains sufficiently resilient overall. The complex nature of species and community responses result from the fact that distribution range, life history, and food web dynamics change with warming in different directions that may not be intuitive to predict a priori. Importantly, we show that neglecting changes in species distribution or life history will lead to biased assessment of species and community responses. Our analyses highlight trophic dynamics, species biomass, and community resilience as three emergent properties that our framework can uniquely quantify. This integrative framework is readily applicable to other communities of interest and can be scaled up for multi‐regional or global analyses.

Accounting for biotic interactions is important for predicting species and ecosystem variation under changing climate but difficult to achieve in practice. The proportion of geographical overlap between species, called species geographical sensitivity (SGS), could be used to gauge the potential for species interactions. Species with increasingly high SGS could have the potential to experience more interactions with other species and vice versa, which might have important implications in ecological assessment, particularly at a community level, in the face of climate change. We compiled fish occurrences in the North Sea from 1983 to 2020 and calculated annual mean SGS (mSGS) to systematically evaluate their temporal changes and to estimate influences of species traits on the relative temporal changes in mSGS. The results showed that 36.3% of species significantly changed their mSGS over time, with high correlations between changes in species range size and overlap with other species. The species’ averaged mSGS before warming was highly correlated with the relative change in mSGS. Depth range, body length, and age at maturity together explained most variation in mSGS among these species. Contemporary climate change is expected to reorganize species distributions and interactions and substantially alter marine ecosystem functioning. Our assessment opens a new avenue for evaluating climate change impacts on species geographical interactions, and such geographical changes may be contingent on species traits.

Environmental and climatic changes are expected to redistribute species, altering the strengths of species interaction networks; however, long-term and large-scale evaluations remain elusive. One way to infer species interaction networks is by analyzing their geographical overlaps, which provides indices of species interdependence, such as mean spatial robustness (MSR), which represents the geographical impact of a species on other species, and mean spatial sensitivity (MSS), which indicates how a species is influenced by other species. Integrating MSR and MSS further allows us to assess community coexistence stability and structure, with a stronger negative relationship between MSR and MSS (i.e., species are unequally dependent on each other) within a community at a given time suggesting a more stable community. Here, we assessed multidecadal changes in adult marine fish communities using bottom trawl datasets across latitudes from 1982 to 2011 in the Eastern US Continental Shelf, North Sea, and Eastern Bering Sea. Consistent, significant long-term increasing temporal trends of MSR and MSS were found in all three large marine communities. MSR exhibited strong correlations with species’ range sizes, especially in high-latitude communities, while MSS was strongly positively correlated with species’ median proportion of overlap with interacting species. The relationships between MSR and MSS were generally negative, indicating stably coexisting fish communities. However, the negative relationships weakened over time, implying that the coexisting fish communities gradually became unstable. Our findings provide an assessment of changes in spatially geographical aspects of multiple species, for decades and at mid- to high latitudes, to allow the detection of global ecological changes in marine systems by alternative estimation of geographic overlaps of species interaction networks. Such species co-occurrence estimation can help stay vigilant of strategies for accelerating climate change mitigation particularly at coarser spatial scales.