Globally observed impacts of climate change on marine organisms and ecosystems highlight the need to assess the risks and benefits of international mitigation commitments, such as the goal of limiting global warming to 1.5°C above pre-industrial levels. This requires information on species-specific thermal tolerance thresholds, lifecycle bottlenecks and the sensitivity of critical life stages to additional climate factors such as ocean acidification (OA), the CO2-driven decrease in seawater pH. However, this information is not available for many important fish species, including Atlantic cod (Gadus morhua) and Polar cod (Boreogadus saida). A general assumption is that species adapted to variable climates (e.g., Atlantic cod) have larger tolerance windows than those adapted to relatively stable conditions (e.g., Polar cod), and that egg-stages (embryos) are more vulnerable to temperature changes and OA than adults with fully functional organ systems for oxygen supply and acid-base homeostasis. As adults become sexually mature, temperature windows may narrow due to additional metabolic loads associated with gonad development. Accordingly, there is a risk that future warming and OA will affect the suitability of spawning habitats by exceeding the tolerance thresholds of embryos and/or spawning adults. In this thesis, experimental and meta-analytical investigations on lifecycle bottlenecks were used to describe physiological principles and to identify mitigation pathways that minimize climatic risks regionally (for Atlantic cod and Polar cod) and globally (for marine and freshwater species). The objective of the experimental part (Publication I-III) was to investigate the effects of OA (−0.4 pH, 400 vs 1100 μatm CO2) on embryonic thermal tolerance in Atlantic cod and Polar cod, and to use those embryonic tolerance windows for projections of spawning habitat suitability under different climate change (emission) scenarios. The meta-analysis (Manuscript IV) encompassing data from several hundred species explicitly tested two hypotheses: (i) Thermal tolerance increases from spawning adults and embryos to larvae and non-reproductive adults; (ii) the temperature dependence of physiological rates (i.e., thermal responsiveness) is higher in organisms with narrow temperature windows (stenothermal species or life stages) than in organisms with wide temperature windows (eurythermal species or life stages). Finally, impact risks associated with different global warming scenarios were assessed based on stage-specific tolerance thresholds of species from various climate zones. Experimental results confirmed that embryonic temperature windows are wider in Atlantic cod than in Polar cod. Embryo mortality increased especially above the species-specific spawning temperature range (Atlantic cod: ≥9°C, Polar cod: ≥3°C), most likely due to constraints on aerobic energy (ATP) by mitochondria. The effects of OA intensified these thermal constraints, resulting in a narrowing of the temperature window for embryonic development and thus reproduction. Detailed experiments with Atlantic cod, including biochemical analyses, revealed that embryo vulnerability to additional effects of OA was highest during gastrulation, which is an early period characterized by high developmental complexity and low homeostatic capacity (i.e., low activity and expression of acid-base relevant ion transporters). Enhanced embryonic tolerance after this critical period was probably associated with a rapid (exponential) increase in capacity for ion transport and ATP production. The potential for acclimatization to warming and OA was evidenced through temperature- and OA-dependent changes in protein expression and enzyme activity, especially in larval stages. However, additional costs and developmental trade-offs associated with capacity adjustments during acclimatization (e.g., increased enzyme activity and ATP synthesis under OA) were reflected by increased embryonic oxygen consumption rates and reduced larval size at hatch in both species. Collectively, four experiments consistently showed that OA negatively affects embryonic thermal tolerance and energy efficiency in Atlantic cod and Polar cod. Spawning habitat projections based on embryonic tolerance windows suggest that under the high emission scenario (Representative Concentration Pathway 8.5), both species could lose many important spawning habitats in the northern Northeast Atlantic due to a decrease in embryo survival probability of more than 50%. Reduced emissions (RCP4.5) may avert dangerous climate impacts on Atlantic cod, but still leave few spawning areas for the more vulnerable Polar cod. However, strong emission cuts (RCP2.6), in line with the 1.5°C target, could minimize the risk of spawning habitat loss for both species. The meta-analysis revealed a globally consistent pattern of stage-specific thermal tolerance, indicating that spawners and embryos are less tolerant than larvae and non-reproductive adults. More specifically, it was shown that the tolerance windows of spawners and embryos are generally more than 10°C narrower than those of larvae and adults, possibly reflecting ontogenetic shifts in aerobic and homeostatic capacity. In addition, thermal responsiveness was found to be higher in stenothermal species and life stages with narrow temperature ranges than in eurythermal ones, indicating a mechanistic link between organismal thermal tolerance and the kinetic coordination of metabolic functions. These results clearly identify the temperature requirements for reproduction (gonad and embryo development) as a critical lifecycle bottleneck with respect to the climate change vulnerability of marine and freshwater fish. The global risk assessment revealed that if warming continues unabated (RCP8.5), approximately 50% of the investigated species (N = 107 out of 211) could be confronted with water temperatures exceeding the current tolerance limit of spawners and/or embryos. This means that many species would have to relocate their spawning activity into cooler seasons or regions, which may be particularly problematic for polar species and those dependent on specific habitats (e.g., reef fishes). A positive perspective is that limiting global warming to 1.5°C could reduce the number of species at risk to less than 10%. The results of this thesis clearly demonstrate the importance of integrating life cycle bottlenecks into physiology-based risk assessments for fish stocks. Habitat models and other modelling approaches thus become more effective tools not only to inform societies and policy makers about potential climate change impacts on fish populations and ecosystems, but also to develop effective mitigation strategies. For example, spawning habitat projections indicating potential refuge areas for Atlantic cod and Polar cod can help to establish marine conservation zones and other proactive measures against additional human perturbations such as overfishing and pollution.