In order to understand the effect of global change on marine fishes, it is imperative to quantify the effects on fundamental parameters such as survival and growth. Larval survival and recruitment of the Atlantic cod (Gadus morhua) were found to be heavily impaired by end‐of‐century levels of ocean acidification. Here, we analysed larval growth among 35–36 days old surviving larvae, along with organ development and ossification of the skeleton. We combined CO2 treatments (ambient: 503 µatm, elevated: 1,179 µatm) with food availability in order to evaluate the effect of energy limitation in addition to the ocean acidification stressor. As expected, larval size (as a proxy for growth) and skeletogenesis were positively affected by high food availability. We found significant interactions between acidification and food availability. Larvae fed ad libitum showed little difference in growth and skeletogenesis due to the CO2 treatment. Larvae under energy limitation were significantly larger and had further developed skeletal structures in the elevated CO2 treatment compared to the ambient CO2 treatment. However, the elevated CO2 group revealed impairments in critically important organs, such as the liver, and had comparatively smaller functional gills indicating a mismatch between size and function. It is therefore likely that individual larvae that had survived acidification treatments will suffer from impairments later during ontogeny. Our study highlights important allocation trade‐off between growth and organ development, which is critically important to interpret acidification effects on early life stages of fish.

In all stomiid genera there is an occipitovertebral gap between the skull and the first vertebra bridged only by the flexible notochord. Morphological studies from the early 20th century suggested that some stomiid genera have 1-10 of the anteriormost centra reduced or entire vertebrae missing in this region. Our study reviews this previous hypothesis. Using a new approach, we show that only in Chauliodus, Eustomias and Leptostomias gladiator vertebral centra are actually lost, with their respective neural arches and parapophyses persisting. We present results from a comparative analysis of the number and insertion sites of the anteriormost myosepta in 26 of the 28 stomiid genera. Generally in teleosts the first three myosepta are associated with the occiput, and the fourth is the first vertebral myoseptum. The insertion site of the fourth myoseptum plays an important role in this analysis, because it provides a landmark for the first vertebra. Lack of association of the fourth myoseptum with a vertebra is thus evidence that the first vertebra is reduced or absent. By counting the occipital and vertebral myosepta the number of reduced vertebrae in Chauliodus, Eustomias and Leptostomias gladiator can be inferred. Proper identification of the spino-occipital nerves provides an additional source of information about vertebral reduction. In all other stomiid genera the extensive occipitovertebral gap is not a consequence of the reduction of vertebrae, but of an elongation of the notochord. The complex structure and ontogeny of the anterior part of the vertebral column of stomiids are discussed comparatively.