Search

Stereo-video scuba transects were conducted during daylight hours from June to September 2013 within a proposed marine protected area (MPA) in the Firth of Clyde, west of Scotland. More juvenile Atlantic cod Gadus morhua of fork length (LF) range 6–11 cm were observed in substrata containing mixed gravel, including maerl, than in boulder-cobble substrata with high algal cover, or sand with low density seagrass. Community composition was significantly different between substratum types. A decrease in G. morhua abundance was observed over the period of data collection. Over time, mean and variance in G. morhua LF increased, indicating multiple recruitment events. Protecting mixed gravel substrata could be a beneficial management measure to support the survival and recruitment of juvenile G. morhua; other substrata might be important at night given their diel migratory behaviour. Stereo-video cameras provide a useful non-destructive fisheries-independent method to monitor species abundance and length measurements.

Variation in the survival of herring between the egg stage and the age of recruitment to the spawning population has been examined for seven Atlantic herring populations in northern European waters. The sources of the variation have been partitioned between density-dependent and density-independent factors. The magnitude of the density-independent component was found to be related to the scale of the processes controlling the dispersion and distribution of larvae from the various populations. We conclude that the spawning strategies of some populations are adapted to oceanographic conditions to maximize the probability of a particular larval transport pattern, although no single model could account for all spawning strategies.

The mean properties of larval fish populations do not necessarily reflect the properties of the mean individual. For example, the change in mean length in a population with time may not reflect the average individual growth rate, since individual growth rates and survival probability are linked so that slow growing individuals suffer higher mortality. Hence, mean growth rate indicated from population data could be biased upwards. Factors which influence the magnitude and variability of individual growth rates can exert nonlinear effects on population survival. Two categories of process must be considered: first, the variability in exposure of the average individual as a consequence of individual variability in dispersal through a patchy environment; and second, the intrinsic variability between individuals expressed even under equal exposure conditions. These two aspects have been addressed independently, the first by lagrangian modelling of individual fish larvae linked to spatially resolved hydrodynamic models, and the second by strategic biological modelling. In this paper, progress towards the goal of individually based larval fish ecosystem models is reviewed, highlighting the space and time scales which may be important in such systems, and identifying the gaps in current knowledge of larval biology.

The daily otolith increment growth of individuals in a cohort of fish larvae was simulated by a simple individual-based model over 30 days. The daily otolith growth of an individual larva was dependent on past growth, within fixed limits common to all larvae. The survival of a larva at the end of each day was either a linear function of larval growth or a random outcome, simulating growth-dependent and growth-independent mortality, respectively, The combined effect of the external environment on growth was also studied. Eleven environmental scenarios, favouring or hindering growth at different stages, were tested and compared to runs with no environmental effect on growth. Growth-dependent mortality induced an increase in the average otolith daily increment width amongst surviving larvae. Such an outcome, however, could be negated by an unfavourable environment. The increase in mean growth rate of the population generated by growth-dependent mortality was directly related to the inherent variability in daily otolith growth. With increased variability, the influence of the environment became relatively less important. The effect of the environment on growth was more critical during the early stages of development. A comparison of results generated by the model with patterns observed in data from a field survey of larval herring was consistent with the occurrence of growth-dependent mortality in the sea. The simulation model provided a useful insight into the way in which various processes controlling larval growth interact.

The timing of the development of schooling behaviour in Atlantic herring Clupea harengus and the quantitative and qualitative evolution of such behaviour in the larval and post-metamorphic stages were investigated with a simulation model and experimental observations of laboratory-reared fish. Herring larvae started to develop schooling behaviour between the lengths of 35 and 40 mm. This coincided with the beginning of metamorphosis, characterized by ontogenetic changes which require a switch to a different antipredator strategy from that employed by early larvae, while providing the sensory and anatomical mechanisms necessary for the formation of schools. Schooling behaviour was established at the end of metamorphosis (50–55 mm) and its characteristics continued to evolve throughout the early juvenile stage. A critical period in terms of vulnerability to predation is expected between the end of the larval and beginning of the juvenile stages, before schooling behaviour is fully developed. The implications of some of the results for the methodology of future studies are also discussed.