
ABSTRACT One of the most pressing challenges confronting the scientific community today is understanding the consequences of global change on ecological communities and mitigating their impacts. Cleaning mutualisms are essential to the health and stability of coral reef ecosystems, influencing individual fitness, parasite dynamics, and community composition. However, these interactions are increasingly threatened by anthropogenic pressures. This review describes the behavioural diversity and ecology of dedicated cleaners, ranging from cognitively sophisticated wrasses and efficient, low‐conflict gobies to visually signalling shrimps. We then review how the intricate interactions between cleaners and clients respond to environmental change, including chronic stressors (e.g., acidification), extreme events (e.g., marine heatwaves), and direct anthropogenic disturbances (e.g., underwater noise). Mounting evidence demonstrates that such stressors can disrupt cleaner and client dynamics, diminishing interaction motivation and quality, and altering underlying neurobiological mechanisms. Despite recent advances, significant gaps remain in our understanding of the long‐term resilience and recovery of cleaning mutualisms across ecological, behavioural, cognitive and physiological pathways. As ocean climate‐related drivers of change intensify and increasingly co‐occur, it becomes urgent to investigate whether cleaners can continue to fulfil their ecological roles, or if reefs will face a future with compromised mutualistic services.
ABSTRACT Monitoring diverse, dispersed small‐scale invertebrate fisheries remains challenging because taxonomic diversity, informal trade and dispersed harvesting limit sustained programs that can capture standardised data used for catch identification and measurement. Computer vision is increasingly being used to overcome these challenges, but most affordable and accessible single‐camera systems face a trade‐off: close‐range imagery required for high fidelity species identification while increasing parallax error for three‐dimensional specimens. We tested a hardware‐agnostic calibration method to improve an automated fisheries‐monitoring platform. Nationally standardised coins placed on specimens provided a local, distance‐independent scale reference. Manual, coin‐corrected and uncorrected computer‐vision measurements of carapace length in spiny lobsters ( Panulirus penicillatus ) were compared using linear mixed‐effects models. Coin correction removed systematic bias, reducing mean percentage error from 10.77% to 1.20% and mean absolute error from 8.6 to 2.3 mm. By contrast, linear‐regression correction for Panulirus penicillatus under‐corrected small lobsters and over‐corrected large individuals, biasing Length‐Based Spawning Potential Ratio estimates. Where remote observers could verify species‐level labels, computer‐vision classification exceeded 90% accuracy, with other performance metrics scoring above 80%. Performance on unfiltered operational data depended on image acquisition: close‐range board imagery substantially outperformed greater‐distance mat imagery for taxa requiring fine diagnostic detail, particularly bivalves and gastropods. Coplanar calibration is therefore necessary for 2D computer‐vision systems to provide reliable species identification and morphometric data simultaneously, supporting scalable, field‐appropriate monitoring of multi‐species invertebrate fisheries in data‐limited small‐island developing states.
ABSTRACT Critical swimming speed (Ucrit) is widely used to quantify prolonged swimming performance in fishes, yet available measurements remain concentrated in particular species, life history stages, morphotypes and experimental contexts, leaving broad‐scale patterns of Ucrit variation poorly synthesized. Here, we compiled a global dataset of experimentally measured absolute and relative Ucrit (absolute Ucrit standardized by body length) from 1345 experimental cases across 361 species and 387 publications. Body length was strongly associated with Ucrit, but this association varied between absolute and relative Ucrit and among life history stages. Absolute Ucrit increased with body length across all stages, more steeply in larvae than in adults and juveniles, whereas relative Ucrit declined with body length in adults and juveniles but increased weakly in larvae. Morphology provided further explanatory insight in adults and juveniles: elongated body forms and forked or lunate caudal fins were generally associated with higher Ucrit, while caudal peduncle throttling (CPt) was negatively associated with both absolute and relative Ucrit, and body lateral shape (BLs) was negatively associated mainly with relative Ucrit. Water temperature was positively associated with both Ucrit metrics after accounting for body length, life history stage and ecosystem type, although temperature slopes varied among ecosystem types (freshwater, marine, migratory). Overall, these results show that body‐size scaling differs between absolute and relative Ucrit and that Ucrit variation is jointly associated with life history stage, morphology and temperature, providing a clearer basis for comparing swimming performance across fishes and interpreting Ucrit reference values in applied ecological and management contexts.
ABSTRACT Fishing‐induced evolution (FIE) has emerged as a key concern in fisheries science, yet it remains unclear how different fishing sectors contribute to evolutionary change. Here we examine how large‐scale fisheries (LSF) and small‐scale fisheries (SSF) may differ in the pathways through which fishing mortality becomes trait‐dependent in fish populations. We synthesise how fishery characteristics—including fishing intensity, spatial and temporal targeting, gear diversity, technological capacity, target breadth and governance structures—may influence the timing, consistency and targets of selection, thereby shaping evolutionary outcomes. We suggest that SSF may generate more heterogeneous selection regimes, acting on individual variation in behavioural, ecological and vulnerability‐related traits through partial harvesting, diverse gear use and localised fishing practices. In contrast, LSF may generate consistent patterns of trait‐dependent mortality because of operational scale, greater catch capacity, advanced technologies and standardised management frameworks. These differences do not necessarily imply stronger or weaker overall selection, but suggest that SSF and LSF may differ in the mechanisms through which phenotypes become exposed to fishing mortality and the traits most likely to be affected. Differences in target specificity, bycatch and discard practices may also contribute to divergent evolutionary pathways between sectors. While high mortality remains an important driver of life‐history evolution, sector‐specific mechanisms may shape trait trajectories, including behaviour, growth, habitat use and migration. We highlight key knowledge gaps and emphasise the need to integrate evolutionary perspectives into both LSF and SSF management, especially given technological advancement and potential future regulation and standardisation of SSF practices.
Biomass-based reference points are used to define stock status and set sustainable exploitation levels. Risks should be equivalent across stocks, regardless of the data and knowledge available to assess them by maintaining comparable probabilities of achieving targets and avoiding limits across agreed time scales. This study introduces Brebuild$$ {B}_{\mathrm{rebuild}} $$ , the lowest biomass from which a stock can rebuild to a target within an agreed time-frame under a specified fishing scenario, and Trebuild$$ {T}_{\mathrm{rebuild}} $$ , the time required to rebuild from a given state. Once Brebuild$$ {B}_{\mathrm{rebuild}} $$ (or an equivalent biomass threshold), Btarget$$ {B}_{\mathrm{target}} $$ , and a harvest control rule defining Frebuild$$ {F}_{\mathrm{rebuild}} $$ are agreed, Trebuild$$ {T}_{\mathrm{rebuild}} $$ is a random variable with a probability distribution. Brebuild$$ {B}_{\mathrm{rebuild}} $$ can serve as a limit or a trigger in harvest control rules. Using age-structured projections for 82 Northeast Atlantic stocks, we summarise the relationship between Trebuild$$ {T}_{\mathrm{rebuild}} $$ and life-history covariates ( SPR0$$ {\mathrm{SPR}}_0 $$ , steepness) across stock-recruitment scenarios. Recovery is faster for stocks with higher per capita productivity (i.e., SPR0$$ {\mathrm{SPR}}_0 $$ ) and steepness, with little difference between Beverton-Holt and Ricker forms once these characteristics are accounted for. The approach shifts emphasis from recruitment impairment to rebuilding capacity, makes the trade-off with yield during rebuilding explicit, and provides a common currency for risk-equivalent advice across assessment models and data tiers.