LAUSR

Temporal environmental variability causes behavioural and physiological responses in organisms that can affect their spatial location in time, and ultimately drive changes in population and community dynamics. Linking ecological changes with underlying environmental drivers is a complex task that can however be facilitated through the integration of physiology. Our overarching aim was to investigate the association between physiological performance and habitat utilisation patterns modulated by short temporal fluctuations in environmental factors. We used in situ monitoring data from a system experiencing extreme environmental fluctuations over a few hours and we selected four fish species with different habitat utilisation patterns across dissolved oxygen (DO) fluctuations: two commonly observed species (Siganus lineatus and Acanthopagrus pacificus), including at low DO (40 and 50% saturation, respectively), and two reef species (Heniochus acuminatus and Chaetodon vagabundus) never recorded below 70% saturation. We hypothesised that these patterns were associated to species' physiological performance in hypoxia. Therefore, we measured different metabolic variables (O2crit, incipient lethal oxygen (ILO), time to ILO, index of cumulative ambient oxygen deficit (O2deficit), maximum oxygen supply capacity (α)) using respirometry. Physiological performance differed among species and was intrinsically associated to habitat use patterns. S. lineatus had a lower O2crit than H. acuminatus, A. pacificus and C. vagabundus (13, 18.7, 20 and 20.2% saturation respectively). Additionally, S. lineatus and A. pacificus displayed better capacity for survival below O2crit than C. vagabundus and H. acuminatus (lower ILO, higher O2deficit and longer time to ILO) and higher α. Field monitoring data revealed that DO temporarily falls below species' O2crit and even ILO on most days, suggesting that short temporal variability in DO likely forces species to temporarily avoid harmful conditions, driving important changes in ecosystem structure over a few hours. Our results support the hypothesis that organismal physiology can provide insights into ecological changes occurring over a few hours as a result of environmental variability. Consequently, integrating physiology with ecological data at relevant temporal scales may help predict temporal shifts in ecosystems structure and functions to account for ecological patterns often overlooked and difficult to identify.