Figure 1.

Number of publications (total = 207) by decade from 1940 to 2019 relevant to the thermal physiology of marine mammals. Refer to the supplementary file for the studies included. Studies are classified by setting (i.e. where it was performed) (left bar) and by the state or condition of the animal (right bar; modeling studies that did not make measurements on an animal were not included). Numbers inside the field bar (light gray) indicate how many of those field studies (if ≥1) involved animals in water (vs. land). Venn-diagrams on top of the bars indicate how many studies during that decade used biotelemetry for remote measurements of physiological variables (e.g. body temperature; top), or behavioral variables (e.g. diving behavior; bottom), or both. Biotelemetry is defined broadly here and includes remote sensing of physiologically relevant data (e.g. infrared thermography) and the use of data loggers (i.e. biologgers) to record continuous measurements on captive animals. Categories for study settings were defined as: lab (i.e. using an experimental approach in a captive and controlled setting), field (i.e. using wild animals in their natural setting), biophysical modeling (e.g. heat transfer models). Categories for animal state were defined as: ex vivo (i.e. excised tissue measurements), carcass (i.e. in situ measurements on a dead animal), peri-mortem (i.e. measurements taken at or near death), restrained (including sedated), captive (and unrestrained or freely-behaving given surrounding constraints), trained (i.e. accustomed to experimental protocol or performing a task on command), free-ranging (i.e. freely-behaving wild animals). Classification into these categories was inclusive (i.e. one study could be classified into multiple categories) and a study’s contribution to multiple categories was weighted evenly. Only research articles were included; book and encyclopedia chapters, theses, reviews, or comments were excluded. Papers that did not include any discussion of thermoregulation even if relevant parameters were measured (e.g. blubber lipid content, body composition, movement patterns in relation to sea surface temperature) were excluded.