Aquatic vertebrates that emerge onto land to spawn, feed, or evade aquatic predators must return to the water to avoid dehydration or asphyxiation. How do such aquatic organisms determine their location on land? Do particular behaviors facilitate a safe return to the aquatic realm? In this study, we asked: will fully-aquatic mosquitofish (Gambusia affinis) stranded on a slope modulate locomotor behavior according to body position to facilitate movement back into the water? To address this question, mosquitofish (n = 53) were placed in four positions relative to an artificial slope (30° inclination) and their responses to stranding were recorded, categorized, and quantified. We found that mosquitofish may remain immobile for up to three minutes after being stranded and then initiate either a “roll” or a “leap”. During a roll, mass is destabilized to trigger a downslope tumble; during a leap, the fish jumps up, above the substrate. When mosquitofish are oriented with the long axis of the body at 90° to the slope, they almost always (97%) initiate a roll. A roll is an energetically inexpensive way to move back into the water from a cross-slope body orientation because potential energy is converted back into kinetic energy. When placed with their heads toward the apex of the slope, most mosquitofish (>50%) produce a tail-flip jump to leap into ballistic flight. Because a tail-flip generates a caudually-oriented flight trajectory, this locomotor movement will effectively propel a fish downhill when the head is oriented up-slope. However, because the mass of the body is elevated against gravity, leaps require more mechanical work than rolls. We suggest that mosquitofish use the otolith-vestibular system to sense body position and generate a behavior that is “matched” to their orientation on a slope, thereby increasing the probability of a safe return to the water, relative to the energy expended.

Respirators must be properly used to be effective. In an experimental protocol, 145 subjects were trained and then observed donning and doffing respirators. Filtering facepiece and dual cartridge half face mask types were studied. Subjects were then tested for knowledge and for proper performance using video recording analysis. Knowledge tests showed adequate learning, but performance was often poor. Inspection, strap tension (half mask), seal checking, and avoiding mask contact during doffing were particularly problematic. Mask positioning was generally well done. Correlation between knowledge and performance for specific items was generally poor, although there was a weak correlation between overall knowledge and overall performance (rho = 0.32) for the half mask users. Actual unprompted performance as well as knowledge and fit-testing should be assessed for user certification. Respirator design approval should consider users' ability to learn proper technique.

Video training for proper respirator use can be effective when direct training from an occupational health professional is unavailable. These methods are particularly relevant to "rapid rollout" situations, such as natural disasters, epidemics, or bioterrorism concerns.