Many species from several different families of fishes perform mouthbrooding, where one of the sexes protects and ventilates the eggs inside the mouth cavity. This ventilation behaviour differs from gill ventilation outside the brooding period, as the normal, smallamplitude suction-pump respiration cycles are alternated with actions including near-simultaneous closed-mouth protrusions and highamplitude depressions of the hyoid. The latter is called churning, referring to its hypothetical function in moving around and repositioning the eggs by a presumed hydrodynamic effect of the marked shifts in volume along the mouth cavity. We tested the hypothesis that churning causes the eggs located posteriorly in the mouth cavity to move anteriorly away from the gill entrance. This would prevent or clear accumulations of brood at the branchial basket, which would otherwise hinder breathing by the parent. Dual-view videos of female Nile tilapias (Oreochromis niloticus) during mouthbrooding showed that churning involves a posterior-toanterior wave of expansion and compression of the head volume. Flow visualisation with polyethylene microspheres revealed a significant inflow of water entering the gill slits at the zone above the pectoral fin base, followed by a predominantly ventral outflow passing the ventrolaterally flapping branchiostegal membranes. X-ray videos indicated that particularly the brood located close to the gills is moved anteriorly during churning. These data suggest that, in addition to mixing of the brood to aid its oxygenation, an important function of the anterior flow through the gills and buccal cavity during churning is to prevent clogging of the eggs near the gills.

149 first-year university students. Our results revealed that teaching of animal behaviour mainly emphasizes ontogeny and, to a lesser extent, causation, and that Tinbergen's framework is not explicitly referred to. Students, in both groups, typically addressed only one or two questions, with the majority addressing causation or both causation and ontogeny when explaining behaviour in general, but function or causation and function when explaining specific animal behaviours. The latter high prevalence of function may be largely due to intuitive teleological thinking. Evolution was generally completely neglected, even in university students who had recently completed an evolution course. Our results revealed that transfer of the concepts of ontogeny and evolution was (almost) absent, although the skill of knowledge transfer is an important goal in science education.We argue why Tinbergen's framework should best be an integral part of any biology curriculum.