SUMMARYWe investigated the distribution of serotonin (5-HT)-containing neuroepithelial cells (NECs), the putative O 2 sensing cells, in the gills of four species of fish: trout (Oncorhynchus mykiss), goldfish (Carassius auratus), trairão (Hoplias lacerdae) and traira (Hoplias malabaricus) using immunohistochemical markers for 5-HT, synaptic vesicles and neural innervation. We found that all fish had a cluster of innervated, serotonergic NECs at the filament tips, but there were species-specific distributions of serotonincontaining NECs within the primary gill filaments. Trout gill filaments had a greater number of serotonin-containing NECs than both trairão and traira, whereas goldfish primary filaments had none. Serotonin-containing NECs in the secondary lamellae were most numerous in goldfish, present in trairão and traira, but absent in trout. Those found in the primary filament were generally associated with the efferent filamental artery. Innervated, serotonin-containing cells (NECs or Merkel-like cells) were also found in the gill rakers of trout and goldfish although vesicular serotonin was only found in the gill rakers of goldfish. These differences in serotonergic NEC distribution appear to reflect paracrine versus chemoreceptive roles related to hypoxia tolerance in the different fish species.

SUMMARY We investigated the influence of oxygen demand (developmental stage) and supply (hypoxia, water flow rate, the chorion and body movements) on the oxygen concentration within the boundary layer next to the chorion of embryos or skin of larvae of rainbow trout (Oncorhynchus mykiss). Oxygen microelectrodes were used to measure dissolved oxygen (DO) within the boundary layer of trout embryos and larvae. As the embryos and larvae developed, the DO gradient and the thickness of the boundary layer increased. The DO concentration within the boundary layer next to the chorion or skin surface decreased as the DO concentration in the free-stream water decreased. A decrease in water flow rate increased the magnitude of the gradient and thickness of the boundary layer. In normoxia, the DO in the perivitelline fluid inside the chorion was 16±3.0% saturation at 31 days post fertilization, indicating that the chorion was a significant barrier to oxygen diffusion. The number of body movements did not change when embryos were exposed to hypoxia before hatching, but after hatching, hypoxia resulted in a decrease in body movements of the larvae. Taken together, our data indicate that the oxygen boundary layer around trout embryos and larvae depends on both the oxygen demand and supply. The factors that significantly impacted boundary layer oxygen were developmental stage, free-stream oxygen levels, water flow rate, and the presence of the chorion.

Growth was measured in dechorionated and chorionated embryos of rainbow trout Oncorhynchus mykiss exposed to three oxygen treatments: hypoxia, normoxia and hyperoxia (nominally 5, 10 and 15 mg l À1 ). Dechorionated embryos grew faster than chorionated ones in all three oxygen treatments. Similar patterns, however, were found in both dechorionated and chorionated groups; embryos exposed to hypoxia grew less than normoxic and hyperoxic exposed embryos. Moreover, there was no significant interaction between the effect of oxygen levels and presence or absence of the chorion on growth after hatching. These results suggest that the effect of oxygen on growth was not affected by the presence of the chorion.

The acute responses of fish to hypoxia are well established, however the time course for the hypoxic ventilatory response in chronic hypoxia is not well understood. In the present study we assessed the ventilatory and cardiovascular responses of the facultative air breather, Amia calva, to severe hypoxia for 3 days. During prolonged exposure to hypoxia, there was an initial increase in gill breathing frequency, amplitude and total gill ventilation over the first three hours of hypoxic exposure, but this was not sustained. After the initial acute response, there was a decrease in total ventilation that lasted until about 24 hours of hypoxic exposure, when a secondary increase in ventilation started taking place. There was also an immediate bradycardia that was not sustained. These changes were accompanied by changes in the size, and density but not distribution of putative gill neuroepithelial cells (homologous to carotid body glomus cells) after 3 days of hypoxic exposure indicating the importance of these cells in the hypoxic ventilatory response. The time domains seen in gill breathing in Amia are similar to those reported in mammals however the time course of these responses is slower with a ventilatory "roll off" developing in hours not minutes. (funded by NSERC of Canada)