SummaryPrevious morphometric methods for estimation of the volume of components, surface area and thickness of the diffusion barrier in fish gills have taken advantage of the highly ordered structure of these organs for sampling and surface area estimations, whereas the thickness of the diffusion barrier has been measured orthogonally on perpendicularly sectioned material at subjectively selected sites. Although intuitively logical, these procedures do not have a demonstrated mathematical basis, do not involve random sampling and measurement techniques, and are not applicable to the gills of all fish. The present stereological methods apply the principles of surface area estimation in vertical uniform random sections to the gills of the Brazilian teleost Arapaima gigas. The tissue was taken from the entire gill apparatus of the right-hand or left-hand side (selected at random) of the fish by systematic random sampling and embedded in glycol methacrylate for light microscopy. Arches from the other side were embedded in Epoxy resin. Reference volume was estimated by the Cavalieri method in the same vertical sections that were used for surface density and volume density measurements. The harmonic mean barrier thickness of the water-blood diffusion barrier was calculated from measurements taken along randomly selected orientation lines that were sine-weighted relative to the vertical axis. The values thus obtained for the anatomical diffusion factor (surface area divided by barrier thickness) compare favourably with those obtained for other sluggish fish using existing methods.

The stomach of Pterygoplichthys anisitsi has a thin, translucent wall and a simple squamous epithelium with an underlying dense capillary network. In the cardiac and pyloric regions, most cells have short microvilli distributed throughout the cell surface and their edges are characterized by short, densely packed microvilli. The mucosal layer of the stomach has two types of pavement epithelial cells that are similar to those in the aerial respiratory organs. Type 1 pavement epithelial cells, resembling the Type I pneumocyte in mammal lungs, are flat, with a large nucleus, and extend a thin sheet of cytoplasm on the underlying capillary. Type 2 cells, resembling the Type II pneumocyte, possess numerous mitochondria, a well-developed Golgi complex, rough endoplasmic reticulum, and numerous lamellar bodies in different stages of maturation. The gastric glands, distributed throughout the mucosal layer, also have several cells with many lamellar bodies. The total volume (air + tissue), tissue, and air capacity of the stomach when inflated, increase along with body mass. The surface-to-tissue-volume ratio of stomach varies from 108 cm(-1) in the smallest fish (0.084 kg) to 59 cm(-1) in the largest fish (0.60 kg). The total stomach surface area shows a low correlation to body mass. Nevertheless, the body-mass-specific surface area varied from 281.40 cm(2) kg(-1) in the smallest fish to 68.08 cm(2) kg(-1) in the largest fish, indicating a negative correlation to body mass (b = -0.76). The arithmetic mean barrier thickness between air and blood was 1.52 +/- 0.07 microm, whereas the harmonic mean thickness (tau(h)) of the diffusion barrier ranged from 0.40 to 0.74 microm. The anatomical diffusion factor (ADF = cm(2) microm(-1) kg(-1)) and the morphological O(2) diffusion capacity (D(morphol)O(2) = cm(3) min(-1) mmHg(-1) kg(-1)) are higher in the smallest specimen and lower in the largest one. In conclusion, the structure and morphometric data of P. anisitsi stomach indicate that this organ is adapted for oxygen uptake from air.