Three types of bottom samplers were used in this study: slurp gun, Shipek and IKU grab sampler. The slurp gun is a vacuum cleaner-like device that sucks in the uppermost layer (ca. 1 mm) of the seafloor and thus samples only what is present at the surface or very close to it. The Shipek grab sampler, a spring loaded "clam shell" type sampler, obtains samples of surficial seafloor sediments. The samples collected cover an area 20 cm x 20 cm to a maximum depth of about 10 cm. The IKU grab sampler is a large volume (0.5 cubic metre) grab sampler developed by IKU (Institutt for Kontinentalsokkelundersøkelse - Norway) specifically for sampling continental shelf seabed sediments. The sampler penetrates to a depth of 50 cm into the substrate as the sample is obtained and retains the stratigraphic relationships of the surficial materials sampled. The grab sampler operates much like a construction excavator, employing large and widespread spring-loaded jaws that close as the sampler is retrieved from seafloor. The closing force of the jaws is generated through a system of pulleys attached to the retrieval cable. See the Sponge Project website for pictures of the Shipek grab sampler and the IKU grab sampler.

All samples other than piston core samples were stained in Rose Bengal and preserved in a mixture of water and methanol. All samples were sieved with a 63 µm sieve, and in most cases an additional 1 mm sieve was used to retain coarser material. A separate count was made of both residue sizes after residues had been split into countable aliquots. Wet samples were split with a wet sample splitter (Scott and Hermelin 1993), and dry samples were split with an ordinary desk sample splitter. Larger sponge fragments found in the 1 mm sieve were set aside, and a separate count was made of the specimens attached or clinging to them. This constitutes the "sponge fraction" referred to herein. Smaller fragments were left in the >1 mm fraction and except for those foraminifera conveniently positioned at the outer surface of sponge fragments, most specimens had to be extracted manually. This procedure was achieved by holding a fragment with the fingers (when large enough) or a needle and by breaking away the individual sponge lattice silica rods surrounding a foraminiferal specimen until it could be removed with a wet brush. Because this process was time consuming, the exploration of the sponge fragments was not comprehensive. The examination of the surface of large fragments was often more complete than that of the deep interior, and the counts are therefore not perfectly representative. As this paper is primarily dedicated to the non-quantitative aspects of the foraminifera found in the sponge ecosystems, an additional qualitative examination of the >1 mm fraction as well as of the sponge fractions, was carried out to find specimens attached to sponge fragments that would provide valuable information on sponge-foraminifer relationships.