Foraminifera
Methods

Most surveys on the distributional patterns of planktic Foraminifera were based on sediment samples, rather than on plankton tows. Both methods are useful for the study of distributional patterns (see chapter on "Radiolaria Polycystina"). However, comprehensive biological studies must be based on plankton tows and/or hand sampling by SCUBA methods are required.
(Changes in abundances), latitudinal changes in the relative abundances of planktic Foraminifera.
(Globigerinoides ruber 3), abundance of this species in the eastern South Atlantic.

A third way of obtaining information on distributional patterns is by means of sediment traps. Time-series sediment trap materials have several advantages over planktonic and/or sedimentary samples, e.g., larger sample-size, integration over preselected depth and time ranges.

None of these 3 approaches alone suffices to yield an adequate picture of the world-wide distribution patterns of planktic Foraminifera, as all three have advantages as well as drawbacks (see "Radiolaria Polycystina" in this volume). Information combined from various sources should therefore be used in order to maximize the potential of these organisms for oceanographic and paleoceanographic surveys.

—SCUBA sampling
Sampling planktic Foraminifera by SCUBA diving, in which specimens are captured by hand in solid containers (Hemleben et al., 1989), is very time consuming, but it is the best method for obtaining healthy specimens for culture experiments and cytological and ultrastructural studies in the transmission electronic microscope (TEM). However, SCUBA methods require trained divers and well-equipped facilities to maintain the specimens in laboratory cultures (Hemleben et al., 1989), or to prepare specimens for TEM research.

—Plankton tows
Planktic Foraminifera have been collected in the past with a wide variety of nets made with mesh of various sized openings, towed through the water vertically, horizontally, or obliquely. Some have been equipped with opening and closing devices to collect material from descrete depths, sometimes fitted with flowmeters to measure the amount of water represented by the samples. Mesh sizes have varied between 10 and 300 µm in the last several decades. Hence, many data sets are difficult to compare with one another, especially samples collected in the 60's and 70's compared to more recent decades. Today, most investigators use special opening and closing nets, such as the Multinet (0.25 mö2 opening) or the MOCNESS system (Wiebe et al., 1976), and usually with mesh openings of either 64 or 100 µm. The Multinet is usually towed at a speed of approximately 0.3-0.5 m sö-1. Samples are preserved with HgCl2, formaldehyde, or methyl alcohol, and stored at 4°C on board. In the laboratory samples are split with a rotary liquid splitter. Wet subsamples may be used without further processing, or they may be combusted (<150°C) to remove the cytoplasm and other organic components. If living-to-dead ratios in the water-column assemblage are sought, foraminiferal cytoplasm can be stained using rose Bengal, Sudan black B, or eosin (Boltovskoy and Wright, 1976), in which case ashing is not possible. However, after some practice "living" shells can be identified without staining because specimens containing cytoplasm are easy to distinguish from empty ones by their color. Nevertheless, cytoplasmic remnants may persist in dead specimens for weeks and months (Boltovskoy and Lena, 1970; Bernhard, 1988), especially if they sink without contact with larger phytoplankton aggregates with dense bacterial populations.

Samples are thoroughly wet-sieved through a 30 µm mesh sieve, oven-dried, and sieved again into suitable grain size fractions (e.g., >250 µm, 150-250 µm and <150 µm). Counting is usually performed on splits containing approximately 300 specimens. A random check should be carried out in order to assure correct aliquot size.

One has to be aware that plankton hauls are only spot checks in space and time, not accounting for the complete dynamics of species distributions. Thus, daily to weekly surface and subsurface plankton tows over extended time periods are needed to achieve a reasonable picture of distribution patterns in the productive zone. Deeper (below ca. 80-100 m) plankton tows can help evaluate processes in the flux zone where remineralization of most juvenile specimens occurs. On the other hand, comparison of plankton tows and sediment samples can aid in revealing the effects of currents and environmental conditions in the region.

—Sediment traps
Sediment traps are specially designed containers that are deployed at various depths in the oceans for periods usually of several months to one or more years to accumulate material settling from the water column above (Honjo et al., 1980; US GOFS, 1989). These samples describe quantitatively the seasonal or subseasonal fallout usually below the productive layer plus organisms advected laterally from elsewhere (Siegel and Deuser, 1997), both modified by processes that affect settlement and preservation (Boltovskoy, 1994). Their shortcomings include restricted effectiveness when deployed at depths less than 500-700 m, the integration of the flux from several biologically dissimilar layers, distortion of the near-surface patterns due to selective and non-selective destruction (dissolution, grazing), and lateral advection of particles (see chapter "Radiolaria Polycystina").

—Sediment coring
For paleoceanographic studies, the Multicorer is the best gear for obtaining undisturbed sediment samples from the ocean floor. Usually, a Multicorer core covers the last 3 to 4 kyr. If longer cores are desired, a piston core or other coring devices can be used (see chapter "Radiolaria Polycystina" in this volume).