Tintinnoinea
Methods

Tintinnids have traditionally been analyzed on the basis of water-column samples, the only exceptions being a few studies on cysts found in Recent intertidal and offshore marine sediments (Reid and John, 1981), and some investigations on calpionellids, extinct late Jurassic to early Cretaceous forms (Colom, 1948; Tappan and Loeblich Jr., 1968; Remane, 1969; 1971; Echols and Fowler, 1973; Corliss, 1979). The most common sampling techniques for tintinnids are indicated below with brief comments on their performance (for further information on each type of gear, see Boltovskoy, ed., 1981, and Paranjape, 1991).

Plankton nets have long been the most widely used gear for qualitative studies. Improvements in their design, and the introduction of new ancillary devices (opening-closing mechanisms, flowmeters, etc.) have allowed their use in ecological studies requiring assessments of abundance and horizontal and vertical distribution patterns. However, as with several other microzooplanktonic organisms, standard plankton nets often fail to yield adequate results because tintinnids are small in size, but unlike many comparably-sized phytoplankters, are seldom abundant. Indeed, their small size requires the use of phytoplankton-sized meshes (around 20-30 µm). However, these clog fast, not only complicating the task of filtering enough water, but also making flowmeter measurements imprecise.

The use of suction pumps with the outlet driven through a net or another sieving device allows direct control over clogging even with very small-sized meshes, very precise estimates of the volumes of water filtered, and under static conditions, very high spatial resolution (since no towing is involved). Furthermore, damage due to turbulence and pressure can also be controlled to a certain extent (although detachment of the cell from the lorica due to mechanical disturbance is common). Because tintinnids are usually concentrated in the upper water layers the use of pumps does not pose major operational problems. However, vertical profiles to depths in excess of 100-200 m are increasingly difficult (see Beers, 1981).

Bottle sampling provides another very useful tool for tintinnid studies, especially when vertical profiles of overall abundance and species-specific distributions are sought. The volume yielded by standard (Niskin, Go-Flo) bottles (ca. 5-10 l) is usually too low for adequate sample-sizes, for which reason 6-3 bottles from each level are required. Replicate samples are pooled and gently filtered through a 20 µm mesh. Disturbances are minimal with these method, which ensures the suitability of the specimens for detailed morphological and cytological studies.

Sediment trap techniques have undergone major improvements in the last years, thus providing a useful tool for the collection of various microplanktonic organisms (US GOFS, 1989; Lange and Boltovskoy, 1995; see chapter "Radiolaria Polycystina" in this volume). Yearly abundance and specific composition cycles are also best assessed by means of sediments trap samples, which is of particular importance when flux rates are compared with primary production fluctuations in the upper mixed layer (Honjo et al., 1982, 1988; Deuser et al., 1983, 1990; Wefer, 1989). However, sediment traps also have important shortcomings for tintinnids: efficient sediment trap sampling is usually restricted to depths in excess of 500-700 m (US GOFS, 1989; Lange and Boltovskoy, 1995), and traps integrate the flux from several biologically dissimilar layers. Furthermore, although specific relative compositions may not change significantly due to destruction of the loricae in the process of sinking (Ling, 1992), overall destruction is very high: only ca. 1% of the loricae produced in the upper layers reach 800 m in a recognizable fashion (Boltovskoy et al., 1993a). Furthermore, the very significant qualitative dissimilarities between microzooplanktonic upper-layer assemblages and those retrieved in 800-2000 m sediment trap samples, presumably due to intermediate- and deep-water advection of cells produced elsewhere (Boltovskoy et al., 1996), also hinders the use of these materials for ecologic and biogeographic surveys.

Once concentrated, samples can be preserved with 2% buffered formaldehyde, acid Lugol's iodine solution, mercuric chloride with bromophenol blue (for fixation and staining), or a solution of 3% glutaraldehyde and 1% osmium tetroxide. Laval (1971, 1972), Laval-Peuto (1975), Hedin (1975), Gold and Morales (1976), Laval-Peuto et al. (1979), Gold (1979), Capriulo et al. (1986) and Wasik and Mikolajczyk (1991, 1992) provide useful information on techniques for SEM and TEM preparations aimed at the analysis of the ultrastructure of tintinnid cells and loricae.

Usually the thin cytoplasmic stalk (peduncle) attaching the cell to its lorica (Cymatocylis convallaria intro) can be easily broken either by manipulation during sampling (when using nets and pumps), or by the effects of the fixative. Moreover, the fixative can also cause subsequent distortion (shrinkage or swelling) of the cell (Gilron and Lynn, 1989). As a consequence, most of the loricae are found empty in fixed samples, only very few retaining the (distorted) cell inside. This may be a serious problem not only for taxonomic studies, but also for ecologic-biogeographic ones, since one cannot ensure whether empty loricae represent individuals still alive at the time of sampling, or were transported empty from elsewhere by currents. Also, empty loricae prevent direct estimation of the cell's contribution to total tintinnid biomass (Alder, 1990; see Verity and Langdon, 1984, for alternative methods). Thus, whenever possible potential biases resulting form the above-described artifacts should be assessed on the basis of fresh, unpreserved samples (including observations, counts and measurements in vivo).

Regardless of the collecting method, samples are concentrated before storage. Thus, when estimates of absolute abundance are sought, samples often need to be diluted again prior to counting under the inverted microscope. It is strongly recommended that absolute abundance assessments be based on counts (Utermöhl, 1858, technique) of several dilute counting chambers rather than a single one with more concentrated materials, where other particles, especially phytoplanktonic cells, may strongly interfere. Furthermore, counts of several subsamples yield more precise results (Frontier, 1981), and also permit estimation of counting errors.