Coccolithophores in Polar Waters: HOL Revisited Papposphaera sagittifera HET and Author Thomsen, Helge A. Author Østergaard, Jette B. Author Heldal, Mikal text Acta Protozoologica 2016 55 1 33 50 https://www.mendeley.com/catalogue/0ed4da50-fdd6-3f27-a928-a1e7a637b9c4/ journal article 10.4467/16890027AP.16.005.4046 1689-0027 10994318 P. sagittifera HET ( Figs 4–16 ) The Arctic material from West Greenland ( Fig. 7 ), North East Water Polynya ( Figs 4–5 , 13 ) and Svalbard ( Figs 6, 8–12, 14–16 ) basically confirms what is al- ready known about P. sagittifera HET from this region ( Thomsen 1981 ). The rationale behind illustrating three complete cells from TEM is to document the consistent presence of flagellation. The flagella and the haptonema typically curl up densely which renders measurements of these structures impossible. A haptonema is visible in e.g. Fig. 4 (arrow) where it is differentiated from the flagella by being significantly thinner. The apparent ab- sence of flagellation in the numerous SEM illustrations of complete cells ( Figs 10–12, 14 ) is bound to be an ar- tefact somehow caused by the manipulation of the cells during filtration and subsequent drying. All complete cells illustrated from SEM clearly display organisms that have the typical cell polarity to be expected from a flagellated individual. Dimensional issues across the different regions are accounted for in Table 1 . Figs 4–9. Papposphaera sagittifera HET (TEM / Figs 4 – 5, 7; SEM / Figs 6, 8 – 9) from the Arctic (West Greenland / Fig. 7; NEW / Figs 4 – 5; Svalbard / Figs 6, 8, 9). 4 – whole cell with curled up flagella and haptonema (arrow); 5 – detail of calyx; 6 – lateral view of coccolith displaying the two cycles of elements in the rim and details of the stem and the calyx; 7 – complete cell; notice the difference in length of coccolith processes from one end of the cell to the other; 8 – details of central area calcification; notice the short stub-like central proboscis and also possibly tilted pentagonal elements from the distal rim circle (arrows); 9 – detail of two coccoliths one with a short central process and the other with just a central upheaval of the arms of the crossbar. Table 1. Numerical details of P. sagittifera HET from Arctic sites.

Coccolith length		Coccolith width
Collection site	Reference	Cell dimensions (L × W)	Mean ± sd	min/max	n	Mean ±sd	min/max	n	Ant. Spine	Post. Spine
Arctic (NEW)	Fig. 4	6.1 × 4.2	1.4–1.5	0.75–0.8	2.7–3.3	2.3–2.4
Fig. 13	6.2 × 4.8	3.0–3.7	2.0–2.5
Arctic (Disko)	Fig. 7	8.4 × 6.1	1.3 ± 0.054	1.20–1.39	14	0.69 ± 0.036	0.60–0.73	14	2.5–2.9	1.5–2.1
Arctic (Svalbard)	Fig. 10	6.8 × 6.7	1.12 ± 0.082	0.95–1.31	31	0.65 ± 0.056	0.56–0.81	30	2.5–2.7	0.7–1.2
Fig. 11	6 × 5.7	1.10 ± 0.085	0.98–1.27	19	0.67 ± 0.072	0.56–0.85	19	2.6–2.8	1.1–1.5
Fig. 12	6.2 × 6.9	1.13 ± 0.076	0.10–1.35	30	0.68 ± 0.037	0.62–0.75	27	2.2–2.4	1.0
Fig. 14	7.7 × 6.5	1.11 ± 0.071	0.89–1.27	41	0.63 ± 0.046	0.54–0.71	38

Central area calcification is invariable an axial cross in combination with a number of longitudinal bars even- ly spread across the coccolith. The highest number of such bars observed is 7 ( Fig. 8 ) which creates a densely packed central area. Most coccoliths tend to have 3–5 of these longitudinal bars (see e.g. Fig. 14 ). The calyx takes the shape of a four-winged rosette with the four wings positioned at right angles to each other ( Figs 5–6, 9 ). Each wing has parallel sides. The exterior edge is straight, whereas the interior edge has a variable number of steps that gradually decrease the width of the wing towards the distal end ( Fig. 3A ). There seems to be a significant variability with respect to both the width of the wing and the number of steps on the interior edge. The degree to which the four-winged rosette diverges from the stem is dependent on the an- gle between the exterior edge and the line that proxi- mally connects the inner and exterior edges of the wing. Papposphaera sagittifera HET is generally con- ceived as having monomorphic coccoliths yet with a tendency towards a varimorphic state in as much as there is variability among coccoliths in the length of the central process and with the longer ones prevail- ing in a circum-flagellar cluster. The SEM images ( Figs 10–12, 14 ) clearly and consistently show that the actual pattern of variability goes beyond trivial dimensional issues. The major part of the coccoliths covering the cell surface – with the exception of the apical and an- tapical poles of the cell – are in fact devoid of a central process. All that is left is a short stub-like feature ( Figs 8 , 14 ) or just a low mound created by the lifting up of the arms of the axial cross where they meet centrally ( Fig. 9 ). It is of course possible and also relevant to interpret these as being just very much reduced stand- ard calicate coccoliths without a central appendage. However, it does anyway represent a differentiation among coccoliths that is almost equal to having dimor- phic coccoliths. TEM images (e.g. Figs 4, 7 , 13 ) are less informative with respect to illustrating the distri- bution of calicate and non-calicate coccoliths within a coccosphere. The electron beam does not adequately penetrate the middle part of the cell leaving the massive appearance in this area of non-calicate coccoliths dif- ficult to resolve. The two-dimensional TEM image and a certain repositioning of individual coccoliths caused by the handling of the cells during processing, further add to the interpretational challenge with respect to ac- counting for the actual positioning of calicate versus non-calicate coccoliths. Disregarding these reserva- tions the cells illustrated in Figs 4 and 13 do appear to have a few calicate coccoliths positioned in the middle part of the coccosphere. This might indicate a possible regional or seasonal variability among specimens of P. sagittifera that we do not at present have material to possibly account for.