Berkeleya cf. hyalina (Round & M.E.Brooks) Cox

Lobban, Christopher S. & Witkowski, Andrzej, 2023, Marine benthic diatoms of Guam: new records, Dictyoneis apapae sp. nov., and updates to the checklist, Micronesica 2023 (2), pp. 1-75 : 4-5

Berkeleya cf. hyalina (Round & M.E.Brooks) Cox Figs 15–25

Ref. illus.: Round & Brooks 1973, figs 1–4, 6–16; Cox 1975a, figs 7, 33, 35; Cox 1975b, figs 9, 18, 22, 23; Chastain & Stewart 1985, figs 1–14; Lobban 1985, figs 3–5; Witkowski et al. 2000, p. 156, pl.62, fig. 22

Samples: GU77A-2

Observations: Colonial (tube-dwelling) species forming wide, branched colonies with many narrow tubes bundled together, each narrow tube containing a single file of cells ( Figs 15–17); plastids Hshaped ( Fig. 18). Valves hyaline, 45–51 µm long, 6 µm wide, striae parallel except convergent near apices, 37 in 10 µm. Cells linear-lanceolate with broadly rounded ends ( Fig. 19). In SEM, striae parallel except radiate near the apices, areolae quadrangular except a row of larger areolae along one side of the raphe ( Figs 20, 21). Central raphe endings separated by 7.1–8.6 µm, curved slightly away from the side with the larger areolae.

Comments: These microscopic colonies are very much smaller than those in southern California but that is not unusual for Guam tube-dwelling diatoms (so far only Homoeocladia [ Nitzschia ] martiana C.Agardh colonies found at seaweed size; Lobban & Tsuda 1993). Our cells appear to differ from B. hyalina in larger size, wider separation of central raphe endings and shape of the apex, from B. fragilis in shape and longer central area, and from B. micans , which is longer but narrower with

coarser striae and very long central area ( Table 1). However, variability in some of these characters warrants caution.

First, there are discrepancies between Chastain & Stewart’s (1985) California populations compared to type material, some of them noted by those authors, i.e., the stria density, the shape of the apex, and the extent of the axial row of enlarged areolae. The difference in stria density might be significant but Chastain and Stewart’s statement is broad and imprecise (“approximately 32–38” in 10 µm), and Round & Brooks’ is even more vague (“more than 40”). Lobban (1985) in a floristic survey including southern California, reported striae of B. hyalina to be “ca. 45 in 10 µm… visible only in the electron microscope” but no SEM was published. Perhaps a more significant difference lies in the hemilanceolate depression formed by the very long areolae in San Diego populations, which Chastain & Stewart noted as a “signal feature,” though saying that “the row of enlarged pores… appears to be somewhat more prominent” than in the type material. All species of Berkeleya in Table 1 are reported in the literature to have a row of enlarged areolae along one side of the raphe, so what is distinctive in the San Diego populations is the width and the depression. Round & Brooks (1973, figs 7, 8, 15) noted that the larger pores appeared to be in a “slight depression;” the pores are trapezoidal, wider toward the axis. Chastain & Stewart (1985, figs 6, 10) show a single row of larger areolae in their TEM (fig. 6) but in the SEM (fig. 10) most in the widest part are divided. We have seen such depressed zones in a few samples that we could not identify to species ( Figs 24, 25) and even, rarely, in the present colony. In all these, the distance between central raphe endings was <7 µm but perhaps not significantly different from the majority of the colony (in the two specimens in Figs 24, 25 = 4.2 and 6.2 µm but in Guam specimens up to 6.7 µm). This brings into question whether the depressed area is a consistent enough character to use for taxonomy, and how much distance between central raphe is needed to separate species. Chastain & Stewart noted the great difference in latitude and water temperature between Togo (type locality) and San Diego, to which we might add, the different ocean basins.

Second, there are the differences between our material and both of the other two populations which require additional material from Guam and re-examination of type material (or better, more samples from Togo as well!). At this point both the Guam and Togo populations can be considered inadequately studied. The biggest problem with the Togo material is the vague stria density given by Round & Brooks; we tried counting it from the published images ( Round & Brooks 1973, figs 8, 15) and got answers very much higher than the>40 given, but only magnifications are given for the images and perhaps the plates were reduced in publication without correction of the captions. Thus, we have identified our specimens only as “ cf. hyalina ” and shown their morphology for future reference .