On the distribution of diatoms on cover slips: cause for concern

Abstract

We investigated the effect of frustule morphology on the distribution of various diatom taxa on microscope cover slips, using an artificial assemblage. Seven diatom morphotypes were processed separately in nitric acid and an eighth morphotype (Chaetoceros muelleri) was processed in Lugol’s. These were mixed into a single assemblage and dried on a cover slip. We then mapped the locations of all 1,664 valves in the debris area, which covered 244.4 mm2. The inner half of the debris area (A = 122.2 mm2, r ≤ 6.237 mm,), the “Central Region,” contained 1,303 valves (78 % of the total). Each of the eight morphotypes was significantly (p ≤ 0.035) more numerous in the Central Region than in the Marginal Region. In addition, the assemblages in the two Regions were quite distinct: C. muelleri accounted for 19 % of the Central Region diatoms, but 57 % of the Marginal Region ones, whereas Surirella peisonis accounted for 17 and 6 %, respectively. We also considered valve distribution relative to aliquot volume by comparing the number of valves under the proximal 50 μl of the aliquot (r ≤ 4.758 mm) with the number of valves under the distal 50 μl. Four morphotypes were significantly (p < 0.02) more numerous in the Proximal Region. In contrast, valves of C. muelleri were significantly more numerous in the Distal Region (p < 10−43). Valve distribution may reflect valve morphology, such as spine length and degree of silicification. Also important may be other environmental variables, such as the abundance and size of sediment particles. Although preliminary, these data indicate that some taxa are not distributed randomly on cover slips, especially valves that have long spines, thin walls, or thick walls. Whereas many natural assemblages lack such morphotypes, investigators must be aware of such potential heterogenous distributions when counting.