Rodríguez-Juiz et al. 1996

Rodríguez-Juiz AM, Torrado M, Méndez J. 1996. Genome-size variation in bivalve molluscs determined by flow cytometry. Marine Biology 126: 489-497.

These authors measured nuclear DNA contents in 10 species of bivalves of commercial importance. Genome size variation in plants and poikilothermal animals had previously been associated with life-history and ecological traits, suggesting links between genome size variation and speciation events. There had been few previous studies of DNA content in molluscs. Early examples of such studies include Mirsky and Ris (1951), Hinegardner (1973, 1976), and Cavalier-Smith (1978); several earlier papers on molluscs and other poikilotherms related genome size to “specialization”, particularly the work of Hinegardner and colleagues.

Unlike most studies of DNA content, these authors included relatively large samples of each species, using 20 individuals in each species. All individuals were purchased from commercial shellfish sellers, either in Spain (9 species) or the Netherlands (Mytilus edulis), and maintained alive in the laboratory until dissection of gill tissue; the assumption that gill tissue is diploid is never stated explicitly but was used in the calculations of genome sizes. Tissue was placed in filtered, autoclaved seawater and subjected to mechanical shaking for 30 minutes; these authors do not describe in detail this shaking, is there a standard rate and magnitude of mechanical shaking of mollusc tissue? The presence of isolated cells was verified using a microscope, and cell preparations were strained through 15mm mesh and sonicated for two minutes to remove cell membranes. The nuclei were then centrifuged, resuspended in buffer that appears similar to Galbraith’s buffer (Galbraith et al. 1983), and fixed with 0.1% formaldehyde on ice. Finally, aggregations of nuclei were disrupted using a 26-gauge needle, pumped three times.

Two internal standards were employed: Capsicum annuum and chicken red blood cells (CRBCs). Isolated nuclei from these species were added to bivalve nuclei suspensions before staining, thus standards and specimens here were co-stained rather than co-prepared. The CRBCs produced a peak in the flow cytometry histograms overlapping seven of the 10 bivalve species, thus the introduction of the plant nuclei. C. annuum nuclei were employed after checking for consistent measurements with the three species of bivalves that did not overlap in peak area with CRBCs, and comparison between C. annuum and CRBCs to determine a C. annuum diploid nuclear content of 8.4 pg, larger than any bivalve measured in this study. 10 000 nuclei were measured per histogram, presumably this means total events recorded above the debris cut-off, and each specimen was measured three times.

These authors report significant intraspecific genome size variation in all 10 bivalve species. Interspecific (and between higher taxa) was much greater than intraspecific variation, but the intraspecific variation was statistically significant under 2-way ANOVA and “GSD” calculations based on the work of Gold and Amemiya (1987) and Alvarez-Fuster et al. (1991).

The discovered and possibly unexpected intraspecific variation is used to bolster an argument made in the discussion that large samples are necessary for accurate determination of genome size and genome size variation in species. This explains the difference between these results and the no-intraspecific-variation results of some previous authors that did not use large samples per species.

Following this is a discussion of Hinegardner’s (several papers in the 1970s) “specialization” assumption / hypothesis. It is described as one, then the other. These data do not support this hypothesis, which is not surprising considering how vague and taxon-specific the terms “specialized” and “generalized” are, and their underlying assumptions about species and lineage ages and rates of evolution.