Chromosome counts

I have been trying to get chromosome numbers for a number of scleractinian coral species, but in each of my preparations I obtain a wide range of chromosome counts, typically ranging from 24 to 30. The mode of the distribution seem to indicate in most cases a basic number of 28, but as this mode represent less than 50% of the observations it is far from being totally convincing. Here is the protocol I've been using: 10-11 hours old coral embryos were cultured with colchicine 0.02% in seawater for 2 hours, followed by a 20 minutes 65:35 seawater:tapwater hypotonic treatment. Treated embryos were then fixed in three changes of freshly mixed, absolute ethanol:50% glacial acetic acid (1/1, v/v), and refrigerated during storage. To remove lipids prior to staining, fixed embryos were soaked in diethyl ether for 4-6 hours, then briefly returned to fixative. Embryos were stained with 2% lacto-aceto-orcein on a glass slide for 15 minutes, gently rinsed with tapwater, and squashed under a cover slip. Dehydratation of the quash preparations was retarded for several days by sealing the cover slip edges with clear nail polish. (from Kenyon, Evolution 51(3), 1997, pp. 756-767).

I found in the litterature that the same problem occured with sea anemones (e.g., Fukui, J.mar.biol.Ass. U.K. (1993), 73, 971-973) using Meredith's method for mammalian tissue (Chromosoma 26 (1969), 254) and in hydra (Rahat et al, Experientia 41 (1985), 282-283) using also this method (Pieces of tissue were transfered into a Dryer tube containing 0.5ml 60% acetic acid. Within 5 minutes the tissue loses its coherence, and a cell suspension could be formed by aspiration with a finely-drawn pasteur pipette. A drop of cell suspension was then transferred onto a microscope slide on a hot plate at about 60oC and immediately withdrawn. The same drop of cell suspension was applied to the slide in this manner 5-10 times before discarding. The procedure was repeated until all the fixed material was applied unto the slide. The slides were places vertically for 10 minutes in a 5% aqueous solution of Giemsa stain, rinsed with tap water and air-dried.).

I wonder if anybody out there has experienced this kind of problem doing chromosome counts in any species ; any suggestion is welcome! I can also send some pictures of the preparations I obtain if it can help to indentify the problem. I am not an histologist, but a graduate student with a background in biochemistry trying to get my protocols working with no technical expertise available in my university...