Skip to main content
ARS Home » Research » Publications at this Location » Publication #189186


item Jones, Gretchen

Submitted to: Grana
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/12/2006
Publication Date: 3/8/2007
Citation: Jones, G.D., Bryant, V.M. 2007. A comparison of pollen counts: Light versus scanning electron microscope. Grana. 46:20-33.

Interpretive Summary: Scientists who study pollen mainly use a light microscope (LM) to count and identify pollen grains and a scanning electron microscope (SEM) to look at differences among pollen grains. There are no published reports comparing the LM and SEM pollen counts and pollen identifications of the same sample. The purposes of this paper were 1) to examine the advantages and disadvantages of each microscope, and 2) to compare the pollen counts and pollen identifications from the same surface soil sample using both microscopes. Pollen from 73 plant species was found in the sample. Three pollen types occurred most frequently regardless of the microscope used. However, significantly more pollen types were found with SEM. Sample preparation and the time needed to count, analyze, photograph, and print the pollen pictures were similar for both microscopes. The microscope needed depends on the sample, desired information content, and available funding. A SEM is more appropriate when very specific information is needed, and a LM is appropriate when pollen data is needed very quickly. This is the first time pollen data from LM and SEM analyses of the same sample have been compared. Results of this study demonstrate the unique value of LM and SEM for pollen identification, which is invaluable for applications including identification of food resources of harmful and beneficial insects and identification of allergens.

Technical Abstract: Palynologists rely on a compound light microscope (LM) as the main microscope for pollen identification and the interpretation of the pollen spectrum of a particular sample. A scanning electron microscope (SEM) is mainly used for morphological comparisons and taxonomy where the increased resolution of SEM makes differentiation among pollen grains and taxa more obvious. Because we are unaware of any published reports comparing LM and SEM counts and pollen identifications of the exact same sample, we decided to test the advantages and disadvantages of each microscope. A surface soil sample collected from a city in Southern Europe was used for this experiment. Standard palynological techniques were used to extract the pollen. Just before adding a mounting media (glycerin), a small sub-sample was removed and used for SEM analyses. Two, 300-grain counts were made using LM, and two, 300-grain counts were made using SEM. Pollen grains viewed with SEM were not only counted and identified but also divided into three categories, “identifiable,” “obscured,” and “virtually impossible to identify." The majority of the pollen grains counted with SEM (86%) were classified as either “identifiable” or “obscured.” Pollen concentration values ranged from 385,714 (LM Count #2) to 900,000 (SEM Count #1) per gram of soil. Regardless of the technique (LM or SEM), the same three taxa (Ligustrum spp., Myrtaceae-type, and Tilia spp.) had the greatest number of pollen grains. A total of 73 unlumped taxa were found in the four counts. A scan of the unexamined portion of the stubs resulted in 20 unencountered taxa bringing the total number of unlumped taxa to 93. There were no significant differences among the counts or between the two techniques (ANOVA, P > 0.05, F = 0.18, df = 76). However, there were significantly more taxa found with SEM than with LM (T-test, T = 0.05). Sample preparation, and the time needed to count, analyze, photograph and print the micrographs are about the same regardless of which technique is used. The type of sample, the information needed from the sample, and the funding level available for the research will in most cases determine which technique to use. A comparison of a greater sample size is needed to determine if the differences we found are consistent across the vast diversity of palynological applications.