Pollen: not just amorphous blobs that make people sneeze

Last semester, I wrote about the process of bee identification as I used DiscoverLife to work on sorting our bee inventory. This semester, I’ve been doing something similar: pollen identification. Pollen identification has a range of applications, from fossil dating to track the evolution of plants to plant physiological studies to studies on the foraging behavior of pollinators. In our recent pollen foraging project, we are looking at the diversity of pollen collected by honeybees in different treatments – that is, a normal colony, and a colony with an experimentally impaired communication system. Our goal is to see whether honeybees use communication to collect a great diversity of pollen, or if its function is more like how it works in nectar foraging, where the colony shares information to converge collectively on the best resources. There are reasons to suspect either strategy – pollen diversity seems to be valuable in honeybee diet, but some pollen sources could be more fruitful than others – so it will be interesting to see in which direction our data points.

So, basically, we have many samples of pollen that were collected from each colony during the experiment using a pollen trap, and I made microscope slides of those samples to get a representative sense of their diversity. For the past few weeks, the task has been to distinguish between and count the types of pollen on the slides, and to identify those types. But unlike the process of taxonomically sorting bees or other biota, the world of pollen sorting doesn’t avail me of nice tools like DiscoverLife. And since this was really my first time looking at pollen under a microscope, I often found myself asking some of the same questions as Jingnan in his computer vision project: is this blob-looking-thing different from that other blob-looking-thing and why?

As it turns out, pollen grains that generally look like blobs actually have a range of detailed surface characteristics and different shapes. The image below from the Dartmouth Electron Microscopy Facility gives a sense of just how unique different pollen types can be.

Image 1. Pollen diversity viewed through an electron microscope

Unfortunately, the bee lab does not have an electron microscope to play with. Our microscope does show some level of detail, and we can clearly see different types of pollen. These images are representative of what I was looking at in my slides. You can see that in some cases, discrimination is pretty easy; in others, not so much.

Image 2. California buckwheat (Eriogonum fasciculatum) pollen in one of our samples

Image 3. Solanum umbelliferum pollen

Image 4. Erodium botrys pollen

Image 5. An example of a frame from one of the sample slides I made. Using image 6 as a reference, I would distinguish the leftmost three grains from the rightmost grain on the basis that they appear to be tricolporate, and also have a more distinct border. 

You can also see that even pollen grains of the same type may appear different in size, detail, and so on. What kind of criteria could we use to draw the line between normal variability within a “type” (family, species, etc.), or variability that defines different  types? There are a few basic characteristics visible under our microscope that I used to discern pollen types: some examples include the shape of the grain, the presence of one or more furrows, surface texture, and the number and shapes of apertures (image 4). For example, I would have been able to distinguish the E. fasciculatum pollen (image 2) from S. umbelliferum (image 3) because E. fasciculatum appears tricolpate. Even though these characteristics don’t necessarily always correspond to a certain taxon (e.g. many taxa are monoporate), they are good indicators of similarity to use when looking at reference slides of previously collected and identified samples. After doing some sketches of pollen collected during summertime surveys at the BFS in 2015 and 2016, I pick a few possible matches among the reference slides for each slide that I made, and use those characteristics to decide if they could be the same. Some characteristics that seemed less reliable included small variations in the size of the grain, though general size is a helpful grouping factor.

Image 6: a range of pollen types based on furrow and pore combinations.  

This method, while time-intensive, worked quite often. As expected for southern California during the summertime, I identified a lot of buckwheat pollen. But most of the pollen types I identified got catalogued as “unknown #1, #2, #3…”. They didn’t match anything in our reference slides, though they appeared multiple times in our own samples. Since we’re mostly interested in looking at pollen diversity, we can deal with counting unknowns as long as we recognize what’s the same and what’s different, but it might be good down the road to get a sense of what exactly the bees are foraging on. It could be that they’re foraging on flowers in people’s gardens, which would explain why they don’t match any of our references. After all, the foraging radius of bees is known to reach, or perhaps even exceed, seven miles, and our field station is surrounded on all sides by suburbia. In the future, we might be able to do some genetic analysis on our samples to identify what’s in them.

Image Credits

[1] image from the Dartmouth College Electronic Microscope Facility

[2], [3], [4] image taken by Saul Gonzales

[5] image taken by Tessa Finley

[6] image from the Paleoecology department website of Universität Bern, originally published in G. Lang (1994).

Further Reading

“Tiny Grains of Fossil Pollen Tell a Big Story.” National Museum of Natural History Unearthed. Accessed December 7, 2016. http://nmnh.typepad.com/100years/2015/06/tiny-grains-of-fossil-pollen-tell-a-big-story.html.

Pernal, Stephen F., and Robert W. Currie. “The Influence of Pollen Quality on Foraging Behavior in Honeybees (Apis Mellifera L.).” Behavioral Ecology and Sociobiology 51, no. 1 (December 1, 2001): 53–68.

Williams, Joseph H. “The Evolution of Pollen Germination Timing in Flowering Plants: Austrobaileya Scandens (Austrobaileyaceae).” AoB Plants 2012 (January 1, 2012): pls010.

Wilson, Erin E., C. Sheena Sidhu, Katherine E. LeVan, and David A. Holway. “Pollen Foraging Behaviour of Solitary Hawaiian Bees Revealed through Molecular Pollen Analysis.” Molecular Ecology 19, no. 21 (November 2010): 4823–29.