Identification, Images, & Information
For Insects, Spiders & Their Kin
For the United States & Canada
Clickable Guide
Moths Butterflies Flies Caterpillars Flies Dragonflies Flies Mantids Cockroaches Bees and Wasps Walkingsticks Earwigs Ants Termites Hoppers and Kin Hoppers and Kin Beetles True Bugs Fleas Grasshoppers and Kin Ticks Spiders Scorpions Centipedes Millipedes

Fall Fund Drive

Potential use of Bugguide data for phenological studies

It has been suggested that the accumulated data in Bugguide can be put to use in a number of ways in the future, for instance in studies of populations, geographic distribution and phenology. Despite the lack of a coherent methodology in the gathering of data, such use may be possible and it may be possible sooner than expected. To test this hypothesis I used the Advanced Search to construct monthly charts of several bee genera.
I used the raw data. For a more advanced study it would be necessary to discard irrelevant images, such as empty nests, leaf damage, etc. It would also be preferable to use only one entry for each specimen rather than variable numbers depending on the number of images of the same individual. In many cases, only adults or immature stages would have to be selected. This didn’t seem necessary for the purpose of this work. Despite some possible "noise" in the data, certain patterns seem apparent.
The graphs include all the images of the respective genera submitted to Bugguide from its start in February 2004 to March 27, 2011.

Graph 1 illustrates the monthly numbers of images for two bee genera: Andrena and Megachile. Andrena populations peak in April and May, while Megachile are most numerous between June and September.
Graph 2 shows the monthly distributions of populations of Megachile and Coelioxys. The cuckoo bees of the genus Coelioxys specialize on Megachile bees. The rise and fall of their populations follow those of their hosts very closely and their numbers are smaller just as expected.
Graph 3 shows the phenology of Andrena (blue), the cuckoo bee Nomada (red) and the parasitic beefly Bombylius (yellow). The populations of Nomada and Bombylius follow those of Andrena. The numbers of the two parasites are rather high when compared to those of Andrena. If Andrena were the only hosts of these parasitic bees and flies, they would not be able to sustain such large populations. However, Nomada parasitizes Agapostemon and Eucera in addition to Andrena; Bombylius uses Colletes and Halictus as hosts in addition to Andrena.

None of this is news; it just confirms other observations. In fact these charts can be compared to the ones developed by Sam Droege (The Weekly Phenology of Bees of the Mid-Atlantic States: MD, VA, WV, DC, PA, DE) using collections from the USGS bee database. The results are similar.
It is nice to see how Bugguide data can be used. Its potential will grow in the future. It would also be possible for the advanced search to generate data sheets and charts like the ones presented here, instead of copying the numbers manually as I did for this little study.

Update, Dec.20/2016. Sam Droege has published newer Phenology of Bee Genera, Midatlantic States. USA.

slippery slope
This goes back to the old (and often recurrent) arguement of the value of state data points on BugGuide. Images on BugGuide are random images take of random insects by random photographers who happen to be at a particular spot where a particular insect happens to be at that moment.

It can be dangerous trying to extrapolate such data based on image submissions. I am not saying it cannot be done, one must just be very cautious, especially is someone plans on using such data for publications, etc.

In the examples above, you took into account things like not using damage evidence, empty nests, etc. But what about geographic distribitions? Do you have data from the South intermingled with data from the North?

I Wish It Was Random
Hi Blaine:

Isn't randomness what we want?

I'm more concerned that BG data is full of bias. When compared to, say, Christmas Bird Count data, which is more controled and/or corrected for area covered, time in field, weather, other conditions and is more of a true surve in nature, BG data may be even more biased with regard to detectability, thoroughness, identifiability, asthetic apeal and whatever biases one to frass data.

I really appreciate your concerns and Sam's comments.

random data points
At the scale of al this data doesn't randomness, become only another form of distribution that can be accounted for when compiling? So a query could take range into account. And the data should be compared against other queries for normalization. Not sure what bugguide uses on the back end, but even MySQL could do this type of thing.
At some point the data may be useful. Even if only for presented as an example of the use of random/anecdotal information about ecosystem change. Someone else may pick up on the data and use it in a way that is novel.
While I applaud research based on hard data, excluding anecdotal evidence precludes thinking about our world in a way that is time tested by aboriginal cultures. We need both and the real usefulness comes from using them supplementally.
References for integrating the two types of evidence could include,


Very nice work!

I appreciate the discussion of the noise.

We'll get to track change (if any) in phenology over the years.

How did you do this? Did you take pictures of the graphs you made or screen capture or what?

I use Power Point. It has a tool for creating a graph from a chart. All I do is feed the numbers and choose the color of the bars. Then I just copy the graph by clicking on it and I save it as JPG. Probably there are better ways but this works for me.

Phenonlogical data visualization
Though this is a year old, my 2cts worth. Excel would probably be a better tool for this. Using Excel gives one more powerful filtering choices as per images, dupes, searching. Then you can create the graph as well as export as a web page to allow people to use the collected data in ways they would need or want.

Good to know. I confess that I never learned much about Excel. Maybe somebody will be willing to try that with other sets of data. As data keep accumulating the info becomes more valuable all along.
I know an author that has been relying on Bugguide maps to check the distribution of moths, even though the data is still terribly incomplete in most cases.

Great idea
I love it.

Comment viewing options
Select your preferred way to display the comments and click 'Save settings' to activate your changes.