Community-wide Comparative Identification Experiment

Welcome to the community-wide project on testate amoeba identification!

We are leading a community-wide project to assess the effect of identification bias on ecological interpretation of testate amoeba data.

Invitations have been sent to all colleagues working on testate amoebae, if you have not been contacted and wish to participate, please contact Liza: elizaveta.ermolaeva at

All data contributors will be co-authors on a joint paper. However to qualify you need to analyze 10 samples (100 image each).

All files can be found here:


Please read carefully these instructions. Please read also the questions and answers at the end of the document.

1. For each sample pictures of 100 individuals were taken. In some cases, two pictures of the same individual were taken (focusing on different parts) to facilitate the identification. Be careful to check the picture number from time to time to make sure you are entering data for the correct picture number. There are 100 specimen per sample but most species will appear more than once. This is typically what we have when we do community ecology work. With this we can then calculate the relative proportion (%) of each species in the sample.

2. We suggest you analyze the samples in the order indicated as the first ones should be easier (the palaeo samples come last).

3. For each sample, please identify the specimen one by one and for each specimen indicate an identification confidence level as follows: A: 90-100% certain, B: 50-89% certain, C: 10-49% certain, D: 0-9% certain.

4. Although ideally each image should be identified as a separate object, we know this is not how things are typically done in (palaeo)ecological analyses. Therefore, you are allowed to go back to previous specimen to correct an identification. This can happen for example if the first image of a species is confusing but then you see several specimens of the same species which lead you to revise your identification of a previous image. However, in any case, the confidence level of each identification should reflect your confidence of identification for that given image and not for the species in general.

5. We kindly ask you to select just one name for each image, as you would do in an ecological study. If you hesitate between two or more names for a specimen, please chose just one and select the confidence score according to your level of confidence. For example, you may think that a specimen could be species X or species Y, but more likely species Y. You would evaluate that there is a 60% chance that it’s species Y and a 40% chance that it’s species X. You would thus call it Species Y with a score of 60% (hence a confidence level B).

6. Please give names only to testate amoebae. If you think that something else is shown, please write “NA”. You can also write your guesses in parentheses, for example, “NA (chicken)”.

7. If you cannot think of any specific names but you still think that it is a testate amoeba, please make a guess (your confidence score will typically be the lowest in this case). For example, in some cases, it will be difficult to select one species within a genus and thus the name will be genus sp. (e.g., Difflugia sp.), or you would typically call it species X-type. In more extreme cases, even identifying a genus may be impossible but you are quite confident that it’s a testate amoeba, and thus the name could be “testate amoeba sp.”.

Please follow this overall framework while putting names on a sheet:
• Full species name, e.g. “Nebela tincta”
• For subspecies or infrasubspecific taxa such as forms and varieties, use just three words, e.g. “Nebela tincta minor”
• Morphotype, e.g. “Nebela tincta type”
• Genus name, e.g. “Nebela sp.”
• Testate amoeba sp.
• “NA” or “NA (something)”

8. Please note that there can be several individuals on the same picture. If your target is not obvious, it is most likely to be in focus, located close to a centre, or marked with a red arrow. For palaeo samples, specimens are marked with a red circle.

9. The majority of pictures were taken at 40x magnification. However, some images were taken at 20x magnification. These pictures have a warning sign.

10. Please take your time and check your spelling to prevent confusion between several similarly spelt names.

11. If you feel that some comments are needed regarding your identifications, feel free to write these in a separate “comments.txt” file.

12. All identifications for the 10 samples are to be entered in a single spreadsheet. With the "freeze planes” option you will always see the name of the sample and photo (lines 1 & 2) as well as the species names (1st column).

13. Each time you see a news species type in its name in the first column. This way you only need to enter the species names once for the 10 samples. As the number of species increases you may find yourself looking for the names and losing a bit of time doing this. If you want to sort the names (e.g., alphabetically) to make data entry easier, please make sure to select the whole lines otherwise you will mess up your datasheet. Unless you are familiar with this procedure it is perhaps best to just keep the names in the order you have entered them.

14. There is only one file for all samples. Please enter the data as you proceed with the identification (save the file regularly!) and once you are done there will be just one file for all samples. This way we also have only one list of species to deal with, which is much less work for us than having to compile 10 separate files.

15. Please note that the blue cells are filled automatically with formulas, please don’t edit these. These columns will be hidden (collapsed) in the document you receive:

Done! 😊

Questions and answers

Q: Deadline: can we have an extended deadline?
A: Yes it's fine if you do this until end March. However it would be nice to start getting some answer soon so that we can evaluate how well this project goes and do some first stats. If end March is still too early for you please let us know. We need the results rather fast as we hope to present this study at the next ISTA!

Q: Do I need to send a separate file for each sample?
A: No all results should be given on a single excel file which was provided. It is a large file indeed, but having just one file means that we will only have one liste of species per participant, which makes our life much easier!

Q: Is there only one species per sample?
A: No Every image is potentially another species! For each picture you need to determine which species it is. For example image 1 will be species 1, image to may be species 1 or another species and so on. You could potentially have 100 species in a sample as you have 100 images. However this is not the case as the same species will appear several times. Most likely you will have something like 10-30 species in a sample.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License