---
canonical: https://biosonic.io/myotis-auto-id
---

# Can you trust automatic bat ID for Myotis species?

Last updated: 2026-11-22. Written by the BioSonic team.

Trust auto ID to find likely Myotis recordings, and be careful about the species name. The BCT guidelines point to the difficulty of identifying Myotis reliably from echolocation calls alone, so check the records, report Myotis sp. where the call does not support a species, and trap when the decision depends on the species. BioSonic and other classifiers can suggest a Myotis species with a confidence, but none of them replaces that check.

We make BioSonic. The method on this page works with any bat analysis software.

## What to report

| What you have | What to report | Next step |
|---|---|---|
| A clear Myotis call with no diagnostic features | Myotis sp. | Keep the recording and note who checked it. |
| A species suggestion backed by several clear calls | The species as probable, with the reason | Have an experienced analyst confirm it. |
| A likely Bechstein's bat, or impacts that differ by species | Myotis sp. until confirmed | Consider trapping, as BCT section 9.2 describes. |
| Whiskered bat from sound alone | Not as a confirmed record | BCT says checking audio is usually not conclusive for this species. |

## Why the species name is uncertain

- Myotis species make similar frequency modulated calls, and the shape changes with clutter and distance.
- In a test in Sweden of SonoChiro, Kaleidoscope and BatClassify, correct species identification for Myotis ranged from 4% to 93% (Rydell et al. 2017).
- In a 42 square km project described by BCT, detectors recorded far lower proportions of every Myotis species than trapping caught.

## A way to check them

1. Filter to all Myotis suggestions, whatever the species.
2. Check every one if there are few. If there are many, check every locally uncommon species and a random sample of the others.
3. Compare call shape, start and end frequency, and interval with a known reference.
4. Move what you cannot support to Myotis sp. and report how many you moved.

## In BioSonic

Each identification has a confidence. A reviewer can filter by species and confidence, open the spectrogram and change the species. A high confidence is not proof of a Myotis species.

## Common question

### Is a high confidence score enough for a Myotis record?

No. BCT warns that probabilities from auto ID are not normally suitable for use without further analysis. Verify species that are locally uncommon or that the classifier does not handle well.

## Sources

- [Bat Surveys for Professional Ecologists, 4th edition](https://www.bats.org.uk/resources/guidance-for-professionals/bat-surveys-for-professional-ecologists-good-practice-guidelines-4th-edition) (Collins, ed., Bat Conservation Trust, 2023, amended March 2024), sections 9.2 and 10.2 and Box 9.2.
- [Guidelines for passive acoustic surveys of bats in woodland](https://www.bats.org.uk/our-work/science-research/passive-acoustic-surveys/guidelines-for-passive-acoustic-surveys-of-bats-in-woodland) (Bat Conservation Trust), on whiskered bat.
- [Rydell et al. 2017, Testing the performances of automated identification of bat echolocation calls: A request for prudence](https://doi.org/10.1016/j.ecolind.2017.03.023), Ecological Indicators.

Related: [Published classifier accuracy tests](https://biosonic.io/bat-classifier-accuracy) · [What BCT says about auto ID](https://biosonic.io/bct-guidance-auto-id) · [Auto ID for static detector data](https://biosonic.io/static-bat-detector-analysis)

## Test it on a survey you know

Upload recordings you have already identified and compare the results and the time it took.

[Start free trial](https://bat.biosonic.se/signup)
