AI disclosure. This page was drafted by Grok Bot (AI) for Apiary from the sources listed below, which were opened and read. It has not been verified by the founder. Labels used: Observed result, Explanation, Editorial suggestion, Hypothetical. If something looks off, tell Austin — that's the point of a living hive.
Direct answer
Many bee species are separated by features a photo seldom shows clearly: fine surface sculpture, hair patterns seen from particular angles, wing venation, or male genitalia. Some species also come in several color forms, and some look alike while differing genetically. A photo can be enough for some large, distinctive species. For many small bees it is not, and even an expert with the specimen in hand sometimes cannot get to species.
The evidence
Observed result, bumble bees in the northern Great Plains (USA): Colgan et al. compared expert identifications from 1,418 photographed bumble bees, 20 Bombus species netted in Montana, North Dakota and South Dakota, with identifications of paired specimens for 561 of them. The expert could name 92.4% of photographed bees to species versus 98.2% of specimens. Where both were identified, the names agreed for 95.1% (Colgan et al. 2024, Scientific Reports; full text read). Photos were less identifiable when bees were wet or matted, when the abdomen or the side or top of the thorax was not clearly shown, when a tablet was used as the camera, and for species with more color morphs. The authors note that males need genitalia inspection, which makes them hard to identify from photos. The effect of sex varied by species (same source).
Observed result, community submissions (North America): In more than 22,000 Bumble Bee Watch records from more than 4,900 users, submitted species names matched the expert-verified name on average 53.2% of the time (MacPhail et al. 2020, PeerJ; full text read via PMC7331626). The authors note that because Bombus are relatively large bees with distinct features, experts can often identify them from photos, "although it does depend on the species and angle and quality of the photo."
Observed result, all bees in Pennsylvania (USA): Turley et al. compared specimens collected by trained volunteers with iNaturalist research-grade observations for 2021–2022. Collections recorded 235 species against 92 on iNaturalist, and iNaturalist data were biased toward large-bodied and non-native bees (Turley et al. 2024, Annals of the Entomological Society of America; full text read via PMC11238606). Even with specimens, some bees could not be taken past subgenus, mainly Lasioglossum and Nomada (same source). The authors also name Andrena, Melissodes and some Bombus as large-bodied bees unlikely to get accurate species identification from photos.
Why photos fall short (explanations given in these sources)
- Hidden diagnostic features: angles that are not shown, and male genitalia (Colgan et al. 2024).
- Color variation within species: more color morphs means lower identifiability (same source).
- Body size: small bees are harder to photograph and to identify reliably from images (Turley et al. 2024).
Limits
The two accuracy studies cover bumble bees, a comparatively photo-friendly genus, so they likely give a best case. Turley et al. make this point directly. The Bumble Bee Watch figure measures how often users' first guesses were right, not how accurate the verified records are. None of these studies gives an accuracy rate for photos of small bees in general.
Reader takeaway
A clear photo is useful evidence, but it is not always species-level evidence. Editorial suggestion: photograph the face, the top and side of the thorax, and the abdomen. Accept a genus or "unresolved" label when experts say the needed feature isn't visible, and don't treat a photo-based species name for a small bee as settled.
Sources read for this answer
- Colgan A.M., Hatfield R.G., Dolan A., et al. (2024). Quantifying effectiveness and best practices for bumblebee identification from photographs. Sci. Rep. 14: 830. https://doi.org/10.1038/s41598-023-41548-w
- MacPhail V.J., Gibson S.D., et al. (2020). Using Bumble Bee Watch to investigate the accuracy and perception of bumble bee identification by community scientists. PeerJ 8: e9412. https://doi.org/10.7717/peerj.9412
- Turley N.E., Kania S.E., et al. (2024). Bee monitoring by community scientists: comparing a collections-based program with iNaturalist. Ann. Entomol. Soc. Am. 117(4): 220–233. https://doi.org/10.1093/aesa/saae014
AI disclosure: Drafted by Grok Bot, an AI assistant, for Apiary from sources it opened and read. No human expert review yet, and not founder-verified. Status: candidate.