How do you people identify all these critters?

Question

I am always amazed at the answers in species-identification, especially for insects and plants. It seems like a grainy photo, a vague description, and very broad geographical location (often just a continent) is all that it takes to pinpoint a species. Yet there are millions of species! That seems remarkably efficient from an information theory standpoint to me.

Before asking online I always try to identify my critter myself first, and I never have much luck.

• Searching by traits ("large green head with black oval eyes and slender body") turns up too much irrelevant information
• Binary tree style identification flowcharts seem hard to find, and having used those to identify microorganisms I get the impression that they tend to be very imprecise
• Phone apps and similar software seems absolutely useless 99% of the time

But somehow the answerers are able to identify species so easily and so accurately. How do you do it? What's your secret? Do you just have encyclopedic knowledge of all the common species out there? Are there just lots of clade nerds lurking this site, who each know everything about one cluster of species, and it's just that whatever the species there happens to be someone on this site who knows it? Or is there some kind of system for IDing even species you haven't heard of before?

Answer 1

Some thoughts....
In general, having good overall knowledge of groups (orders/families etc) and their characteristics can quickly narrow down the answer set, which makes searching for traits more efficient and practical. Therefore, having a good overall background in e.g. entomology (or botany etc...) can allow you to relatively quickly arrive at the correct order, family genera etc more easily. Without more knowledge, dichotomous identification keys are often relatively useless, since they are often useful only in specific regions, or can only determine individuals to e.g. the level of order.

However, we definitely also get overconfident answers, where the answerer underestimates the difficulty of getting correct ids to the species level only from one or two pictures (I may also be guilty of this). There are often sets of species that require deep knowledge and day to day familiarity with the species (i.e. looking at them regularly) to id them correctly. Also, the information you find on the internet is scattered (at best). Just because the overall appearance fits with one particular species present on the internet (with a couple of pictures, some basic info etc) doesn't exclude that there might be several other extremely similar species, only known to specialists on that particular taxon. Even so, providing a suggested ID isn't a problem, as long as it is done in a humble way, acknowledging that there might be species unknown to you and clearly stating your level of expertise. In that way, the OP can interpret the reliability of the answer in a better way.

So to your bullet points:

• Yes, but together with e.g. Megachilidae found in Germany you might have more luck.

• I agree, in the sense that they are either only useful for very broad classifications (e.g. orders), or very narrowly (Longhorn beetles in Scandinavia). Also, most really good determination keys (of the latter kind) are found in specialist books that are not found openly on the internet.

• There exist some good ones, but, again, they are usually only useful on specific regions or countries. I haven't used that many species-ID apps though.