With so many unidentified Sawfly, Horntail and Wood wasp larvae it has become necessary to group them together.
"For some large genera such as Tenthredo and Macrophya, almost nothing is known. It is easy to run across and photograph a nice looking larva, but ID is a different thing. Then, the photos don't show characters I need to see to place them to subfamily, genus, or even sometimes family. I need to see some or all of the following: number of prolegs, number of annulets on the abdominal segments, antennae, mandibles, and if with tubercles or spines on the body, which annulets they are on, which ones are or are not forked, what they look like on the 10th tergum, etc. As many people say in the remarks, the host plant is also important. Early and late feeding stages are commonly different, and the last stage (prepupal or non-feeding stage) is sometimes very different from the feeding stages. Also, the brilliant color of all the specimens in the photos is lost in the alcohol specimens, so preserved specimens sometimes look a little different.
"Larvae of 20% or fewer N. American species are known. Most of these are economic species or species that can be easily reared. For most, there is a one year life cycle, and very difficult to rear. [For rearing, I commonly put sand/soil, corks, corrugated cardboard into jars because I never know what the larvae prefer for pupation sites.] If one can get larvae to pupate, they are put aside and hopefully they will survive a year an adult or two will emerge the following year. Maybe now with DNA work, adults and larvae can be associated easier, without rearing, matching the DNA of adults and larvae. A lot to be done - and more than I can do.
"One character missed in remarks by people separating sawfly larvae from leps is that sawfly larvae do not have crochets on the prolegs whereas lep larvae do. Usually only the number of ocelli and number of prolegs are mentioned." (Comment by Dr. David Smith)