You can use this as a general guide for how to describe new species, except you have to look out for the relevant taxonomic characters (but if you’re describing a new species, I guess you already knew that). Think of this as some form of template.
In the introduction, always start out by mentioning the genus and citing the original description – you have no idea how annoying it is to read taxonomic papers that throw names around without citing them. And simply naming the author and date isn’t enough, make sure you include the full citation in the reference list (it often occurs that the name and date is given in the text, but not cited fully at the end)! Anyway, a summary of the diversity, biogeography and habitat preferences is standard for the first paragraph. A historical overview of research into the genus/species group and biogeographical area is always welcome, to allow further reading. Think of it as common courtesy :)
In the materials and methods, you list several things: specimen preparation (if applicable), description method (i.e. stereomicroscope + drawing) and specimen sources (depending on the journal, you can also make a table with specimen IDs, etc.). The most important thing in this section is to cite the sources for your terminology. While most of the time there are standard reference works that are cited by everyone, sometimes papers published after those works provide updates (this is the case with the terminology of fly antennae, for example). If you’re going to deviate from any published works, always provide a diagram detailing what you mean (often the case for measurements).
Next comes the actual taxonomic part. Name the family and genus hierarchically (with original description citation). If redescribing a species, make sure to list the original description and past synonymies (and their citations as well. Seriously. Do not skimp on original descriptions). When describing a new species, do not forget to say that it’s a sp. n. or sp. nov. or spec. nov. (or some variation on that theme; same applies for new genera (gen. nov.), families (fam. nov.), etc).
For every species, there are several subsections. First, list the type material’s location and what is written on the type’s label (it’s an unwritten rule that labels are standardised to include precise location, ecological data, date of collection and collector). List any other material of that species you examined and the sexes.
Then comes the diagnosis, i.e. the precise and concise morphological description of the characters that allow us to differentiate the species. Often, there are no complete sentences here. This is essentially a list, but not in bullet form. It saves space :) Preferably, this is supplemented by at least one diagram. Note that diagrams are always preferrable: you learn a lot drawing them, since your eye is focussed on the specimen, and the amount of information is much greater than from a photograph; note that this is changing nowadays with the advent of high-quality photography, but if you must use photographs, make sure the lighting is not deceptive and allows us to see structures clearly, also making sure that the entire picture is sharp (there are programs that can combine pictures at different DOPs for this purpose). As a sidenote, if you’re writing a fieldguide, my unscientific polling shows that photos are preferred to diagrams, but that’s a different matter.
As you all know, species have a binomial name, the first being the genus, the second being the species. These are not chosen at random, and the etymology has to be explained (although the reasons can be silly. Some name them after their children, I’ve even seen a dinosaur named after Hogwart’s.) Also, these have to be declined properly in Latin, paying attention to gender.
In the description section, you elaborate on the diagnosis and provide a full description of the animal, providing information on variability or colouration. Finally, a note on ecology is always included; if these are specimens from a decades-old museum collection, you can still elaborate, as collectors always keep archived notes on their specimens.
While not necessary, a key is a very nice finishing.
In the Acknowledgements section, besides listing grants, etc. as normal, do not forget to thank the museums, institutions and private collectors (as applicable).
A special addendum for fossil species: Another section must contain a description of the geological locality, with at least the stratigraphy and type of rock and, if available, a date. A citation to a geological overview of the locality is expected, but a summary of what is currently known in the actual paper saves the reader much hassle.