patawa-malaise-2 (c) Jean Cerda, 2009

The project

In 2000 I brought back a small sample of tachinids from French Guiana and was given a larger sample collected by a group of entomologists living in the country. The 150 or so specimens sat in a box until 2008 when an idea dawned on me which would allow me to do meaningful work on them without the inherrant problems associated with identifying them.

The aims of the project are:

  1. To work out a rough number of species for a small region of French Guiana.
  2. To identify taxa as accurately as is possible.
  3. To describe any undescribed species.
  4. To broaden my knowledge of neotropical tachinid morphology and source all available literature and keys for the region. Ulitimately it would be nice to think that I could create a key for the region but this is probably too much to ask for and quite presumptious for me to think that I have enough knowledge to do this.

The project involves dividing the specimens into morphologically distinct taxon-groups, based on features that are easy to see (see below). These taxon-groups sometimes resolve to individuals that are assumed to belong to one species but others just group similar taxa together. In the future the latter will be sub-divided into more and more specific groups. Taxon-groups and individual specimens are assigned unique numbers and these are all databased in an Excel spreadsheet.

How do you sort taxa without keys?

This was my main problem but really it isn’t as hard as it sounds – really – it is just very time consumingand involves a lot of thought and invention.

Most palearctic taxa are keyed using external morphology that is fairly easy to find on a fly – we almost never use internal or larval features and it is possible to identify everything using a normal <50x binocular disecting microscope so there isn’t even a need to make slides. The numbers of bristles and their positions (chaetotaxy) are a common feature, as is the hairiness of antennae or the face, or the various patches of dusting. Male genitalia are also commonly used to confirm species-level identifications.

I started by making a list of over 70 features that I could usually find on any tachinid and putting those as column-headers in a spreadsheet. I then just went through the specimens pulling out fairly obvious morphotypes (one similar group at a time) and noting down the 70 features for each one. This group would get a unique taxon number and each specimen gets a unique specimen number. Any significant variations in the answers to the 70 features leads you to sub-group those specimens and often quite similar flies actually resolve into 2 or more actual morphotypes.

Then once you have done the obvious ones (maybe 10-20%) you are left with the more anonymous specimens … usually the black/brown & grey-dusted ones where it would be nearly impossible to pull out similar specimens by eye. For the next stage I have decided to make the job easier by sub-dividing them using very easy/reliable physical features such as:

  • Eyes hairy or bare?
  • Arista bare, hairy or plumose?
  • Hairs on vein r4+5 extending over more than 1/3 of the way to rm?
  • Hairs on r1?
  • Hairs on cu?
  • Bristles over more than 1/3 of the facial ridge?
  • Parafacial bristles decending at least half way down?
  • Median vein with a strong petiole?
  • Smaller than normal antennal segment 3?
  • Presence of strong median discals on tergite 3?
  • Presence of a Sturmia-spot? (possible Sturmiini)
  • Presence of a strongly-angled m-cu (possible Voriini)

Some features wouldn’t be used because they are either:

  • Variable, such as body colouration or
  • Easily destroyed, such as the orientation of the apical scutellar bristles or
  • Often obscured, such as the hairiness of the prosternum or
  • Sex-specific, such as the presence of parafrontal bristles.

The resulting smaller samples can be further scanned visually for similarities and subdivided using finer and finer criteria until everything has been allocated a taxon. I check for any existing taxa that might fit the bill by filtering in the spreadsheet on known criteria and then examining the specimens.

The problems with chaetotaxy

All this sounds great but it has been pointed out, by cleverer people than I, that traditional tachinid identification has relied very heavily (possibly too heavily) on chaetotaxy and that there is a finite limit to how many different permutations are possible before you get 2 different taxa with identical chaetotaxy. I myself have struggled over the neotropical tachinids bacause 2 quite obviously different tachinids can actually have very similar or identical chaetotaxy … but they look different to a human’s eye.

So, we need to be flexible when picking out criteria and be prepared to use unconventional features to help us achieve a determination. Male genitalia will almost certainly prove to be an essential (if slightly inconvenient) feature so this raises the importance of preparing/exposing male genital capsules while they are being pinned. Diagrams or photo-montage images should be prepared of all male genitalia as a matter of course when one gets down to the finer sorting – especially of the black/brown and grey-dusted taxa.

This also means that for some of the more anonymous-looking taxa it will be practically impossible to tackle females and serious workers should concentrate on male specimens to avoid the early onset of insanity!!

I would caution against the use of features that are variable (colour) or which could be distorted by the drying of specimens taken from alcohol (head shape etc).


The project relies on acquiring a very large sample of specimens (2-3,000 or more). This volume is necessary for a number of reasons:

  1. The larger the sample, the more complete the collection will be and I am trying to amass as many taxa as possible.
  2. The more specimens of each taxon you can obtain, the clearer will be the degree of intra-taxon variation and the resulting taxa will be better-defined.

This creates its own problems though – namely just handling and curating that quantity of specimens. So I have acquired (through the generosity of the Oxford University Museum) a number of secondhand museum drawers and unit-trays to help me house them during the project.

The glass tops make it easy to locate the correct drawer when searching for specimens and I actually find myself just peering at the specimens through the glass – it inspires me on a daily basis to think about the project, which I don’t think would happen if the specimens were hidden away in store boxes. The unit-trays make it very easy to arrange/curate taxa and trays of unsorted material, while at the same time it is easier and safer to examine them under the microscope with the minimum of pin handling.

Most large museums are switching over to the unit-tray system too, due to the relative ease of curation – reorganising taxa to allow for new material is as easy as shuffling the boxes around and inserting the new tray in the correct position – no need to move each specimen by hand.


In the process of sorting through the samples I take out various groups of insects and donate them to other workers in those fields:

Theo Zeegers (NL): Tabanidae
Eric Fisher (USA): Asilidae
Menno Reemer (NL): Syrphidae
Martin Hauser (USA): Stratiomyidae
John Smit (NL): Tephrititoids
David Gibbs (UK): Bombylidae & Pipunculidae
Steve Gaimari (USA): Lauxanidae

I keep all Tachinidae, Panthophthalmidae, Ropalomeridae, Hymenoptera & anything that looks “interesting”.

Thoughts so far…

Diversity in neotropical tachinids

It is blindingly obvious to anyone who has worked on neotropical insects that the region contains a vast number of species. Most of these have not been named and those that have been named were worked on many years ago by authors based all around the world (but mainly North America) so the types are often held in widely disparate locations. Tachinids are no exception and in fact are much less studied than groups like the Lepidoptera and Coleoptera.

My initial thoughts are that there are a large number of species from different tribes (and possibly even sub-familes) showing very similar, possibly convergent colouration. This makes it very difficult to spot different species until you get down to very small keying features.

Variability seems to be a major problem too and in many taxon groups I feel that I have located one species but finding stable features will be a problem and any key is likely to include words such as “usually with…” and “often has…”!! This is why such a large number of specimens is necessary and hopefully, after considerable analysis, I will be able to work out where variability finishes and speciation starts.

Differences between the Palearctic & Neotropical, so far…

Obviously I have only examined a relatively small sample from one geographic location but already some interesting patterns have emerged:

  • Palps: Nearly every specimen examined so far has orange palps, which is very unusual.
  • Petiole: Petiolate median veins are a very rare feature. In the Palearctic we might expect them to pop up with fairly high frequency (particularly in Phasinae, Wagneriini and Dexinae).
  • Phasiines appear to be less common than would be experienced in a sample of Palearctic tachinids and their morphology seems centered around a few distinct types.
  • Dexiines on the other hand seem to be much more frequent in the neotropical sample than would be found in the Palearctic.

When I started the project I was expecting to find a far greater level of diversity in the morphotypes than I have actually found. One always imagines that in the tropics you will find specimens with vivid colours or extreme morphology but in reality the vast majority look very ‘normal’. Of course there are many unusual taxa with unusual and exageraged features but in general the huge number of taxa in the neotropics seem to focus on a few general morphotypes.

Index to taxon articles

One Reply to “Neotropical”

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.