Perga dorsalis vs Perga affinis
Workbook
My goal here is to assess whether it is possible to separate the two similar sawfly species Perga dorsalis and Perga affinis, using field photographs of the whole insect.
I have used the catalogue of Schmidt and Smith (2006) to gain information from the taxonomic literature about morphological differences between these species.
Summary
The most authoritative source for separation of Perga dorsalis and Perga affinis is the Perga key in Benson 1939. The morphological characters in the final step of this key (e.g. measurements of eyes/head and tarsi) are only rarely recorded at the necessary level of detail in field photographs, severely limiting its use for this separation.
The claim by Riek 1961 that his key and descriptions enable a reliable separation of these species is questionable. His key uses differences in morphological characters that can be seen in many field photos. However he very likely underestimates the variability in these features even among specimens at the same location. So his key can probably be used with confidence for just a single morph and even then in only a few locations.
I conclude that it will only rarely be possible to separate Perga dorsalis and Perga affinis in iNaturalist observations.
First descriptions
Perga dorsalis and Perga affinis have a long taxonomic history.
Perga dorsalis Leach 1817 is the type species for the genus Perga. Leach provided a colour drawing of the male of this species (table 148, fig. 1) and a brief description (p. 117) of the colour of various body parts.
Perga affinis was first described by Kirby in 1882, again from a male specimen. His paper shows coloured drawings of both the male and female of this species, as well as Perga dorsalis (plate I). I have provided copies of these drawings from his paper in the panel below.
Kirby provided a much more detailed description of Perga affinis (pp. 19-20) than Leach did for Perga dorsalis and also redescribed the latter species (pp. 18-19). However, typical of the approach in taxonomic papers of that time, he confined himself to a description of the colour of various body parts.
The drawings and descriptions suggest that these two species are rather similar. So how can they be separated?
Kirby did not specifically comment on differences between these two Perga species, while he did so for many of the other Perga species he dealt with.
Drawings of male and female Perga affinis and Perga dorsalis from Kirby (1882)
Carne (1962) provides a brief description and drawings of male and female of the Perga affinis subspecies Perga affinis affinis (see Riek’s, 1961 work below for discussion of subspecies).
He describes the adult of P. affinis affinis as “soft-bodied wasps with pale brown, hyaline wings and a shining greenish black body. The labrum, clypeus, antennae, pronotal lobes, scutellum, and legs are light brownish to reddish yellow. Males are smaller than females and comparatively rare”.
He states “there is a general tendency for the insect to be of smaller size towards its drier inland limit, and a corresponding tendency for the pigmentation of the adult female mesepimera to vary from black to pale yellow. Black forms are found exclusively in the Australian Capital Territory; the amount of yellow coloration of the mesepimera increases in samples taken toward the 21-in. isohyet. However, no distinct cline can be demonstrated, both fully black and fully yellow forms coexisting in some localities.”
He also notes that P. dorsalis dorsalis is occasionally found in the ACT. However he states that there is no difficulty in separating the species where they coexist (in the ACT), because P. dorsalis dorsalis “occurs as a final-instar larva when affinis is in the adult or egg stage (February to May with median emergence times of P. affinis between March and April)”.
Carne also provides descriptions and drawings of the larval stages, prepupa and pupa of P. affinis.
Fig. 1. from Carne (1962)
Tait (1962) provides a detailed account of the external and internal anatomy of the adult - both male and female - and larva of Perga affinis affinis. But no mention of Perga dorsalis is made in Tait’s paper.
Figs. 2A, B from Tait (1962)
Comparisons of the two species
Morice
While both of these Perga species were dealt with in several subsequent papers (see Citations on p.68 of Schmidt and Smith, 2006), Morice (1918) appears to be the first to attempt to separate the pair.
He states (pp. 266-267) that “I do not believe that (affinis) is more than an aberration of the next species (dorsalis). It agrees with it exactly in all structural characters, details of "saws", etc . Nor can it be considered as a "subspecies =(local race) since both forms occur in the same locality.”
Benson
Building on Morice’s work, Benson (1939) carried out a major revision of the large genus Perga, splitting it into 8 different genera - Cerealces, Xyloperga, Acanthoperga, Pseudoperga, Paraperga, Antiperga, Pergagrapta and Perga.
He provides a key for the species in his redefined Perga genus (pp.334-338), which includes both P. affinis and P. dorsalis. He points out that the type specimens of those two species are males and that they do not show the differences that are evident in the females of these two species. He also claims that “it is by no means altogether certain that the females have been correctly associated with the males” in the Kirby 1882 study.
Benson separates females of the two species using the features listed below. Incidentally, the most recent catalogue of the Pergidae (Schmidt & Smith, 2006) cites the Benson key for separation of these species.
Perga dorsalis
hind basitarsus about as long as the following three tarsal segments together; the next to last tarsal segment is clearly longer than broad
hind ocelli closer together so that POL is less than OOL, and the postocellar region is longer than its breadth in front, so that the POL is only about half as long as the distance between the hind ocellus and the hind margin of the head (fig. 26 below from Benson, 1939)
wings yellowish-brown infuscate
sawsheath with extreme apices of the valves diverging when viewed from beneath (fig. 29 below); sides of the sheath bearing bristles that are almost racket-shaped (fig. 30 below)
Perga affinis
hind basitarsus greater than the following three following tarsal segments together and almost as long as the four following tarsal segments
hind ocelli further apart so that POL is greater than OOL; postocellar region is shorter than its breadth in front, so that POL is as much as 2/3 the distance between the hind ocellus and the hind margin of the head
wings yellowish
sawsheath with extreme apices of the valves converging (fig. 28); bristles on the sides of the sheath less swollen (fig. 30). A lateral view of a portion of the saw is shown in fig. 39.
The following diagram (modified from fig. 3 of Carne, 1962) shows a ventral view of the ovipositor and saw sheath of Perga affinis.
The red rectangle shows the region in Benson’s fig. 28, while the red arrow indicates an individual bristle, as represented in Benson’s fig. 30.
While Benson’s 1939 key enables the separation of P. affinis and P. dorsalis, it can only be applied with a specimen in the hand. High magnification images would be required to resolve the structures involved.
Only rarely would photographs taken in the field reveal the required detail of the relevant structures.
A reliable method for separation of P. affinis and P. dorsalis?
Riek (1961) claims that Perga affinis and Perga dorsalis can be reliably separated using a suite of characters.
The first step in his key (p. 238) uses the morphology of the ovipositor and sawsheath to separate P. affinis and P. dorsalis.
Perga affinis
“Hairs of ovipositor valves very dense and fine, almost touching one another, tips of hairs not spooned; flattened “saw-bench” on lower margin of saw-sheath very long and thin, without obvious longitudinal striae.”
Perga dorsalis
“Hairs of ovipositor valves not dense, with a considerable space between each hair, tips of hairs spooned; flattened “saw-bench” on lower margin of saw-sheath relatively short and broad, with 7 or 8 obvious longitudinal striae.”
The differences in the morphology of the ovipositor hairs (arrowed) are evident in images of the holotypes of Perga affinis insularis and Perga dorsalis nitida (provided by National Museum Victoria).
This diagram of the terminal abdominal segments of a female Perga affinis affinis shows the fine, thread-like hairs of the ovipositor (va.3), which are spaced finely.
Fig. 6A from Tait (1962)
However as noted above, these differences would rarely be apparent in photos of whole insects taken in the field.
Separation of P. affinis and P. dorsalis subspecies.
Riek’s key then separates three subspecies of both P. affinis (P. affinis affinis, P. affinis atrata, P. affinis insularis) and P. dorsalis (P. dorsalis dorsalis, P. dorsalis nitida, P. dorsalis castanea), which he has newly erected.
Images of two of these subspecies - the holotype of P. dorsalis nitida and P. affinis insularis housed in the National Museum of Victoria - are shown below. (Images downloaded from ozcam.ala.org.au - P. dorsalis nitida here and P. affinis insularis here)
Riek’s key uses features such as colouration of legs and other body parts, but also density of hairs on the gena (cheeks). These features could be seen in many field photos of the whole insect, making this a potentially valuable tool for species separation of iNaturalist observations.
However, the most recent catalogue of Pergidae (Schmidt & Smith, 2006) does not treat Riek’s taxa as distinct subspecies. Schmidt & Smith point out that the three P. dorsalis subspecies overlap in Victoria and that the character used to separate the P. affinis subspecies (legs dark vs. pale) is subject to variation, even among specimens from the same locality.
Indeed, Riek himself states that “one female Perga affinis from Hazlewood, Victoria, has the base, apex and caudal margin of the hind femur darkened but not black. In the other two females examined from this locality the hind femur is entirely pale”. So It is likely that his subspecies just represent normal variation of each species over its range.
Nonetheless, can Riek’s key be used to separate P. affinis and P. dorsalis at the species level by using just the “typical forms” of the subspecies (those with the same name as the nominate species)?
Perga dorsalis dorsalis - Riek states that this subspecies occurs in coastal New South Wales, extending south into Victoria and north possibly into Queensland. (The species extends inland to Canberra, but is not common there and according to Carne (1962), adults are not present at the same time of the year as P. affinis in that location. However this iNaturalist observation of Perga dorsalis in Canberra on February 27th, 2024 might bring this claim into questions. It also shows that the dark colouration of the hind femur in this species is quite subtle).
Female: Coxae black in part; hind femur partly dark; trochanters pale; abdomen with green-black iridescence; mesepimeron with the punctate area pale in part; lower gena with dense short hairs; mesepimeron with a pale spot.
Male: Similar to the female, but legs, except coxae, all pale; coxae mostly dark as in the female; lower gena with dense short hairs; first segment of abdomen generally only narrowly pale at apex or all dark.
Perga affinis affinis - Riek states that this subspecies ranges from the central southern coast of Victoria north through inland New South Wales to southern Queensland and west to the Adelaide region in South Australia.
Female: Legs, except coxae, all pale; all coxae black in part; mesepimeron with the punctate area often pale in part but sometimes all dark; abdomen with a green-black iridescence; lower gena with only scattered fine hairs, punctures large and usually spaced.
Male. Similar to the female, with only scattered hairs on the lower gena; coxae mostly pale, otherwise coloured as in the female. The coxae are mostly dark in the Canberra form.
This combination of features described for the Canberra form of P. affinis separates it from any of the forms of P. dorsalis described by Riek. So, it may be possible to identify this P. affinis form with some confidence in Canberra.
For other locations, Riek lists the following differences between the typical forms of these species.
P. dorsalis dorsalis - the hind femur is partly dark; the lower gena has dense short hairs.
P. affinis affinis - the hind femur is pale; the lower gena has fine, scattered hairs.
As stated above, leg colour in P. affinis is subject to variation even among specimens from the same locality. Indeed this is evident in iNaturalist observations at individual locations where only one form of the species is supposed to occur. So, separating these species at locations other than Canberra using this character would appear to be problematic.
Separation of P. dorsalis and P. affinis based on location?
Given Riek’s claim that Perga dorsalis is “more of a coastal species”, is it justifiable to identify an insect that keys out to Perga dorsalis/affinis as Perga dorsalis if it is found at a coastal location?
Riek has not defined the geographical limits of “coastal” or presented data to demonstrate that specimens close to the coast show a different set of characters to those away from the coast.
In most other locations, Riek states that both species can occur.
So it would appear to be unjustified to separate P. affinis and P. dorsalis on the basis of location alone.
This is supported by an analysis of the geographical distribution of Perga affinis and Perga dorsalis adults on iNaturalist. Observations were selected in which photographs of the ovipositor valves were of a sufficiently high quality to visualise the hairs on the valves. The morphology of these hairs (see “A reliable method for separation of P. affinis and P. dorsalis?” above) was used to separate these species.
The left hand image below shows the distribution of Perga dorsalis observations identified in this way. The observations can be accessed at this link.
The right hand image shows the distribution of Perga affinis. These observations can be accessed at this link.
This data set does not support the claim that Perga dorsalis is “more of a coastal species”.
Separation of Perga affinis and Perga dorsalis by DNA barcoding
DNA barcoding holds great promise for resolving the phylogenetic relationships between sawfly taxa and this approach has been used in a study by Schmidt and Walter (2014). The authors used three gene regions for their analysis: (1) the 5’ end of the mitochondrial cytochrome c oxidase 1 gene (COI), (2) a fragment of the mitochondrial 16S rDNA gene, and (3) the D2 loop of the nuclear 28S rDNA gene.
An adult sawfly identified as Perga affinis was DNA barcoded in this study. This is an image of this specimen.
Perga dorsalis was not included in this or a later molecular study (Schmidt et al, 2017), so no conclusions can yet be drawn about how successfully barcoding could separate the species.
Image of specimen: PERGI043-12downloaded from BOLD Australia
References:
Benson, R.B. 1939. “A revision of the Australian sawflies of the genus Perga Leach, sens. lat. (Hymenoptera, Symphyta)”. The Australian Zoologist 9: 324-357. “ Biodiversity Heritage Library link.
Carne, P.B. 1962. “The characteristics and behaviour of the saw-fly Perga affinis affinis (Hymenoptera)”. Australian Journal of Zoology 10: 1-34.
Kirby, W.F. 1882. “List of Hymenoptera, with descriptions and figures of the typical specimens in the British Museum”. Vol. I. Tenthredinidae and Siricidae. London: British Museum, xxviii. Biodiversity Heritage Library link.
Leach, W.E. 1817. “The zoological miscellany; being descriptions of new, or interesting animals”. Vol. 3. Shoe-Lane, London: R. and A. Taylor, 151 pp. Biodiversity Heritage Library link.
Morice, F.D. 1918. “Notes on Australian sawflies, especially the "Authors' Types" and other specimens in the British Museum of Natural History and the Hope Collections of the Oxford University Museum; with diagnostic Synopses of the Genera and Species, and photographs illustrating their structural characters.” Transactions of the Entomological Society of London. Biodiversity Heritage Library link.
Riek, E. F. 1961. “The distribution and inter-relationships of Perga affinis Kirby and Perga dorsalis Leach (Hymenoptera, Symphyta)”. Proc. Linnean Soc. NSW 86, 237-240. Biodiversity Heritage Library link.
Schmidt, S. & Smith, D.R. 2006. “An annotated systematic world catalogue of the Pergidae (Hymenoptera)”. Contributions of the American Entomological Institute. 34, no 3. link to this catalogue from Pergidae of the World website
Schmidt, S. & Walter, G.H. 2014 “Young clades in an old family: Major evolutionary transitions and diversification of the eucalypt-feeding pergid sawflies in Australia (Insecta, Hymenoptera, Pergidae)” Molecular Phylogenetics and Evolution 74: 111-121.
Schmidt, S. et al. 2017 “Identification of sawflies and horntails (Hymenoptera, ‘Symphyta’) through DNA barcodes: successes and caveats”. Molecular Ecology Resources 17, 670–685.
Tait, N.N. 1962. “The anatomy of the sawfly Perga affinis affinis Kirby (Hymenoptera: Symphyta)” Australian Journal of Zoology 10: 652-683.
This is a workbook page … a part of our website where we record the observations and references used in making species identifications. The notes will not necessarily be complete. They are a record for our own use, but we are happy to share this information with others.