<strong> Revision of the genus <em> Erythromelana </ em> Townsend (Diptera: Tachinidae) and analysis of its phylogeny and diversification </ strong> Author Inclan, Diego J. diegojavier.inclanluna@studenti.unipd.it Author Stireman, John O. john.stireman@wright.edu text Zootaxa 2013 2013-03-11 3621 1 1 82 http://dx.doi.org/10.11646/zootaxa.3621.1.1 journal article 10.11646/zootaxa.3621.1.1 1175-5326 5260220 99CC048E-C555-4178-8932-70E93A3B255C PCA of Erythromelana Mapping Erythromelana s.s. specimens onto the first two PCA axes, based on 45 morphological characters, accounted for the 18.5 and 11.6% of the variance respectively. However, these first two axes allow us to illustrate a relatively clear division of two main groups ( Fig. 109 ), although there is slight overlap in the 95% confidence intervals. Separation of specimens along the first axis (PC1) is due primarily to the number of setae on various sclerites, particularly the number of supra-alar setae on the postsutural scutum, the presence of abdominal discal setae, the number of katepisternal setae, the number of discal scutellar setae, and the number of acrostichal setae on the postsutural scutum ( Table 2 ). Separation of specimens along the second axis (PC2) was mainly the result of contributions from the number of frontal setae, the ratio of the flagellum length to the head height, abdominal coloration, tibia coloration, and the presence of ocellar setae ( Table 2 ). TABLE 2. Factor loadings for the five most important characters on the first two axes of the PCA of 169 specimens based on the analysis of 45 morphological characters of Erythromelana species.

Variable	PCA loading Axis 1	Variable	PCA loading Axis 2
number of supra-alar setae on the postsutural scutum	0.3	number of frontal setae	0.28
presence of abdominal discal setae	0.28	flagellum length to the head height	0.27
number of katepisternal setae	0.26	abdominal coloration	- 0.26
number of discal scutellar setae	0.23	tibia coloration	- 0.22
number of acrostichal setae on the postsutural scutum	- 0.23	presence of ocellar setae	- 0.22

FIGURE 109. Principal components ordination of 169 Erythromelana specimens based on the analysis of 45 morphological characters. Ovals indicate 95% confidence limits. Specimens morphologically similar to E. jaena are grouped towards the upper-left quadrant of the ordination plot, and specimens similar to E. cryptica are clustered along a diagonal from lower-left to upper-right portions of the ordination space. This suggests that Erythromelana species can be separated into two main species groups, which are referred here as the E. jaena and E. cryptica species groups (see species description section). In general, most of the species in the E. jaena group have a bright yellow abdomen (except E. leptoforceps and E. nigrithorax ), 2–3 katepisternal setae (usually 2), and the first postsutural supra-alar absent. In contrast, most of the species in the E. cryptica group have the abdomen mostly black with yellow laterally, usually 3 katepisternal setae, and first postsutural supra-alar usually present. Individual species do not form clear, identifiable clusters in the PCA ordination. One reason for this overlap is that this PCA only includes non-terminalic structures and within the species groups most Erythromelana can only be separated by differences in male terminalia. Similar results have been found in other Diptera , such as Pseudexechia Tuomikoski (Mycetophilidae) flies, where a PCA of 59 nonterminalic characters showed broad overlap between species in the ordination plot and only species groups were identified ( Kjaerandsen 2009 ). Morphologically cryptic species have been commonly found across Tachinidae (e.g., Belvosia Robineau-Desvoidy, Smith et al. 2006 ; Winthemia Robineau-Desvoidy , Anoxynops , Lespesia Robineau- Desvoidy, Smith et al . 2007 ) and many other insect groups (e.g., butterflies, Burns et al . 2008 ; beetles, Monaghan et al . 2005 ; Wasps, Molbo et al . 2003 ), where separation of species has required molecular techniques and/or terminalic characters. The difficulty in separating species of Erythromelana based on external morphology reinforces the usefulness of incorporating terminalic characters and genetic data. PHYLOGENETIC RELATIONSHIPS OF ERYTHROMELANA SPECIES