Aedia, Hubner, 1823

Keegan, Kevin L., Rota, Jadranka, Zahiri, Reza, Zilli, Alberto, Wahlberg, Niklas, Schmidt, B. Christian, Lafontaine, J. Donald, Goldstein, Paul Z. & Wagner, David L., 2021, Toward a Stable Global Noctuidae (Lepidoptera) Taxonomy, Insect Systematics and Diversity (AIFB) 5 (3), pp. 1-24 : 18

Aedia , Copibryophila , and Tyta

Aedia Hübner, [1823] ; Copibryophila Smith, 1900 ; and Tyta formed a poorly supported clade (BS = 21, SH = 83.3, UF = 80) apart from any recognized subfamily, with long branches leading to each genus ( Fig. 5). Aedia has long been recognized as peculiar in Noctuidae , having been placed into its own subfamily by Beck (1960) where it remains today. Copibryophila is currently recognized in Amphipyrinae , and Tyta in Metoponiinae , and both are well supported as being excluded from their respective subfamilies. Our gene coverage for these three genera does not stand out as exceptionally poor and thus we do not believe they have been erroneously placed due to a lack of gene coverage. It is possible that their clustering together is an artifact of long-branch attraction which, although primarily a concern when inferring phylogenies using parsimony ( Felsenstein 1978), is not without effect on maximum-likelihood phylogenetic inference ( Kück et al. 2012). We performed three additional RAxML phylogenetic inferences with the same settings and data as detailed earlier, but with each of these three genera alone and the other two removed from our alignment (results not shown). With Copibryophila and Tyta removed, Aedia grouped in the poorly supported clade of Pseudeustrotiini ? + Condicinae . With Aedia and Tyta removed, Copibryophila grouped (poorly supported) in the same position as shown in Fig. 5. With Aedia and Copibryophila removed, Tyta grouped (poorly supported) in the same position as shown in Fig. 5. The overall resultant tree topology of these three additional analyses did not conflict significantly (with respect to support values) with the tree topology as shown in Figs. 3–6 . We have no reason to believe these genera are closely related to each other or necessarily deserving of recognition in multiple distinct subfamilies. All are worthy of more study.

Figures

Fig. 3. Maximum-likelihood tree inferred by RAxML analysis. Clade labels reflect our hypotheses about the extent of the various clades, and the taxonomic changes suggested in Results and Discussion. Lineages shown as being members of various clades may not have as yet been transferred into them. Bootstrap values ≥ 70 are shown at nodes.The scale bar shows estimated number of substitutions per site.The tree continues into Figs. 4–6. Clades are shown in three alternating colors to aid visibility.

Fig. 4. Continuation of maximum-likelihood tree inferred by RAxML analysis.