Notes on the retractabilty of gill tufts in Pseudolestes mirabilis (Zygoptera: Pseudolestidae)

Pseudolestes mirabilis Kirby (1900), the only member of its genus, endemic to Hainan, continues to provoke speculation on its phylogenetic position. Based on its unique adult characters the species was originally placed in its own subfamily by Kirby (1900), and then in subfamily Lestinae (now Lestidae) by Calvert (1902) and Tillyard and Fraser (1938–1940). Fraser (1957) established the family Pseudolestidae to accommodate Pseudolestes and several other genera, an arrangement followed by Davies and Tobin (1984) and van Tol (2006). According to recent molecular studies it either nested in “Megapodagrionidae” (Bybee, Ogden, Branham, & Whiting, 2008) or deserves its own family Pseudolestidae (Dijkstra et al., 2014). Nevertheless some researchers have placed emphasis on larva characters. Kalkman, Choong, Orr, and Schütte (2010) divided megapodagrionid larvae into four groups and supposed that the larva of P. mirabilis would fall into group 1, i.e. having sack-like gills. The first detailed description of the larva of P. mirabilis was provide by Yu and Bu (2011) in which the sack-like caudal gills were confirmed. However this study also revealed the presence of a remarkable character, namely the gill tufts, which might implies that it may be close to Amphipterygidae, in the restricted sense defined by Novelo-Gutiérrez (1995) and Rehn (2003). It is in this traditional sense we will use the term amphipterigid henceforth. Unfortunately, Yu and Bu (2011) did not check if the gill tufts of P. mirabilis larva are retractable. Larvae of amphipterygids possess paired gill tufts (Corbet 1999; Watson 1966) which are retractable in life (Orr, 2008). Larvae of P. mirabilis also possess paired gill tufts which strongly resemble those of amphipterygids. If closer examination should prove them to be essentially the same as the structures known from the amphipterygids group (retractable), it would be strong evidence for a