Node Calibrated: (4) Crown Anisoptera. This node represents crown Anisoptera (dragonflies).

Species. Sinacymatophlebia mongolica (Cymatophlebiidae) Nel and Huang, 2009

PALAEO- ELECTRONICA.ORG

Specimen. NIGP 148312 (Nanjing Institute of Geology and Palaeontology, China): hind wing of male, holotype of Sinacymatophlebia mongolica.

Phylogenetic Justification. The phylogenetic relationship of this fossil with Aeshnidae and hence Anisoptera is strongly supported by many wing venational synapomorphies (Bechly et al., 2001; Nel and Huang, 2009). These include: the presence of Rsp1 vein, RP1 and RP2 basally parallel up to the pterostigma, and RP3/4 and MA undulating.

Minimum Age. 168 Ma.

Age Justification. Found in Middle Jurassic Jiulongshan Formation (= Haifanggou Formation) near Daohugou village, Wuhua township, Ningcheng county, Chifeng city, Inner Mongolia, north-east China. The Daohugou-Biota, resting on the Jiulongshan Formation, has been dated with SHRIMP U-Pb zircon age of 168-152 Ma, giving a minimum age for Jiulongshan Formation of 168 Ma (Liu et al., 2006). There is debate about the age of the Jiulongshan Formation, as radioisotopic studies have not been congruent (Liu et al., 2012). However, because the Daohugou Formation overlies the Jiulonshan formation, we can conservatively use the age range of the former to provide a minimum age as 168 Ma for the calibrating fossil.

Discussion. This fossil calibrates the crown group node Anisoptera (Figure 3), which we are considering to include modern families of Anisoptera but not the Liassic stem group families Henrotayiidae and Liassogomphidae (Fleck et al., 2003; Davis et al., 2011). This fossil also calibrates the crown group node for Aeshnoidea (Austropetaliidae + Aeshnidae), a well-supported clade called Aeshnoptera by Bechly et al. (2001). This has not yet been used in molecular estimates. Ware et al. (2007) and Ware (2008) set the maximum age of Anisoptera to be 250 Ma, with a minimum age for Anisoptera of 144 and 120 Ma, respectively. The current fossil calibration will provide more accuracy and precision at the root of the dragonfly tree.

Nel and Huang (2009) placed this species in Cymatophlebiidae based on various alleged synapomorphies, such as the presence of undulating IR2 that is parallel to RP2 and RP3/4 and MA being more strongly undulating compared to other families. We consider the original attribution of Sinacymatophlebia to Cymatophlebiidae to be erroneous and assert that it rather belongs to Paracymatophlebiidae, because it shares all synapomorphies of Paracymatophlebiidae and only differs from this extinct family in the plesiomorphic presence of a second distal lestine oblique vein. The strongly undulating RP3/4 and MA are not only shared with Cymatophlebiidae but also with Paracymatophlebiidae. Contrary to Nel and Huang (2009, p. 203), the published figures of Sinacymatophlebia show that it lacks the main synapomorphy of Cymatophlebiidae, namely the undulating course of IR2 parallel to the undulating RP2. As stated by Nel and Huang (2009), Sinocymatophlebia differs from Cymatophlebioidea (Cymatophlebiidae + Rudiaeschnidae) by the plesiomorphic presence of only two rows of cells between IR2 and Rspl. This latter character state is again shared with Paracymatophlebiidae. According to Bechly et al. (2001), Paracymatophlebiidae is even more closely related to Aeshnidae than Cymatophlebiidae, sharing a larger set of synapomorphies. An attribution of Sinacymatophlebia to Paracymatophlebiidae rather than Cymatophlebiidae would thus not affect the calibration of this node and Sinacymatophlebia mongolica represents the oldest crown group Anisoptera under either scenario.