Cordulegastridae - Spiketails

This family is found in both the Old and New worlds, but only Cordulegaster is found in North America, with records of five of the eight species of Cordulegaster found in North America having been recorded in Michigan.

C. erronea is appears to be very rare in Michigan, and only one record appears to have been reported from the western part of the LP. This is mostly a southern species, and Michigan probably represents the northernmost (accidental?) part of its range. This species is on the State's Special Concern List. In south-central Ohio (Hocking Co.), C. erronea inhabits sand substrates of forested seeps and spring-fed brooks with considerable slope (Dan Riggs, personal communication). C. obliqua tends to be found only in first-order, small sandy forest brooks, at least in the southern part of its range (Tennessen et al. 1995). Based on observations of adult C. bilineata and C. diastatops in Michigan, habitat tends to be sandy substrates of downstream pools and runs of hillside springs and brooks, both in forested as well as open or bushy pastures. This appears to agree with habitat descriptions for these species from other workers (e.g., Walker 1958, Tennessen et al. 1995). The rarity of specimens (particularly larvae) from our state reflects insufficient sampling of these habitats. Conversely, C. maculata, which inhabits larger forested creeks and streams with good water quality, is by far Michigan's (and Eastern North America's) most common species of Cordulegaster and is found throughout the state.

Larvae are lie-and-wait predators, usually found mostly submerged in sand and silt protected under pieces of large woody or leafy detritus. These species are long-lived and probably require at least three years in which to develop into adults, at least in the northern part of its range: three different size classes of larval C. maculata have been found in streams from Marquette Co., UP (pers. obs.). Emergence of this species in the north appears to occur in early to mid-June, with females observed ovipositing in early July. Emergence occurs close to the waters edge, usually in June. Johnson (1982) studied prey selection of C. maculata from a stream in the Adirondack Mountains in New York, USA., and found that mayflies (Baetis sp. and Paraleptophlebia sp.) were the principal prey, with chironomids, simulids and caddisflies of various trophic relationships also important components. The author speculated that C. maculata in this stream are active foragers (apparently at night), as their principal prey (in this study, the mayflies) where not well represented in Surber samples that were used to collect larvae in the deeper sections of the stream. Unfortunately, the author did not account for drift in prey species, thus the observation must be considered speculative.

Work over the past two decades (e.g., Carle 1983, Lohmann 1992) to place the species of Cordulegaster in an accepted phylogenetic scheme has not yielded universal support. In describing Zoreana bilineata, Carle (1983) elevates to genus the subgenera Cordulegaster (or Thecophora), Taeniogaster and Zoraena. Going even further, Lohmann (1992) argues that Cordulegaster is an Old World genus, and that the North American species are surviving and individual remnants of formerly speciose, distinct ancient groups. He grouped the eight North American species into the six genera (see table below). However, many researchers - especially in North America - are reluctant to accept these proposals (but see May and Carle 1996), and the key below follows the conservative practice of relegating Taeniogaster, Thecophora and Zoraena as subgenera of Cordulegaster (e.g., Westfall and Tennessen 1996). The table below summarizes the placement of Cordulegaster species: