No abstract

Phylogenetic analysis was performed on a set of 242 morphological characters. The taxon sample included 31 Libellula, and representative species from selected libeluline tribes, from all libellulid subfamilies, from all libelluloid families, from all anisopteran superfamilies, and Epiophlebia. Corduliinae was shown to be paraphyletic even among genera characterized by a well developed anal loop bisector. Sympetrini was found to be polyphyletic with Crocothemis the sister group to Libellulini. The traditional placement of Trameini, far from Libellulini is in doubt, because it is here placed as the sister group to Crocothemis + Libellulini. Kennedy's phylogeny of Libellula was largely corroborated, with the following exceptions: the subgenera Libellula, Eolibellula, and Syntetrum form a monophyletic group which is the sister group to a clade including Belonia, Holotania, Neotetrum, and Eotainia subgenus nov. [type species Mesothemis composita Hagen]; and Eurothemis is determined to be the sister group of Ladona instead of Neotetrum. In addition we confirm Belonia to be monophyletic, and find Platetrum + Plathemis to form a monophyletic group, sister to Ladona + Eurothemis; these four subgenera together form the sister group to Libellula sensu stricto (s.s.).

No abstract

No abstract

A Circular Depletion Sampler (CDS) was designed to allow the removal of consecutive subsamples from a sample area, while minimizing immigration and emigration. Equal sampling effort was expended during each removal period to permit estimation of both the probability of capture, and the number of individuals not captured from a sample area. Average probabilities of capture from 270 CDS samples ranged from o. 9 per minute for Antocha (Tipulidae) to o.96 for Paratendipes (Chiroal (3 removal periods), Surber, and kick-net methods. Removal population estimates were consistently higher and less variable than those of the other methods. Estimates of diversity from samples taken with conventional methods were biased by differences in catchability, and by emigration. Separation of benthos from detritus took about twice as long per unit sample area for Surber and kick-net methods, although sampling time per unit area was nearly equal for all three methods. Removal estimates using only the first two catches were comparable to estimates based on three catches, indicating a further possible increase in efficiency.