Observed diversity of bryozoans within Devonian stages is not significantly different from range-through values for species because very few species have ranges longer than a single stage. Generic diversity differs significantly between observed stadial and stadial range-through values except for the diverse Givetian faunas. Among families, observed stadial and range-through comparisons differ significantly in the lower diversity Early and Late Devonian but are not significant in the taxonomically diverse Middle Devonian. The patterns of Devonian generic and familial diversity are apparently robust proxies for specific diversity.

Givetian specific and generic extinctions, based on generated bootstrap distributions, are significantly higher than the other six Devonian stages even when modifications for different stage durations are considered. Based on present data, we conclude that a major change in Devonian diversity occurred between the Givetian and the Frasnian stages. Givetian specific and generic extinctions are not “smeared” with respect to adjacent stages and should qualify as a mass extinction among bryozoans pending more accurate data on bryozoan ranges and more precision in radiometric dating of Devonian stadial boundaries. Devonian bryozoan familial extinctions are not numerous and do not exhibit a significant number of extinctions for any stage. The Givetian extinction, which for the time being ranks as the largest bryozoan extinction in the Phanerozoic, is a global event with a marked local effect in the Hamilton-Tully fauna of New York, Pennsylvania, and Ontario.

Givetian specific originations are significantly higher than bootstrap distributions for both raw data and data modified for stadial durations. A Givetian generic high in originations for raw data is not significant when modified for stadial durations and only Eifelian familial originations are significantly higher than bootstrapped distributions.

Gap-coding permits the use of continuous metric characters in cladistic analyses. Character means are converted to integer equivalents by placing character state divisions in the locations of phenetic breaks between specimen clusters, under the assumption that these breaks represent the locations of bottlenecks in character distributions. Similarities and differences between specimens from closely related species of cystoporate bryozoans were evaluated for the first time by converting continuous morphometric measurements into gap-coded binary and multistate characters and analyzing them cladistically, rather than just phenetically, across multiple species of Strotopora, Cliotrypa ramosa and Fistulipora compressa.

Our results demonstrate that cladistic analysis of gap-coded morphological characters can be effective in resolving phylogenetic relationships at low taxonomic levels (within and among genera) while objectively highlighting both the morphological features that specimens (taxa) share and those characteristics that differentiate them. Differences in cystiphragm abundances and sizes, especially in the proximal portions of colonies, discriminate between species of Strotopora. Colony size and growth form, abundances and lengths of hemiphragms, and sizes of cystopores discriminate between Strotopora and the closely related genus Cliotrypa. Cladistic patterns indicate that Strotopora foveolata Ulrich is a valid species with Strotopora dermata as its junior subjective synonym. Fistulipora compressa is reassigned to the genus Strotopora whereas a decision on the taxonomic status of Cliotrypa ramosa requires a broader cladistic analysis of fistuliporine genera.

Lack of an adequately calibrated Devonian time scale precludes satisfactory comparisons of Recent and Devonian bryozoan diversities. Nevertheless, Devonian bryozoans apparently are 3–100 times less diverse than Recent bryozoans. This variation is a function of how Devonian time is divided (whole period, stages, per million years) as well as the diverse skeletal architecture of cheilostome bryozoans, the most abundant Recent bryozoan order.

The Givetian has the largest specific, generic, and familial diversity of any Devonian stage and the drop in bryozoan diversity from the Givetian to the Frasnian is greater than the rise or fall of diversity between any other adjacent stages. Diversity hardly changes among bryozoans across the Frasnian–Famennian boundary. Among bryozoans the Givetian–Frasnian extinction was a major event.

Devonian taxonomic diversity is less than nomenclatorial diversity, although both exhibit the same trends from stage to stage during the Devonian. The amount of difference between the two measures of diversity is probably a reflection of the intensity of research on Devonian Bryozoa, especially revisionary studies of previously described faunas. Monographic bursts in diversity are present and are a measure of the episodic character of taxonomic studies on less popular phyla. Examination of the literature commonly provides a finer stratigraphic resolution than indicated in compilations such as the Zoological Record.

Although some significant areas of the exposed continents have not received adequate study, Devonian bryozoans show marked geographic changes in diversity that can not be ascribed entirely to lack of study.

Until the late 1960's, most of Nicholson's types of Paleozoic bryozoans were not available for study. We present a set of coded characters of many of Nicholson's types, which should assist in bringing his species into conformity with current taxonomic standards so that his species can be consistently recognized and used in biostratigraphic, paleobiogeographic, and evolutionary studies.

Cladistic and phenetic analyses of these species permit comparisons between inferred phylogenies of Nicholson specimens, adaptive morphospace, and treatise-based systematic relationships. Specimen-based cladistic and phenetic analyses of Nicholson's species both produce clusters that are congruent with existing family-level taxonomic assignments of species in the collection. However, cladistic analysis more fully retrieves the pattern of branching, or degree of relatedness, among higher taxa. Phenetic clusters represent adaptive peaks in morphospace for these specimens, but higher level “phenons” are strongly affected by multiple evolution of the same character states.