\ SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 6 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti¬ tution, Joseph Henry articulated a program that included the following statement: “It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge.” This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, com¬ mencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of profes¬ sional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. These publications are distributed by mailing lists to libraries, laboratories, and other interested institutions and specialists throughout the world. Individual copies may be obtained from the Smithsonian Institution Press as long as stocks are available. S. Dillon Ripley Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 6 Alan H. Cheetham Functional Morphology and Biofacies Distribution of Cheilostome Bryozoa in the Danian Stage (Paleocene) of Southern Scandinavia SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON 1971 ABSTRACT Cheetham, Alan H. Functional Morphology and Biofacies Distribution of Cheilo- stome Bryozoa in the Danian Stage (Paleocene) of Southern Scandinavia. Smithsonian Contributions to Paleobiology, number 6, 87 pages, 1971. Highly diversified assemblages of cheilostome Bryozoa in the Danian Stage of southern Sweden and Denmark represent the culmination of primarily divergent evolutionary trends originating in the first appearance of the group in Early Cretaceous time. Functional relationships between colony and zooid morphology are less likely to have been obscured by vestigial structures and convergent and parallel evolution in these assemblages than in later Cenozoic faunas. The Danian assemblages, then, provide a test of the hypothesis that, in the early evolution of cheilostomes, environ¬ mentally correlated variation in the form of colonies depended functionally upon the structure of their component zooids. Theoretically, the rigidly erect growth form should have an adaptive advantage over the presumed ancestral encrusting form, by virtue of a vastly increased potential zooid density relative to substrate occupied. A rigidly erect colony must be able to resist stresses induced by vertical loading, bending, and twisting and thus appears to require calcified walls, especially on the frontal sides of its zooids. Given the constraints imposed by the cheilostome mode of growing and calcifying zooid walls and of operating the hydrostatic system, zooid morphotypes can be relatively graded for efficiency in structural support of the colony by the degree to which their joint calcification approaches a laterally merging, continuously thickening, distally tapering skeletal mass analogous to the outer walls of an enlarging cantilever beam. These hypothetical relationships are generally consistent with biofacies distri¬ butions of more than 50 species associated with a single middle Danian mound in southern Sweden. This mound is typical of many which accumulated, probably at depths approximating the shelf-edge, in southern Scandinavia during Danian time. It includes three biofacies: (1) the flanks, dominated by bryozoans; (2) the core, rich in octocorals with less abundant colonial scleractinians and bryozoans; and (3) transitional areas, between the two, dominated by octocorals but with abundant bryozoans. Sediments of the three biofacies contain distinctive assemblages of cheilostome species which differ in abundance rather than by presence or absence. The flanks are dominated by species inferred to have had erect colonies and the more complex zooid morphotypes. This group of species constitutes the bulk of the total fauna in weight-abundance but fewer than half the species. Species dominant in the (Continued on page iv ) Official publication date is handstamped in a limited number of initial copies and is recorded in the Institution’s annual report, Smithsonian Year. UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1971 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price $1.60 (paper cover) Contents Page Introduction. 1 Acknowledgments. 2 Functional Morphology of Cheilostomes. 4 Adaptive Significance of Morphologic Differences. 4 General Relation of Morphology to Environment. 4 Colony Form. 5 Structural Analysis of Rigidly Erect Colony. 7 Adaptive Features of Rigidly Erect Colony. 10 Zooid Structure. 11 Relation of Zooid Structure to Colony Form. 15 Morphologic Interpretation of Danian Cheilostomes. 16 Characters Related to Colony Form. 16 Characters Related to Zooid Morphotype. 19 Taxonomic Significance of Morphologic Variation. 22 Paleoenvironments of Danian Cheilostomes . 23 Bryozoan Mounds. 23 Biofacies Analysis of Limhamn Mound II-Nj. 25 Comparison with Other Mounds. 34 Inferences about Mound Formation. 35 Biofacies Distribution of Danian Cheilostomes. 37 Distribution in Mound II-N t . 37 Morphologic Basis of Distribution. 44 Conclusions. 47 Literature Cited. 48 Plates 1-17. 53 IV SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ( Abstract — Continued) core facies make up about half the total number of species and are inferred to have had mostly encrusting colonies with zooids of all morphotypes recognized, including the simplest. The transitional facies includes a mixture in subequal proportions of the two groups of species dominant in the other facies; however, this facies has other distinctive species in abundance and thus may represent an ecotone. Morphologically, the cheilostomes abundant in the transitional facies are intermediate in inferred zooid morphotypes and colony forms. The relation between abundance and morphology of Danian cheilostomes suggests that attainment of the more advantageous rigidly erect colony form was functionally more probable for zooid morphotypes susceptible of heavy frontal calcifi¬ cation than for others. If a minimum amount of frontal calcificadon must have been present before the rigidly erect mode of growth could be assumed, then frontal calcification was associated originally with some other function, such as protection of the lophophore. It is possible that the various further advances in zooid morphotype could also have been made as separate prospective adaptations, but it seems more likely that some or all of them represent direct adaptive improvements for the structural support of rigidly erect colonies. Alan H. Cheetham Introduction The evolutionary role of selection requires that sustained trends of morphologic change be adaptive, that is, advantageous to the organism or to the con- specific population of which it is a part (Simpson 1953:160, 1958:534). Morphologic differences between taxa, usually resulting from sustained evolu¬ tionary trends, therefore suggest differences in adapta¬ tion. Within a multispecific assemblage of organisms occupying broadly similar ecologic niches in closely related habitats, the relative abundances of morpho¬ logically differing populations thus should reflect dif¬ fering degrees of adaptation to the general environ¬ ment (Simpson 1953:161, 1958:523). Highly multispecific assemblages of cheilostome Bry- ozoa are known to have occurred from mid-Late Cre¬ taceous time to the present. Since their origin, ap¬ parently in late Early Cretaceous time, cheilostomes have increased in diversity almost steadily (Figure 1). This increase reflects morphologic proliferation with respect both to changes between time-successive as¬ semblages of species and to the range of differences among coeval populations. To the extent that this proliferation applies to assemblages of species that lived in similar environments, it can be assumed to express increasing degrees of adaptation. Alan H. Cheetham, Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Wash¬ ington, D.C. 20560. Functional Morphology and Biofacies Distribution of Cheilostome Bryozoa in the Danian Stage (Paleocene) of Southern Scandinavia The morphologic diversity of cheilostomes encom¬ passes many characters, few of which have been con¬ sidered in terms of both evolutionary trends and re¬ lation to the environment. The growth form of the colony has received much attention because of its cor¬ relation with substrate type and water agitation, but its evolutionary significance has not been fully consid¬ ered. On the other hand, the structural features of the zooids, which have been shown to follow an evolution¬ ary trend of increasing structural complexity through the Cretaceous, have not been investigated for their possible adaptive correlation with other characters or with environmental variables. Other characters, such as kinds and distributions of avicularia, ovicells, and the like, are poorly known in terms of both evolutionary trends and relation to the environment. The present investigation considers relationships among colony form, zooid structure, and environment in assemblages of cheilostomes of Danian age in south¬ ern Scandinavia. This fauna is important to the study of adaptations in cheilostomes because it has a high species diversity within a geographically and apparently ecologically restricted area and because it represents the culmination of the first major episode in cheilostome evolution (Figure 1). The seeming pause in taxonomic diversification during Danian time suggests that chei¬ lostomes had reached an adaptive plateau by the end of the Cretaceous, and yet all three major groups of cheilostomes are diversely represented in the Danian 1 2 fauna. The functional relationships between colony and zooid morphology are less likely to have been ob¬ scured by vestigial structures and convergent and par¬ allel evolution in these assemblages than in later Ceno- zoic faunas, which consist predominantly of species with the most complex zooid morphotypes (ascophor- ans). The adaptive significance of colony form and zooid structure are inferred here from the general evolu¬ tionary history of cheilostome morphology, as present¬ ly understood, and by analogy with living species. The hypothetical relationships of colony form, zooid struc¬ ture, and environment thus proposed are then tested against the relative abundances of Danian species in a series of samples inferred to represent related habitats. Previous studies of cheilostomes from the Scandi¬ navian Danian deposits, summarized by Berthelsen (1962:43-50), have been primarily taxonomic and have included reports of more than a hundred species from Denmark, Sweden, and glacial erratics in north¬ ern Germany and Holland. Cheilostomes from Den¬ mark have been studied extensively by many authors, including Pergens and Meunier (1886), Lang (1921, 1922), Voigt (1923, 1930, 1956, 1968a), Levinsen (1925), Berthelsen (1948, 1962), and Jiirgensen (1968). Many species are known from the erratic boul¬ ders of Germany primarily through the work of Voigt (1924, 1925, 1928, 1930, 1968a), and several have been recorded from the boulder clays of Holland by Veenstra (1963). Danian cheilostomes from Sweden have received much less attention; 16 species were studied by Hennig (1892, 1894), and a few were in¬ cluded by earlier workers in studies of Cretaceous and Danian faunas in the Scandinavian region. In addition to his taxonomic studies on Danian cheilostomes, Ber¬ thelsen (1962:225-256) also determined weight-abun¬ dances of cheilostome skeletal material and relative abundances in number of fragments of cheilostome species in samples from eight localities in Denmark. His purpose was to characterize different Danian lithol¬ ogies in terms of their bryozoan content and to identi¬ fy stratigraphically distinctive species assemblages. Most of the data on which the present study is based are from field observations and samples collected in 1964-1965 from a single moundlike structure of middle Danian age in the large cement quarry in the Limhamn district of Malmo, Sweden. Additional ma¬ terial from other sediments in the Limhamn quarry, from other Swedish localities, and from some Danish localities (including the classical exposures along SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Stevns Klint and in the quarry at Fakse) was collected and examined in an attempt to extend the relationships inferred from the Limhamn mound. This additional material was not examined in the same quantitative detail as that from the one mound, but the differences between samples from similar parts of different mounds appeared generally to be less than those between dif¬ ferent parts of the same mound. Therefore, the quan¬ titative relationships for the Limhamn mounds are thought to apply to all Danian moundlike sedimen¬ tary structures in southern Scandinavia. Acknowledgments Fieldwork was supported by a visiting professorship at the University of Stockholm in 1964—1965 and a grant- in-aid from the Louisiana State University Research Council in the summer of 1965. I am grateful to the Chancellor of the Swedish Universities, to the Director of the Geological Institute, Stockholm University, and to the Dean of the Graduate School of Louisiana State University for making this work possible. The help of Skanska Cement A.B., Malmo, Sweden, and their geological engineer, J. Hedelin, is also gratefully acknowledged. Laboratory studies were supported by assistance in sample preparation during 1964-1965 from the Geol¬ ogical Institute, Stockholm University; by a partial subsistence allowance in the summer of 1965 from the Wenner-Gren Center, Stockholm; and by a re¬ search grant in 1966-1967 from the Smithsonian Re¬ search Foundation. I thank the many people who made this aid available, including Mrs. Maggie Dodds, then of Arlington, Virginia, who donated half a year’s time as a volunteer preparator in the Smithsonian Insti¬ tution’s National Museum of Natural History. During various phases of this study, I was fortunate to have advice and aid from specialists in various taxonomic groups, including Ivar Hessland, Sten Schager, and Krister Brood, Geological Institute, Stockholm University; Ole Berthelsen, Geological Sur¬ vey of Denmark; Theodor Sorgenfrei, Polytechnical University of Denmark; Ehrhard Voigt, Geological Institute, Hamburg University; G. P. Larwood, Uni¬ versity of Durham; Patricia L. Cook, British Museum (Natural History) ; R. M. Finks, Queens College, New York; R. H. Benson and E. G. Kauffman of the De¬ partment of Paleobiology, Smithsonian Institution; and members of the seminar on Bryozoa in the Na¬ tional Museum of Natural History, Smithsonian Insti- NUMBER 6 3 Figure 1.—Division of evolutionary history of cheilostome Bryozoa into two major episodes, as suggested by changes in number of families from Early Cretaceous through Pleistocene. Data are from Larwood et al. (1967), and time scale is after Casey (1964) and Funnell (1964). Two roughly sigmoid curves, Albian-Paleocene and Eocene-Pleistocene, are each separated into a phase of rapid diversification (cladogenesis) and a following quiescent phase (stasigenesis). Because of the generally long ranges of families and the nearly equal intervals of time over which the census was taken, it seems inappropriate to calculate frequencies as time-frequencies. Changes in diversity at the family level, as shown here, may be an inadequate expression of evolutionary activity at lower categorical levels. The earlier evolutionary episode shows increase in all three major groups of cheilostomes—anascan, cribrimorph, and ascophoran. Most increase in the later episode has been in the ascophoran cheilostomes. One genus, Dacryoporella, included by Larwood et al. (1967) among the ascophorans, has been omitted from consideration of family ranges, as discussed in the text (page 15). tution, including R. S. Boardman, O. L. Karklins, W. C. Banta, D. B. Blake, T. G. Gautier, R. W. Hinds, O. B. Nye, R. J. Scolaro, and R. J. Singh. J. N. Foster, an undergraduate research participant from Earlham College, helped interpret some of the samples from Danish localities not included in the detailed analysis. My acknowledgment is not meant to shun full respon¬ sibility for any shortcomings this paper may show in spite of all this help. I am grateful to M. A. Buzas, Smithsonian Institu- NUMBER OF FAMILIES 4 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tion, for use of his program in BASIC for calculating information-function indices for diversity and domi¬ nance. Other assistance in computation was provided by the Smithsonian Information Systems Division through the development of programs and the process¬ ing of some of the data. Henry Feldman wrote programs in BASIC for converting aliquot weights and counts into sample statistics and for weighted- pair-group-method (WPGM) clustering of distance functions and correlations coefficients; H. D. Roth developed a program in FORTRAN for unweighted- pair-group-method (UPGM) and WPGM clustering of correlation coefficients. The greater part of the work has been in the tech¬ nical preparation and processing of material and data. The following are persons to whom I am especially in¬ debted for assistance: M. Ardai for the preparation of samples at the Geological Institute, Stockholm Uni¬ versity; and, at the Smithsonian, Beverly B. Tate for preparing and processing samples and aid in drafting text-figures; Maggie Dodds for processing samples and data and bibliographic aid, in part as a volunteer as¬ sistant; Lorenzo Ford for preparing thin sections of sediments and bryozoan specimens; and L. B. Isham for designing and drafting the text-figures. Photographs of thin sections were made by the Photographic Services Laboratory of the National Museum of Natural History, Smithsonian Institution. I am grateful for helpful suggestions from the fol¬ lowing persons who read the manuscript: Patricia L. Cook, British Museum (Natural History) ; Ivar Hess- land and Krister Brood, Stockholm University; Eckart Hakansson, University of Copenhagen; W. C. Banta, American University; D. B. Blake, University of Illi¬ nois; and R. H. Benson and R. S. Boardman, Smith¬ sonian Institution. Functional Morphology of Cheilostomes Adaptive Significance of Morphologic Differences The differing states of a morphologic feature can be expected, in general, to differ in adaptive value with respect to any particular environment. In the study of adaptive morphology of fossils, two approaches have been used to infer the relative values of differences in a morphologic feature. In the functional-analysis method (e.g., Rudwick 1968:44), a graded series of states is deduced from the theoretical efficiency of the morphologic feature in performing a postulated func¬ tion—this usually having been chosen by analogy with living organisms—and the actual states of the feature in known fossils are then compared with the hypo¬ thetical constructs. The second method is empirical and consists simply of grading the actual series of states of the morphologic feature according to the rela¬ tive abundances of populations within which each state is shown. Neither method, used alone, can be expected to re¬ sult in a full understanding of the adaptive significance of morphologic differences in fossils. It is apparent that empirically graded character states are only the results of evolutionary adaption and do not yield its basis. Functional analysis may not yield adaptive significance directly, because conflicts among the functional prop¬ erties of a feature may require some functions to remain partly or wholly unrealized (Bock and von Wahlert 1965:274). An example of functional conflict in cheilo¬ stomes is that between rigidity and flexibility of the frontal wall; the advantage of having that wall com¬ pletely calcified (except for the operculate orifice through which the lophophore is protruded), to protect the lophophore and associated organs, is partly offset by the necessity of keeping a portion of the wall depres- sible as part of the hydrostatic system (Harmer 1930: 92-99). As pointed out by Dobzhansky (1956:346), it is the whole phenotype of the organism, not each of its separate features, aspects, or stages, which is adapted to the environment. Pleiotropic and other forms of genetic correlation between features, aspects, or stages of the phenotype can result in the appearance and continua¬ tion in successive populations of states of morphologic features whose functions are never realized because of adaptive conflicts. Thus the adaptive significance of any morphologic feature can be expected to depend on the functions of other features, but to varying degrees (Dul- lemeijer 1958). The object of functional analysis of fossils, then, is to deduce adaptively graded series of morphologic states which can be tested against their relative abundances in populations. General Relation of Morphology to Environment A large part of the variation in skeletal morphology of cheilostome Bryozoa appears to be directly related to the environment. The extent of environmental influ¬ ence on morphology is suggested by some of the differ- NUMBER 6 5 ences within a colony, either among the individual zooids themselves or among major regions of the colony (Stach 1935). These are the approximately continuous differences that remain after patterns of intracolony variation related to growth stages of the zooids (ontog¬ eny) and generational differences among zooids which have budded at different stages of colony development (astogeny) have been taken into consideration (Board- man et al. 1970). More distinctly discontinuous intra¬ colony variation in zooid form (polymorphism) pre¬ sumably has a functional basis, such as feeding and reproduction, and therefore is at least indirectly re¬ lated to the environment, whether or not the poly¬ morphism is directly induced by microenvironmental differences (Silen 1938:646, Powell and Cook 1966). Moreover, each set of a polymorphous series of zooids ordinarily displays a similar remainder of microenviron¬ mental variation after growth differences have been taken into consideration (Cheetham 1966:17-21). Non-growth variation shown by the zooids of a col¬ ony (or by one set of zooids in a colony having polymorphous zooids) may be inferred to reflect en¬ vironmental influence on morphology because of the general genetic homogeneity of a bryozoan colony (Stach 1935:646, Boardman and Cheetham 1969:208, Boardman et al. 1970). Lacking correlated genetic differences, phenotypic variants of the same onto¬ genetic, astogenetic, and polymorphic state within a colony must be interpreted as so many physiologic adaptations to varying microenvironments during col¬ ony growth. Although such non-heritable variation lacks direct evolutionary significance (Bock and von Wahlert 1965:284), it provides a baseline for recog¬ nizing morphologic differences which may be heritable, in senarate colonies within or among populations. Although few comparisons between intra- and inter¬ colony variation have been made quantitatively for cheilostomes, the evidence available suggests that these two kinds of variation, expressed as coeffi¬ cients of variation for the linear dimensions of zooids, may not be significantly different (Boardman and Cheetham 1969:226). The relationship between these kinds of variation is apparently unlike that for some features in other colonial groups (Oliver 1968:27). Environmentally induced differences within some colonies are expressed by groups of zooids making up major parts of the colony as well by individual zooids. For example, Cook (1968a: 124) described a Recent colony of Membranipora arborescens (Canu and Bassler) combining broadly encrusting portions with erect, unilaminar fronds. The zooids on the erect fronds are relatively more elongate than those on the encrusting portions and differ from them in some struc¬ tural features. Other Recent specimens of M. arbor¬ escens show one or the other of the colony forms and correlated zooid morphologies (Cook 1968a: 124, pi. 1c,d). The correlation of colony form and zooid mor¬ phology in this species suggested to Cook (1968a: 123) that the two classes of features are a correlated adap¬ tive response to environmental factors, probably dif¬ ferences in availability of suitable substrates. Similarly correlated differences have been described in other species, e.g., Smittoidea variabilis (Canu) (Cheetham 1966:70). In many previous studies of the functional morphology of cheilostomes, such as those by Stach (1936,1937), Voigt (1939), Silen (1942), and Lagaaij and Gautier (1965), it has been inferred that colony form expresses directly the adaptation of the organism to the environment, whereas zooid structure at best reflects adaptations at the individual level of organiza¬ tion. It has even been argued that structural differ¬ ences in zooid skeletons follow orthogenetic trends of diminishing adaptation (Lang 1916, 1919, 1921). The object of this functional analysis is to deduce some of the possible adaptive relationships between colony form and zooid structure. Colony Form The general correlation between some environmental factors and the form of colonies in cheilostome Bryozoa was recognized by earlier workers, but Stach (1936, 1937) was the first to systematize the relation¬ ship by describing nine basic forms of colonies and adducing the conditions of turbulence and substrate under which each will thrive. Although subsequent authors have made further or different divisions of colony forms, their studies of Recent faunas—such as those of Gautier (1962), Lagaaij and Gautier (1965), Cook (1968b), and Schopf (1969)—have generally verified the relationship described by Stach and there¬ fore have justified its application to the reconstruction of paleoenvironments by Stach (1936) and by others such as Berthelsen (1962), Cheetham (1963), Labra- cherie and Prud’homme (1967), and Askren (1968). Stach (1937:80-82) suggested that cheilostomes are divisible into two groups on the basis of their variability in colony form. The first group comprises species, having a stable form, that produce only one colony type regardless of environmental conditions. Adverse conditions simply preclude their occupying that hab- 6 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY itat. The particular colony form shown by a species in this group can be inferred to represent a heritable adaptation fixed through evolution. The second group comprises species, having an unstable form, that pro¬ duce different colony types in various environments. Their ecologic range can therefore be expected gen¬ erally to be greater than that of stable species. The different colony forms shown by an unstable species, taken together, represent a heritable adaptation, but each form separately is a nonheritable, physiologic modification brought on by particular environmental conditions. Stable colony form was thought by Stach (1937:80) to be restricted to species developing only the most specialized types of colonies: loosely encrusting (petraliiform) ; free-living and discoidal or conical (lunulitiform) ; erect-fenestrate (reteporiform) ; erect- jointed (cellariiform and catenicelliform) ; and erect- flexible (flustriform). All species having non-fenestrate, rigidly erect colonies, with either subcylindrical or bilaminate trunks and branches (vinculariiform and eschariform colonies, respectively), or closely encrust¬ ing (membraniporiform) colonies were regarded as un¬ stable. This is a consequence of Stach’s assumption (1937:82) that any species having vinculariiform or eschariform colonies can become encrusting under ad¬ verse conditions of turbulence and that any species having membraniporiform colonies will tend to assume erect growth under favorable conditions. Studies on Recent faunas—such as those by Gautier (1962), Lagaaij and Gautier (1965), and Cook (1968b)—suggest that even though the great majority of cheilostome species have the colony forms regarded by Stach as unstable, only a small proportion of them assumes more than one of the three major growth forms. Variation, for example, in thickness of trunks and branches in erect forms is common, but this type of variation is also known in so-called stable colony forms (Gautier 1962:386, 387). Therefore, it appears that membraniporiform, eschariform, and vinculari¬ iform colonies may be developed in different species, either as stable forms fixed by evolutionary adaptation or as unstable forms varying plastically through physiologic adaptation. An alternative grouping of Stach’s nine colony forms has been suggested by Lagaaij and Gautier (1965:51) and Schopf (1969: 239). Four major groups are based on relation to substrate and the surrounding water: (1) encrusting (membraniporiform and petraliiform), (2) rigidly erect (eschariform, vinculariiform, and retepori¬ form ), (3) nonrigidly erect (flustriform, cellariiform, and catenicelliform), and (4) free-living (lunuliti¬ form) . Each of these groups is susceptible of further or slightly different division than Stach’s nine forms (Lagaaij and Gautier 1965:51, Cook 1968b: 120, Schopf 1969:239). In the study of morphologic adapta¬ tions in cheilostomes, this method of grouping has the advantage of emphasizing the functional, rather than the genetic, basis of colony form. The adaptations shown by two of the colony-form groups and by parts of the other two, involve special¬ ized colony-wide features. Nonrigidly erect forms have major areas of the colony with reduced calcification; lunulitiform colonies develop specialized zooids or extrazooidal tissue to thicken the basal side; petrali¬ iform colonies have non-calcified attachment tubes; and reteporiform colonies develop calcified anasto¬ moses. The principal morphologic differences between the remaining forms in the other two groups, encrust¬ ing and rigidly erect colonies, include the geometric arrangement of zooids. As these colony forms—mem¬ braniporiform, eschariform, and vinculariiform— characterize the great majority of cheilostomes, they are of special significance in the functional morphology of the group. Encrusting and erect colonies differ in several basic adaptations, some of which were recognized by Stach (1937:82). Encrusting colonies are supported directly by the substrate on a zooid-by-zooid basis (Figure 2a). The size of the colony, consideration being given to factors such as competitive growth of other organisms (Boardman et al. 1970), is directly related to the size of the cohesive and relatively smooth surface available for encrustation. Encrusting colonies thus use the habi¬ tat virtually two dimensionally, they depend nearly entirely on movement of the surrounding water to in¬ crease their exposure to food, and they are highly vulnerable to sedimentation and to limitation of the areal extent of suitable substrate. Erect colonies, on the other hand, are less affected by sedimentation and by substrate limitations because they use the habitat in a more distinctly three-dimensional way, increasing their exposure to the surrounding water and their potential for population density relative to the substrate oc¬ cupied. Lacking extensive supportive contact with the substrate, however, they must provide most of their own support, ultimately through the architectural prop¬ erties of their zooids (Figure 2b). Consequently, the size of an erect colony may be directly related to the combined structural strength of its constituent zooids, NUMBER 6 7 a b Membrane ■Calcified Tissue Figure 2. —Generalized morphology of zooids in an en¬ crusting (a) and a rigidly erect ( b) colony in cheilostomes, shown in diagrammatic sagittal (left) and transverse (right) sections. Lophophores, alimentary canals, and other organs have been omitted for simplicity. The direction of colony growth (distal) is indicated by arrows. whereas that of an encrusting colony may be almost independent of it. The inference that the erect colony form is an adap¬ tive advancement over encrusting growth finds some support in the stratigraphic distribution of those cheilo¬ stomes whose growth form can be interpreted. The oldest genera known (Pyripora, Rhammatopora, Wilbertopora, and Charixa), which occur in rocks of Albian or possibly Aptian age in Europe and North America (Voigt, 1968b: 13), all have simple, ap¬ parently encrusting (membraniporiform) zoaria. It has not been possible to substantiate the presence of pre¬ sumed erect cheilostomes in the Albian of Texas, re¬ ported byLaughbaum (1960:1186-1189, Stamenocella sp. and Thyracella sp.; specimens not found in the Southern Methodist University collections, T. E. Wil¬ liams, personal communication, 1967) ; on the other hand, erect cyclostomes are abundant in the same Texas Albian rocks that yield encrusting cheilostomes and in rocks of comparable age in Europe (Canu and Bassler 1926). The first erect cheilostomes thus are apparently Turonian species of Quadricellaria and Stamenocella (Voigt 1959a) having vinculariiform and eschariform zoaria. The more specialized forms of colonies appeared later in the Cretaceous (lunuliti- form, cellariiform) or in the Tertiary (reteporiform, catenicelliform, etc.). Structural Analysis of Rigidly Erect Colony The structural properties of the rigidly erect colony thus are central to understanding adaptation in cheilo¬ stomes. As an architectural structure, this colony form seems analogous to a pillar or beam, and an approach to its analysis is suggested by D’Arcy Thompson’s (1942:967-985) discussion of the strength of these types of structures. In drawing this analogy, it appears necessary to assume only an economical expenditure of energy in the production of skeletal reinforcement, a minimum amount of which is required for the hydro¬ static function of the individual zooids (discussed below) ; that is, zooecial walls should be thickened in proportion to the stresses distributed through the colony as a result of its subjection to forces in the environment. Two sets of forces probably act on any rigidly erect cheilostome colony: those due to the weight of the colony itself, and those due to the movement of the water in which the colony grows. Even though these two sets of forces are probably very unequal in magni¬ tude, they are probably both factors in structural re¬ sponse and may be considered separately. In motionless water, a rigidly erect colony that rises perpendicularly from a horizontal substrate would be¬ have as an evenly loaded pillar if it is either unbranched or symmetrically branched. If the specific gravity of the nonskeletal parts of such a colony can be assumed not to be significantly different from that of sea water, the vertical load on the base of the colony, due pri¬ marily to the weight (in water) of the skeleton, in¬ creases with the growth in volume of skeletal material in the colony. The increase in load is arithmetically proportional to the increase in total colony volume (skeletal and nonskeletal) in species that calcify zo¬ oecial walls only during early ontogenetic stages, but it is exponentially proportional in species that continue to calcify zooecial walls throughout the life of the zooids. Vertical loads on rigidly erect colonies that grow downward, as from roofs of submarine caves, would be equal to those on identical upright colonies, but 8 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY they would induce tensional rather than compressional stress. The ability of the colonial skeleton (zoarium) to resist stresses set up by the vertical load is proportional to the cross-sectional area of skeletal material. The ratio between the cross-sectional area through weight¬ supporting structures and the volume of the body is a well-known relationship restricting body size in terres¬ trial organisms. The restriction of body size is, of DISTAL GROWING TIPS CRITICAL CROSS- SECTIONAL AREA PROXIMAL ENCRUSTING BASE BROKEN TRANSVERSE SECTION a b ■*- ■> Figure 3. —Structure of rigidly erect cheilostome colonies. a, Erect bilaminate (eschariform) zoarium of a Recent speci¬ men of Metrarabdotos ( Biavicularium) tenue tenue (Busk), showing critical cross-sectional area near proximal base and preferred orientation of branches with long axes roughly parallel to plane of figure (after Cheetham 1968, pi. 9:1). b, Transverse section through a bilaminate branch, showing arrangement and relative thickness of zooecial walls and in¬ ferred directions (double-headed arrow) of bending moments (see Figure 4). c, Transverse section through two cylindri¬ cal branches which together have the same number of zoo- ecia of the same volume as in b, but only about half the strength. course, less significant for organisms that live in sea water. Even though the area-to-volume ratio estab¬ lishes a maximum colony size for any given cross- sectional area through the colonial skeleton of a cheilo¬ stome, stresses induced by movement of the surround¬ ing medium probably cause failure at colony sizes far short of the theoretical maximum that can be sup¬ ported in motionless water. Structural modifications for improving the area-to-volume ratio, however, can be expected to be part of the adaptation shown by rigidly erect cheilostomes. These modifications involve shape changes to increase the cross-sectional area of the colonial skeleton in critical vertical-load support¬ ing regions of the colony. In any rigidly erect cheilostome colony, there is a critical cross-sectional area through the skeleton at which failure would occur if the colony exceeded its theoretical maximum size. This area is the smallest one having the greatest load distal to it and ordinarily lies near the proximal end of the colony (Figure 3 a). In species in which zooecial walls do not thicken ap¬ preciably after early ontogenetic stages, and in which the zooids have a nearly constant volume, the only way that the critical area can be larger is by including more zooids. In some Eocene and younger cheilostomes, such as Kleidionella grandis Canu and Bassler (1920, pi. 78:1-6), zooids were apparently added to the proxi¬ mal trunk and branches during growth of the colony; this was accomplished by frontal budding from proxi¬ mal zooids at the same time that distal budding oc¬ curred at the growing tips (Boardman et al. 1970:304, fig. 5). The earlier cheilostomes considered here ap¬ pear not to have had this capability; instead, in some of them, the shape of the whole colony became bilam¬ inate (Figure 3 b) rather than cylindrical (Figure 3c), so as to include a larger number of zooids in the critical zone. Because the cross-sectional area of the combined zooecial walls is greater than in a cylindrical colony having zooids of the same volume, a bilaminate colony has greater vertical-load-bearing strength for any given total colony volume or for any given number of zooids. For the same number of zooids of the same volume, however, a bilaminate colony has either less height or fewer branches than a subcylindrical one and thus does not exploit the rigidly erect mode of growth as effec¬ tively as can a subcylindrical one (see discussion above). Colonies having either bilaminate or subcylindrical form can be more efficiently strengthened for vertical¬ load bearing by an increase in the proportion of skele- NUMBER 6 9 FREE END FIXED END Figure 4. — Relation between bending stresses and skeletal construction in a rigidly erect cheilostome colony. Idealized lines of stress (left) in a longitudinal section through a can¬ tilever beam fixed perpendicularly to the substrate and sub¬ jected to bending moments in either of the directions (double¬ headed arrow) in the plane of the section (redrawn and modified after Thompson 1942, fig. 460). The parabolic lines converging subparallel to the left and right outer surfaces would alternately express compressional stresses as the load was directed toward the left and right, respectively; simul- tal to non-skeletal volume during growth. This is achieved in species in which zooecial walls continue to calcify throughout the life of the zooids. Even though a bilaminate form has more skeletal material than a subcylindrical one, the proportion of skeletal material to non-skeletal volume is about the same in distal parts of the colony as in the critical region, unless zooecial walls continue to accrete ontogenetically. Continued accretion concentrates skeletal volume disproportion¬ ately nearer the critical support zone where zooids are ontogenetically older (Boardman and Cheetham 1969, fig. 4) and thus minimizes the difference between rates of increase in load and strength. Most cheilostomes having rigidly erect colonies prob¬ ably grew in water in which there was appreciable motion (Schopf 1969:236), and therefore the erect colony cannot be regarded simply as an evenly loaded pillar. Bending and torsion moments produced by water movement around the rigid colony set up stresses that are probably much more significant than those in¬ duced by the vertical load. The strength to resist bend¬ ing or twisting is also enhanced by skeletal reinforce¬ ment, and if parsimony in the production of skeletal material is assumed, then the form and position of skeletal structures should be such as to diminish the risk of breaking without exceeding the limits of the organism’s practical expenditure of energy (Thompson 1942:985, Bowman 1961:151, 229). The distribution of stresses under bending moments in a rigidly erect colony can be approximated by those in a cantilever beam fixed at one end and free at the other (Thompson 1942, fig. 460; Bowman 1961, fig. 63). Such beams are ordinarily depicted in a horizon¬ tal position and evenly loaded, as for example by their own weights (see, for comparison, the vertical canti¬ lever shown on the left side of Figure 4). taneously, the lines subparallel to the right and left surfaces would express tensional stresses. The generalized skeletal structure of a rigidly erect cheilostome colony (right) is shown in longitudinal section. Structures in the plane of section—frontal, basal, and transverse walls of two opposing series of zooids—are shown solid, and those out of the plane of section are stippled. In addition, lateral walls form almost continuous sheets of skeletal material parallel to the plane of section, but are not shown. The frontal walls of the zooecia form flanges which are thickest near the fixed end of the colony where bending stresses are greatest. The flanges thicken in such a way as to leave the zooecial cavity undiminished in volume. The thinner basal, transverse, and lateral walls form a web which takes up the less concentrated stresses in the axial region of the colony. 10 As a cantilever, the rigidly erect cheilostome colony has its fixed end placed downward, against the sub¬ strate, and is subjected to bending moments by loads which may impinge on any of its outer surfaces. Thus, in longitudinal section, as shown in Figure 4 (left), the stress lines subparallel to each outer surface alter¬ nately express tension and compression as the bending force alternates in the directions of the double-headed arrow. Concentration of these stresses near the outer surface of the colony suggests that the most efficient arrangement for resistance to bending would place the greatest skeletal reinforcement in this outer region, a reinforcement corresponding to thickening the flanges on an I-beam (Thompson 1942:970). More¬ over, because stresses induced by the bending moments are greatest at the proximal base of the colony, and because the moments in this region increase as the length of the beam increases, it would be advantageous for this outer flange to increase in thickness throughout the ontogeny of the zooids. This increase would be most efficiently accomplished by the addition of skele¬ tal material on the outside, thus increasing the dis¬ tance from the flange to the axis and giving the highest resistance to bending (Nye et al. 1940:116) while keeping zooecial cavities undiminished in size, as shown in Figure 4 (right). Corresponding to the web connecting the flanges of the I-beam, the generally thinner basal, lateral, and transverse walls of the zooecia (Figures 3 and 4) trans¬ mit the stresses across the axial region of the colony. Although these stresses are less concentrated than those in the flanges, they are more evenly distributed along the colony’s axis (dashed line on left side of Figure 4) from the fixed proximal end to the free distal end. Thus the most efficient arrangement of skeletal ma¬ terial in the axial region is uniformly thin walls formed ahead of the outer flange to resist the vertical shearing stresses set up along the colony axis (Nye et al. 1940: 121). The transverse walls also may take up the shear stresses acting in the transverse sections through the colony due to torsion moments (Nye et al. 1940:153). If bending and torsion moments of similar magni¬ tude operate in any direction around the colony axis, it would be advantageous to have the outer flange of thickened skeleton form a cylindrical tube around the whole trunk and around each branch (Thompson 1942:970). If significantly greater loads were applied from one direction than from the others, or from two directions at 180°, then a bilaminate form would be more resistant to bending. A tube yields first by flat- SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY tening perpendicularly to the plane of bending, before it buckles or ruptures (Thompson 1942:971). Resist¬ ance therefore is increased by flattening in the oppo¬ site direction, that is, in the plane of bending (Fig¬ ure 3b ), because of the increased distance between the flange and the axis (Nye et al. 1940:116). An ability to withstand more powerful bending moments in the direction of flattening could explain the success of bilaminate colonies in more rapidly moving water than that tolerated by subcylindrical ones (Satch 1937:82, Lagaaij and Gautier 1965:52, Schopf 1969:235). The longer axes of the trunk and branches of a bilaminate colony thus could be expected to show a preferred orientation in the direction of prevailing water motion (Figure 3a). In simple cheilostomes lacking extensive skeletal reinforcement on the frontal side of the zooid, the flattening of branches may be even more significant as an adaptation to resist bending than it is in more complex cheilostomes. In summary, a rigidly erect cheilostome colony could be strengthened to resist bending and twisting caused by water movement in two ways, assuming economy in the production of skeletal material. First, concentra¬ tion of skeletal reinforcement near the outer surface of the colony would take up the concentrated stress, with the more central, thinner zooecial walls transmit¬ ting the more spread-out stress across the colony axis. Second, a flattening of the trunk and branches would maximize resistance to bending in the plane of flatten¬ ing. Both of these modifications would also strengthen the colony for bearing the load of its own weight. Adaptive Features of Rigidly Erect Colony The morphologic features of rigidly erect colonies in cheilostome Bryozoa thus seem at least in part to be explicable as adaptations for load bearing and for re¬ sistance to bending and twisting. In general, the bila¬ minate trunks and branches of an eschariform colony offer more resistance to both sets of forces than the subcylindrical ones of a vinculariiform colony do. Even without significant skeletal reinforcement on the outer surface of the colony, eschariform types may grow to appreciably greater volume than vinculariiform ones, especially in more rapidly moving water, if the water movement is essentially in one direction (or in two opposing directions). The volume increase is achieved, however, without a corresponding increase in height above the substrate, and therefore eschariform colonies probably cannot expose themselves as fully to the three- NUMBER 6 11 dimensional habitat as taller, but less voluminous vin- culariiform colonies can, and so do not have as high a potential for population density. The vinculariiform colony also appears to be superior in resistance to weak water movement in many directions. Other, more spe¬ cialized colony forms can also be considered to show adaptations against these forces, but, because of their asymmetrical or one-sided cross sections, they face special problems in the distribution of stresses (Nye et al. 1940:119). The reteporiform colony appears to be able to transmit bending and twisting stresses from branch to branch through struts (anastomoses), thus increasing the number of points of support, and there¬ fore probably is better able to tolerate strong loads from many directions. The various types of non-rigidly erect colonies, also common in shallow, turbulent wa¬ ter, substitute elastic skeletal structures for rigid zooe- cial walls at intervals throughout the colony; thus they may resist bending and twisting largely through tensile strength and may use the force of water motion to com¬ pensate for the loss of load-bearing support. A struc¬ tural and stress analysis of lunulitiform colonies would be much more complex than the one attempted here for erect forms, because the stresses probably change drastically as the colony is tumbled on the substrate (Greeley 1967). In the absence of skeletal reinforcement of its outer surface, the power of a rigidly erect colony to resist the forces acting upon it is restricted. With stiffening, a vinculariiform colony can be stronger than an es- chariform colony lacking it. Whereas the shape of the colony is determined by growth as a whole, reinforce¬ ment of its outer walls is determined by the growth of individual zooids. An important and perhaps even the greater part of colony form thus seems to depend upon morphologic features of the zooids, and, in this sense, it is not an independent functional feature. Zooid Structure At the zooid level of organization, the functional morphology of cheilostomes is complicated by poly¬ morphism. Although the functions of some polymorphs may be obvious—as, for example, the brooding of embryos by ovicelled zooids—those of others, such as avicularia and kenozooids, require further investiga¬ tion (see, for example, Kaufmann 1968:54-55). At least some zooids in every colony, perhaps all of them in some colonies, are autozooids and probably ful¬ fill a variety of roles in the growth, nutrition, reproduc¬ tion, respiration, and other processes of the colony. Among the functions of these zooids, those concerned with the movement and protection of the lophophore and associated organs have been discussed most exten¬ sively (see, for example, Harmer 1930:92-99; Ryland 1967b:1040-1041). The similar mode of protruding and retracting the lophophore in living cheilostomes of all types (Marcus 1926:19-21) permits little variation in the general form of most of the walls of the zooecium. The action of the muscles which eff ect these functions is enhanced by the rigidity of the walls in which they originate. The parietal muscles usually originate in the lateral walls and in most cheilostomes are distributed in multiple pairs on either side of the zooid for most of its length. These muscles insert in the hydrostatic membrane, either the membranous frontal wall or the floor of the ascus, depending on zooid structure (Harmer 1930, figs. 2-4). The medially positioned retractor muscle originates on the distal side of the transverse wall and in many cheilostomes paired lateral occlusor muscles originate on the proximal side. These muscles insert, respectively, in the base of the lophophore and the basal side of the operculum. The seating of muscles must be one-sided on lateral walls but can be two- sided on transverse walls as a consequence (Banta 1968:498-499) of the lineal mode of asexual budding in cheilostomes (Lutaud 1961, figs. 18-19), which is such that each zooid has its own lateral walls (exterior walls of Silen, 1944:470) but shares its transverse walls (interior walls of Silen) with adjacent zooids in the same series. The transverse and lateral walls of cheilostomes as different as the relatively soft-bodied Membranipora (Lutaud 1961) and the robustly rigid-bodied Metra- rabdotos (Cheetham 1968) are monotonously similar in their general form and degree of calcification. The basal wall (also an exterior wall) may remain uncal¬ cified or may be skeletally reinforced and thus serve as origin for one or more sets of muscles—as for exam¬ ple, the parietals and opercular occlusors in Stegan- oporella and Labioporella (Harmer 1926, pi. 17:3; Cook 1964: 57, fig. 4). The greatest variation in the form and extent of calcification, however, is in walls at or near the frontal side of the zooid. In the protrusion of the lophophore, the obvious advantage of having the hydrostatic membrane exposed at the frontal surface of the zooid has been difficult to correlate with the equally obvious proliferation of calcified frontal structures in cheilostome evolution. 12 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY It was apparently this difficulty that led Lang (1916, 1919, 1921) to view the progressive calcification at the frontal side of the zooid and the development of cal¬ careous peri-oral structures as an orthogenetic trend of “uncontrolled super-secretion” amounting to a “disease” to which every “calcareous lineage” must “eventually succumb” (1919:195-196). Although he did not deny the possible momentary usefulness of calcified frontal structures, Lang was convinced of their general inadaptiveness because of the common occurrence in cheilostome colonies of some zooids with completely calcified frontal walls and “the consequent death of the zooid, since even the orifice is sealed up” (1916:76). Most other authors have considered frontal calcification to represent an adaptation for protecting the retracted lophophore and associated organs (e.g., Harmer 1930:94, Silen 1944:41), presumably against predation or mechanical damage. Complete or nearly complete closure of frontally calcified zooids seems to be regarded generally as an individual response to extremely unfavorable environmental conditions (e.g., Canu and Bassler 1920:68). Closed zooids are known, however, to be regularly present in particular parts of the colonies of taxonomically and architecturally different groups of cheilostomes, such as species of Cupuladria and Discoporella (see Cook 1965:159, 162-163) having lunulitiform colonies with apical and peripheral areas of closed zooids, and species of Metro- rabdotos (see Cheetham 1968:9, pi. 15:1) having eschariform colonies with proximal trunks composed almost entirely of closed zooids. The regular placement of these groups of closed zooids in the colonial budding pattern and the continued secretion of calcareous ma¬ terial on the frontal sides of the sealed zooids make it unlikely that the process of occlusion in these examples is generally an individual response to an environmental accident. Rather, as Larwood (1969:179-181) has suggested for certain kinds of lattice-like features in cribrimorph cheilostomes—including branched spines, costae, avicularia, and some types of frontal shields— frontal structures grown by zooids may have a colony¬ wide functional significance beyond that directly re¬ lated to the “needs” of the zooids themselves. In investigating the possibility of a general, colony¬ wide functional role for calcified frontal structures, it is necessary to consider the particular functional prop¬ erties of the known classes of frontal structures. Atten¬ tion here is directed to the major wall-forming struc¬ tures, rather than to branched spines, avicularia, or so-called tertiary frontal walls, some of which Larwood JL MORPHOTYPE I MORPHOTYPE .Membrane — Calcified Tissue Figure 5. —Zooid morphotypes I and II. The relations of skeletal tissues to membranes are shown in diagrammatic sagittal (left) and transverse (right) sections (buu, basal wall; cp, communication pore; cr, cryptocyst; fm, frontal membrane; g, gymnocyst; Iw, lateral wall; pm, parietal muscle; sp, spine; tw, transverse wall; u, vestibule beneath orifice). Lophophore, alimentary canal, and other organs have been omitted for simplicity; the direction of colony growth (distal) is indicated by the arrows. In morphotype i (simple membranimorph or flus- trine) the frontal wall is entirely membranous or shows slight exterior calcification on its proximal and lateral margins in the form of a short gymnocyst; there are virtually no subfrontal or suprafrontal skeletal struc¬ tures, but their presence in even rudimentary form re¬ sults in gradation into type n. In morphotype ii (complex membranimorph or cell- ularine) the frontal wall is more extensively calcified at its proximal end to form a long gymnocyst whose length is limited only by the necessity for keeping part of the frontal wall membranous and depressible; cal¬ cified spines, probably tubular extensions of the gym¬ nocyst containing within their lumina evaginations of the frontal wall, generally project over the frontal membrane from its proximal and lateral margins; from the same margins may project subfrontally a short, shelf-like invagination skeletally reinforced with a cal- NUMBER 6 13 MORPHOTYPE IZ .Membrane — Calcified Tissue Figure 6. —Zooid morphotypes III and IV. Representation as in Figure 5 (c, costa; Ic, lacuna between costae; Ip, lumen pore; op, opesiule; other abbreviations as in Figure 5). cified cryptocyst. Lacking a cryptocyst, this morphotype grades into type in, and, lacking proximal and lateral spines, into type iv. In morphotype m (cribrimorph) a major part of the frontal wall, within a short or extensive gymnocyst, is membranous but is overarched by a calcified shield of spine-like costae usually connected to each other with various kinds of processes; the costal lumina are occupied by evaginations of the frontal wall which may be exposed through uncalcified portions of the frontal sides of the costae; an additional wall of solid calcareous tissue of undetermined origin is present above the costal shield in some extinct genera. In morphotype rv (microporoid) a major part of the frontal wall is membranous but is underlain by an extensive cryptocyst calcified in a subfrontal in¬ vagination from the proximal and lateral margins of the frontal membrane; a gymnocyst may be present on the proximal and lateral margins of the cryptocyst; the parietal muscles pass the cryptocyst either be¬ yond its distal margin or through distolateral pores (opesiules). MORPHOTYPE 3ZL .Membrane — Calcified Tissue Figure 7. —Zooid morphotypes V and VI. Representation as in Figure 5 (a, ascus; em, epifrontal membrane; fs, frontal shield; other abbreviations as in Figures 5 and 6). In morphotype v (umbonuloid) the frontal wall is membranous but is overarched by a continuous frontal shield calcified on the underside of a suprafrontal evagination from the proximal and lateral margins of the frontal membrane; uncalcified areas may remain, commonly on the lateral and proximal margins of the shield, through which the epifrontal membrane com¬ municates with that on the zooid interior; the space be¬ tween the frontal membrane and the shield forms an ascus opening to the exterior near the orifice. In morphotype vi (lepralioid or microporelloid) the frontal wall is reinforced by the development just beneath it of a calcified, continuous shield in an in¬ vagination from the proximal and lateral margins of the frontal membrane; uncalcified areas may remain on the margins of the shield and dispersed over its surface; some of these openings provide communica¬ tion between the frontal membrane and that on the zooid interior (Banta 1970 and personal communica¬ tion) ; a new hydrostatic membrane forms as the floor of an ascus invaginated from the frontal membrane under the shield proximally from the orifice. 419-995 0 - 71 -2 14 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY interpreted as a “coarse outer filter” (1969:180) to keep coarse detritus and organisms from the frontal surface of the colony. Because of the seeming functional conflict between rigidity and flexibility of the frontal side of the zooid, and because of the external position of calcareous ma¬ terial relative to the epidermis, there are limited possi¬ bilities for the arrangement of skeletally reinforced walls in this region. On the basis primarily of the modes of development of calcified walls on the frontal side of the zooids and the concomitant development of the hydrostatic system, Harmer (1902:329-339) grouped living cheilostomes in six major morphotypes (Fig¬ ures 5—7). To the extent that mode of growth can be interpreted from the morphology and microstructure of the skeleton (Boardman and Cheetham 1969:220- 223), these morphotypes are also recognizable among fossil cheilostomes. Criteria specifically applicable to interpretation of Danian cheilostomes are discussed below, in a separate section. Among living cheilostomes, the generally gradational relationships of the basic zooid morphotypes in two or more series of increasingly anatomically complex forms have been recognized by Harmer (1902, 1930), Silen (1942), and others. One major morphologic series begins with simple membranimorph anascans (mor- photype i, Figure 5) having the frontal side of the zooid with a simple, almost entirely membranous fron¬ tal wall; progresses through complex membranimorph anascans (morphotype n, Figure 5) having proxi- mally and laterally calcified frontal walls (gymno- cysts) with hollow, tubular, calcified extensions (spines) projecting over the membranous part of the frontal wall; continues through cribrimorphs (mor¬ photype m, Figure 6) having frontal shields composed of hollow costae, presumably comparable to spines which have fused medially and usually also laterally; and may culminate in umbonuloid ascophorans (mor¬ photype v, Figure 7) having continuous frontal shields calcified on the underside of a fold-like evagination, presumably comparable to completely fused costae, the upper surfaces of which have failed to calcify. In all of these morphotypes, the hydrostatic membrane is the original membranous frontal wall. The second major morphologic series also begins with morphotoype I; it progresses through microporoid anascans (morphotype iv, Figure 6) having extensive subfrontal shields (cryptocysts) calcified within an in¬ vagination into the body cavity in such a way as to permit insertion of the parietal muscles in the mem¬ branous frontal wall; and it culminates in lepralioid or microporelloid ascophorans (morphotype vi, Fig¬ ure 7) having continuous frontal shields developed as cryptocysts beneath which the hydrostatic membrane is invaginated together with attached parietal muscles. In morphotype vi, the position of the hydrostatic mem¬ brane relative to the original membranous frontal wall differs from that of all other known cheilostome mor¬ photypes. A developmental relationship between microporoid anascans (morphotype iv) and lepralioid-microporel- loid ascophorans (morphotype vi) was suggested by Harmer (1902:333), in part on the basis of the anatomy of microporoid genera described by Jullien (1881:276-285, in his anascan suborder Diploder- mata). Subsequently, Harmer’s suggestion was either ignored (Harmer 1930:99, Silen 1942:46) or denied (Ostrcrumov 1903), but recently it has been corro¬ borated by Banta’s (1970, and personal communica¬ tion) detailed morphologic investigations of the lep¬ ralioid genera Watersipora and Schizoporella. A third morphologic series is at least theoretically possible. This is a progression from anascans of mor¬ photype n, in which the gymnocyst is extensive, to ascophorans through invagination beneath the gymno¬ cyst of a hydrostatic membrane. The genus Pseudole- pralia has been assumed to be an ascophoran with a gymnocystal frontal shield (Silen 1942:49-54), but the mode of development of the shield (Silen 1941:39, 1942:50) appears to differ from that of a gymnocyst. The original frontal membrane forms the floor of the ascus rather than having been at the level of the “gymnocyst,” and the frontal shield of Pseudolepralia thus might be comparable to those of umbonuloids (morphotype v), except for calcifying on the upper rather than the lower side of the fold-like evagination. Whether it develops on a fold or on the frontal mem¬ brane, the frontal shield of Pseudolepralia differs from those of morphotypes v “and vi in one respect signifi¬ cant to the following discussion of the possible colony¬ wide functional role of frontal structures: that is, be¬ cause it lies exposed at the outer surface of the zooid, without an overlying secretory membrane (Silen 1942, fig. 50), it lacks the potential for continued accre¬ tionary thickening. Although the recognition of zooid morphotypes in fossil cheilostomes is still tentative, the stratigraphic sequence of first appearances of morphotypes appears to be largely in agreement with the two major morpho¬ logic series described above. Morphotypes i and n NUMBER 6 15 (membranimorphs) appeared in Early Cretaceous time, in the Albian or possibly as early as Aptian (Voigt 1968b: 13). Morphotypes m (cribrimorphs) and iv (microporoids) made their debut soon after the be¬ ginning of Late Cretaceous time, in the Cenomanian (Larwood et al. 1967:388, Voigt 1967). The time of origin of morphotypes v and vi (umbonuloid and lepralioid-microporelloid ascophorans) is more doubt¬ ful, partly because of the difficulty of distinguishing the two modes of development in fossils, and partly because of the incompleteness with which the morphology of some early genera is known. If it is assumed that sup¬ posed porinid genera, such as Rotiporina, developed as do the modern species of Porina, and that supposed exochellid genera, such as Balantiostoma, developed as do the modem species of Escharoides, then these morphotypes date at least from the Turonian and Cam¬ panian, respectively. The supposed hippothoid genus Dacryoporella, which nominally includes species older than Turonian, and which was accepted by Larwood et al. (1967:390) as an ascophoran, could be inter¬ preted as an anascan of morphotype ii (see Voigt, 1968b: 13) ; the oldest (Cenomanian) species of this genus, D. reussi (Lang) (see Larwood et al. 1967: 390) is especially suspect, because Lang (1914:443) only provisionally assigned it to Dacryoporella on ac¬ count of its subcircular (membranimorph-like) zo- oecial openings. In general, therefore, the graded series of morphotypes appear to have evolutionary significance. Relation of Zooid Structure to Colony Form A possible colony-wide function of frontal calcification, which appears not to have been given detailed con¬ sideration previously, is architectural support of an erect colony. As discussed above, the strength of an erect colony, under the load of its own weight in sea water and under bending and torsion moments due to water movement and possibly other forces, is especially enhanced by concentration of reinforcing skeletal ma¬ terial near the outer surface of the zoarium (Figure 2b) . With regard to the origin and mode of growth of calcareous frontal structures, three steps are involved in the attainment of greatest functional efficiency in this role (Figure 8) : (1) if frontal structures become continuous as they extend medially and distally across the zooid, they j oin the vertical walls together, forming a bridge on the outer surface of the zoarium; (2) if, in growing medially and distally, these structures can JOINING intrazooecial walls THICKENING zoarial cover BINDING interzooecial walls Figure 8. —Possible functional steps in development of cal¬ careous frontal structures for colony support. The steps are cumulative in that during zooid ontogeny, joining is followed by thickening, and thickening by binding. accrete frontally during a major portion of the life of the colony, they thicken the zoarium; and (3) if, in thickening frontally, these structures can also spread laterally over the frontal edges of the lateral walls to merge with frontal structures of adjacent zooids, they bind the individual zooecia into a unified zoarial struc¬ ture. These three steps give frontal structures a cumu¬ lative significance (Rudwick 1968:46), increasing the functional effectiveness of frontal structures during the ontogeny of the zooids by which they are grown. This cumulative effectiveness is expressed as an increase, greatest near the proximal end and less toward the distal end, in the critical cross-sectional area of the colony, and also as a concentration of mass farther from the colony axis that, again, is greatest near the proximal end and decreases toward the distal end. Because of limitations imposed by the mode of growth of their frontal structures, not all zooid morphotypes are capable of taking all three functional steps in their ontogeny. Thus, with respect to the postulated function of colony support, the six basic morphotypes can be ar¬ ranged in a graded series of progressive adaptations for the erect mode of life on the basis of their relative effectiveness in each functional step (Figure 9). In zooecia of morphotype I, the virtual absence of frontal skeletal structures makes their functional importance 16 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY ALL FUNCTIONS - 21 - - 2 - - 12 - -m J -n- BINDING - 12 - m- -n- -i- ■m- - 2 - i i i i i i i T -j— 2E- - 1 — 2 - THICKENING -m- ■32- i JOINING r*—12- m- -n- -m- - 2 - i INCREASING EFFECTIVENESS IN COLONY SUPPORT Figure 9. —Inferred potential effectiveness of the six zooid morphotypes in three functional steps involved in structural support of the rigidly erect colony. The scale of effectiveness is arbitrary and intended only to illustrate approximate rel¬ ative differences among morphotypes. The effectiveness in all functions is an approximate average of those for the three steps. negligible in all aspects of colony support. The extensive gymnocyst of morphotype ii places this zooecial type higher on the scale, but the effectiveness of the gymno¬ cyst is subject to two limitations: first, the joining of the lateral walls to each other cannot be as extensive as in more complex morphotypes because of the neces¬ sity for keeping an appreciable part of the frontal wall membranous and depressible; and, second, be¬ cause of the necessity for keeping the body cavity undi¬ minished in size, the gymnocyst, which lies at the exterior frontal surface of the zooid and thus is secreted from within the zooecial cavity, cannot be appreciably thickened ontogentically or merge with gymnocysts of adjacent zooids. The costal shield of type hi zooecia is only slightly more effective than a gymnocyst; even though costae are effective in joining the lateral walls together for a great proportion of their length, costae like gymnocysts apparently are secreted from within and therefore are not susceptible of appreciable thick¬ ening or of merging across lateral walls of adjacent zooids. The so-called tertiary frontal walls, formed in some Cretaceous cribrimorphs possibly by encroachment of interzooecial tissue over the costal shield (Larwood 1962:33, 40-41), may improve significantly the poten¬ tial effectiveness of this morphotype, but the mode of formation of this type of structure has not been deter¬ mined. In type iv zooecia, the extensive cryptocyst has a joining ability greater than that of a gymnocyst, because the gaps in it, aside from that for passage of the lopho- phore which occurs in all types of frontal shields, need be only large enough to accommodate the parietal muscles. In addition, its position beneath the secretory epithelium gives the cryptocyst a greatly increased capacity for thickening. Because the cryptocyst lies beneath the hydrostatic membrane, however, it cannot merge with cryptocysts of adjacent zooids without inter¬ fering with the hydrostatic system. The greatest ca¬ pacity for colony support appears to be embodied in the ascophoran structure of type v and vi zooecia. These have a frontal shield which is continuous for much of the length of the zooid and which, because of its position above the hydrostatic membrane but be¬ neath the secreting membrane, can thicken without apparent limit throughout zooid ontogeny and can merge across the lateral walls of adjacent zooids. There is no apparent difference between the two ascophoran morphotypes with respect to the three functional steps. Thus, on theoretical grounds, the morphotype of the zooid appears to be the primary functional feature on which adaptations, including those expressed in colony form, are based. This hypothesis is tested in the fol¬ lowing analysis of cheilostomes from the Danian Stage in southern Scandinavia. Morphologic Interpretation of Danian Cheilostomes Characters Related to Colony Form Measurement of abundances of different colony forms in assemblages of fossil or Recent cheilostomes re¬ quires morphologic interpretation of the typically fragmentary specimens incorporated into sediment (see, for example, Maxwell 1968:271). Even though NUMBER 6 17 some specimens in an assemblage are ordinarily pre¬ served well enough to indicate their growth form di¬ rectly (see, for example, the encrusting zoarium on Plate 1:4 and the rigidly erect, branching zoarium on Plate 11:2), it cannot be assumed that all conspecific colonies had the same form. In addition, some colony forms are nearly always fragmentary in sediments; jointed colonies, for example, can be expected to be found as disarticulated internodes with only their in¬ conspicuous nodal structures to indicate their form. Therefore, the interpretation of colony forms in an assemblage requires the application of indirect criteria of zooecial morphology and its pattern of variation within the zoarial fragments (Cook 1968b: 119, Board- man et al. 1970:309). As remarked upon by Hennig (1899:38-39), the Danian cheilostomes in Scandinavia are typically skele¬ tal fragments rarely larger than 1 cm and ranging down to particles just recognizable as cheilostome skeletal material *4 mm in diameter. For the most part, these fragments appear not to have retained their positions relative to each other or to the substrate (see Plate 17:1-3). Among the Danian cheilostomes of eastern Den¬ mark, Berthelsen (1962:238) recognized four forms of colonies: vinculariiform, eschariform, membrani- poriform, and lunulitiform. The last of these forms is represented by a single species, Lunulites saltholmensis Berthelsen, which occurs chiefly in the upper part of the Danian at a few localities (Berthelsen 1962:155). This species and two similar ones, Lunulites sp. and Vibracella (Discovibracella) oculata Voigt, have been found in rocks of about the same age, or slightly younger (Dano-Montian), in Poland (Voigt 1964: 441-443, Maryanska 1969:112-113). Lunulitiform colonies have not been reported from most of the Danian sediments in Scandinavia (Berthelsen 1962: 238), and no zoaria that could be interpreted as lunu¬ litiform were encountered in the present study. Voigt (1964:457-458, pi. 14:1-3) described and il¬ lustrated from the Dano-Montian of Poland specimens of Pavobeisselina oblita (Kade) having tapering proxi¬ mal ends perforated with numerous small openings which he interpreted as indicating “attachment by radicells or by a chitinous stalk.” Similar interpreta¬ tions of openings in the proximal ends of zoarial frag¬ ments of other species, such as Onychocellaria rhombea (Hagenow) and Smittipora? canalifera (Hagenow) from the Maestrichtian, have led to the conclu¬ sion that such specimens represent segments of jointed (cellariiform) colonies (Voigt 1957:15-17; 1968: 27-29). A proximally tapering shape, even without openings in the proximal end, has been considered evi¬ dence of a cellariiform colony in the Maestrichtian species Micropora transversa (d’Orbigny) (Voigt 1968b: 32). Pavobeisselina oblita and a species similar to Smittipora? canalifera, S.? prismatica (Hagenow), are known from the Danian in Scandinavia but have been interpreted to have had a rigidly erect colony form (Berthelsen 1962:67, 201). Some Danian speci¬ mens of S.? prismatica (Plate 7:1) have tapering prox¬ imal ends occupied by zooecia whose frontal sides are occluded except for a small, suboral pore, or slit. Though these closed zooecia are shown by the pres¬ ence of opercular scars to have been parts of auto- zooids, the openings left in them may have held noncalcified tubes. Such tubes, however, are known in living chielostomes having erect, non-jointed colonies. In Zeuglopora arctata Harmer (1957:756), Siphoni- cytara formosa Harmer (1957:893), and Cleido- chasma biavicularium (Canu and Bassler) (Harmer 1957:1048), non-calcified tubes form rootlets anchor¬ ing the otherwise rigidly erect colonies to the substrate. In jointed colonies of living and fossil species of Nellia, Poricellaria, Cellaria, Margaretta, and other typical cellariiform genera, openings for the tubes connecting internodes are present not only at the proximal end of the internode, but also at the distal end or at some position intermediate between the two ends. Distally placed structures of this sort appear not to have been described from the Cretaceous and Danian fossils dis¬ cussed here and have not been observed in the present study. Therefore, cellariiform colonies are not inferred to have been present in the Danian fauna, and the zoarial fragments in this study have been classified in the three forms: vinculariiform, eschariform, and membraniporiform. Vinculariiform Colonies (Plates 7:1-3, 5-6; 11): Zoarial fragments inferred to have been parts of vin¬ culariiform colonies are cylindrical or subcylindrical. Locally, in regions of branching (Plate 11:2-3), the stem may be flattened, and fragments from these re¬ gions of a colony might be mistaken for eschariform specimens if some of them did not show gradation into a cylindrical form distally (see Jiirgensen’s, 1968, pi. 1:3-4, separation of two forms of Floridina gothica ). On growing tips (Plate 11:1) and in transverse sections (Plate 12:3-4), zooecia are arranged radially. This arrangement is more conspicuous in species in which the zooecia are grouped in verticillate whorls (Plate 18 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 11:1-4), than in species in which they are in alter¬ nating series (Plate 7:1-3, 5-6). Zooecia typically lack basal walls because their lateral walls reach the axis of the stem (Plates 3:8; 12:3-4). In Floridina gothica (d’Orbigny), however, narrow basal walls are present, and the stem axis has a small, triangular hol¬ low (Plate 3:7). This hollow might be the remnant of a soft structure encrusted by the colony in the fashion described by Stach (1936:62) as pseudovin- culariiform, but it has a constant size and form in all zoarial fragments of this species and lacks any signs of the patterns normally present on the basal surfaces of encrusting colonies. Moreover, the zoarial form of F. gothica seems to be more regular than known pseu- dovinculariiform colonies in species such as Metrarah- dotos ( Uniavicularium ) unguiculatum Canu and Bass- ler (Cheetham 1968, pi. 10:2). Eschariform Colonies (Plates 2; 3:1-5; 7:4, 7; 8-10; 12:5-6) : Bilaminate zoarial fragments have been interpreted in this study as parts of eschariform colo¬ nies. There is some variation in the width of branches in the bilaminate Danian cheilostomes, but, because of the fragmentary condition of the material, it has not been possible to separate consistently the narrow- branched (adeoniform) colonies from the others. In some species, such as Pithodella cincta Marsson, uni¬ laminate fragments commonly occur together with bilaminate ones, and, because of the regularity of both frontal (Plate 2:4) and basal surfaces of these uni¬ laminate specimens, they have been inferred to be parts of eschariform colonies rather than membraniporiform ones. The proximal parts of some bilaminate fragments are subcylindrical (Plates 2:2, 3:5, 8:2), and such specimens might be mistaken for vinculariiform colo¬ nies. None of these fragments, however, shows a dis- talward gradation into subcylindrical shape. Typical bilaminate arrangement of zooecia can be observed on growing margins (Plates 8:2, 10:4) or in transverse sections (Plates 3:2-3, 5; 9:1, 3; 12:5-6). Most or all of the zooecia have wide basal walls, those of the two layers of zooecia being juxtaposed to form a more or less regular median lamella. Structurally, the median lamella is double-walled, consisting of the basal walls of the two layers of zooecia, and it is discontinu¬ ous, being interrupted by the boundaries between lat¬ eral walls of each pair of contiguous zooecia. Actually, the basal wall of each zooecium is structurally con¬ tinuous with its lateral walls, and many of the bilami¬ nate colonies appear to have broken into the two constituent layers of zooecia, with each layer broken further into lineal series of zooecia. The zooecia within lineal series would tend to resist breaking because of sharing transverse walls. At the lateral margins of a bilaminate specimen, the zooecia, which may be autozooecia or heterozooecia, either have basal walls essentially like those of the more centrally placed zooecia (Plates 9:3; 12:5-6) and thus continue the median lamella to the lateral margin of the zoarium, or lack basal walls (Plate 3:2-3, 5) and thus cut off the median lamella short of the margins. Subcylindrical portions of colonies having shorter median lamellae (Plate 3:5) closely ap¬ proximate the vinculariiform arrangement. Among the zoarial fragments from the Danian in the cement quarry at Limhamn, Sweden, interpreted as vinculariiform and eschariform colonies, specimens preserving growing extremities are common. On the the other hand, as with the erect cyclostome Bryozoa from the same locality (Brood 1970, personal commu¬ nication), specimens showing proximal attachment surfaces are scarce. In any assemblage of large, many- branched colonies, the growing extremities would of course greatly outnumber the attachment surfaces. Moreover, in large, erect colonies, such as in Metrarab- dotos moniliferum (Milne Edwards) (Cheetham 1968, pi. 15:1-2), the proximal zooecia may be occluded and no longer comparable morphologically with the more distal ones. Comparable ontogenetic differences between zooecia in proximal and distal parts of Danian cheilostome colonies are suggested by fragments of Coscinopleura angusta Berthelsen (Plate 8:2-3) and some other erect species. Therefore, many of the es¬ chariform and vinculariiform cheilostomes may have had large colonies despite the small size of the frag¬ ments incorporated in the sediments sampled, and the proximal parts of the colonies bearing the attach¬ ment surfaces may be included among the unidenti¬ fied cheilostome fragments in each sample (see Figure 22). Alternatively, the paucity of recognizable attach¬ ment surfaces could indicate transportation of colony fragments away from the site of their growth (Brood, personal communication), though evidence discussed below suggests that this is not the case. Membraniporiform colonies (Plates 1, 4-6) : Those unilaminate zoarial fragments which show direct or indirect evidence of adherence to the substrate by their whole basal surface have been inferred in this study to represent membraniporiform colonies. Small zoaria (Plates 5:la-b; 6:1) may adhere entirely to larger single objects, such as barnacle valves or octocoral NUMBER 6 19 axes. Larger zoaria (Plate 1:2) may remain attached to only one or more small objects, such as echinoderm ossicles. In membraniporiform colonies, the basal wall of each zooid is in contact with the substrate, and in many of the Danian zoaria the basal walls apparently were uncalcified (Plates 1:1-4, 6:1) so that the sub¬ strate is visible in frontal view. In others the basal walls were completely calcified (Plates 5: la; 6:2), and in many of these forms the topography of the basal wall conforms to that of the encrusted object. Where the colony appears to have become detached from its sub¬ strate, the basal surfaces of the zooecia may show these topographic characteristics, as in the specimen of Callopora sp. (Plate 6:2) having pits suggesting growth on a calcareous sponge. Some zoarial fragments reflect the overall shape of their still adherent substrate by having the whole layer of zooecia roughly conform to it, as the subcylindrical zoarium of Aechmella pindborgi Berthelsen (Plate 5:1) on an octocoral axis, the domed specimen of Onychocella ravni Berthelsen (Plate 5:2) on a calcareous sponge, or the arched specimen of the same species (Plate 5:3) on the outer ends of the septa of a scleractinian corallite. Still another morphologic expression of substrate topography is the alternation of budding directions in the zoarium of Micropora hennigiana Berthelsen (Plate 6:3) on a ribbed brach- iopod shell. Many free, unilaminate fragments (Plate 4:1-3) have shapes suggesting growth on these types of substrates. Characters Related to Zooid Morphotype Frontal surfaces of zooecia are generally well preserved in Danian cheilostomes, and growing extremities of colonies preserving ontogenetic stages of frontal struc¬ tures are common. The disappearance of membranes during fossilization, however, makes reconstruction of some zooid morphotypes doubtful. Consequently, it is not possible to group specimens on any single mor¬ phologic criterion; emphasis has been placed instead on possession of a majority of characteristics typical of each morphotype. Morphotype i (Plate 1:1, 3) : These zooecia are broadly open frontally within narrow, calcified rims, inferred to represent gymnocysts, which end at distinct, usually grooved zooecial boundaries, assumed to mark the position of the intercalary cuticle between lineal series of zooids. Basal walls commonly are uncalcified; spines, cryptocysts, and adventitious avicularia are vir¬ tually lacking. Morphotype ii (Plates 1:2, 4-5; 2; 3:1-5) : These zooecia have moderate to extensive frontal calcification in the form of a convex lamella, inferred to represent a gymnocyst. In longitudinal section (Plate 3:1, 4), this lamella has about the same thickness as the other zooecial walls The mural rim, marking the inner margin of the gymnocyst, commonly supports spine bases, each with a conspicuous lumen (Plate 2:3-^4). Within the mural rim, a slightly to moderately de¬ veloped, depressed shelf, inferred to represent a crypto¬ cyst, may be present (Plate 2:1-4). Adventitious avicularia occur on the gymnocysts of some zooecia (Plate 2:1-2, 4). Zooecial boundaries are usually marked by furrows in frontal view and are discernible in transverse sections as distinct lines, presumably the position of the intercalary cuticles in the living colony, running from the basal wall to the frontal surface of the zoarium (Plate 3:2-3, 5). Morphotype hi (Plate 4:2-5) : These zooecia show differing kinds of frontal costation, extending over virtually the whole frontal surface or limited to the central part of it. The costae may be fused only at their medial ends or show numerous lateral fusions. Species of Tricephalopora (Plate 4:5) have tertiary frontal walls, but in general these structures are less common in Danian species than in Cretaceous cribrimorphs. Morphotype iv (Plates 3:6-9; 5-9) : These zo¬ oecia have extensive frontal lamellae, inferred to represent cryptocysts. These structures are centrally depressed, at least in early ontogenetic stages on zoarial fragments preserving the growing extremities of col¬ onies (Plate 8:2). In specimens inferred to represent later ontogenetic stages, the cryptocyst may be much thicker and not depressed (Plates 8:3, 9:1-3). The boundaries between type rv zooecia, inferred to mark the intercalary cuticles, remain distinct in frontal view, even where the cryptocyst apparently has been greatly thickened ontogenetically (Plate 8:3); however, in transverse sections (Plate 9:1,3) the boundaries may become progressively less distinct frontally, possibly through calcification of the intercalary cuticle (see Banta 1968:499). The distal part of the cryptocyst in type iv zooecia reflects structurally the passage of the parietal muscles, either through proximolateral indentations in the opesia (Plates 5; 6:4; 7:1, 3-5, 7; 8) or through lateral opesiules proximal to the opesia (Plates 6:1-3; 7:2, 6). In Floridina gothica (Plate 7:2-3), zooecia having opesiules appear to occur in the same zoarial fragment with those having indentations in the opesia, Table 1. —General morphologic and distributional characteristics of cheilostome Bryozoa in middle Danian Mound II-Nj, Limhamn Quarry , Sweden. Species are ranked in order of maximum abundance in coarse fraction of any sample. Weight- Rank Species Inferred zoarial form (if more than one, dominant first ) Inferred Zooecial morphotype Total occur - rences Occurrence in Denmark—Zones B c D 1 Coscinopleura angusta Berthelsen. . ESCHAR IV 13 X X X 2 Floridina gothica (D’Orbigny). . VINC IV 12 X X X 3 Aechmella pindborgi Berthelsen. .... MEMB IV 16 X X X 4 Membraniporidra lacrymoporoides Berthelsen. . ESCHAR II 10 X X X 5 Callopora spp. . MEMB + ESCHAR II 15 ? ? ? 6 Columnotheca cribrosa Marsson. . VINC V or VI 13 X X X 7 Floridina spp. . MEMB IV 16 ? ? ? 8 Pithodella cincta Marsson. . ESCHAR II 15 X X X 9 Porina salebrosa Marsson. . ESCHAR V or VI 16 X X X 10 Onychocella ravni Berthelsen. . MEMB IV 15 X X 11 Aechmella tenuis Berthelsen. . ESCHAR IV 14 X X 12 Porina cylindrica Voigt. . ESCHAR V or VI 11 X X X 13 Onychocella? columella Berthelsen. . ESCHAR IV 17 X X X 14 Pachythecella lundgreni (Pergens & Meunier). . . . . ESCHAR V or VI 10 X X X 15 Smittipora? prismadca (Hagenow). . VINC IV 11 X X X 16 Puncturiella sculpta (D’Orbigny) . . VINC + MEMB IV 11 X X X 17 Pachythecella anhaltina (Voigt) . . ESCHAR V or VI 17 X X X 18 Semiescharinella complanata D’Orbigny . . ESCHAR -|- MEMB IV 17 X X X 19 Balantiostoma vallata Maryanska . . MEMB V or VI 14 ? p ? 20 Floridina voigti Bassler . . ESCHAR IV 6 X X 21 Smittipora? sp. . ESCHAR IV 10 p ? p 22 Membraniporidra huckeana Voigt . . MEMB II 14 X X 23 Pliophloea spp. . MEMB III 13 p p ? 24 “Cellepora” daniensis Voigt -f- spp. . MEMB V or VI 13 p ? ? 25 Micropora hennigiana Berthelsen + spp. . MEMB + ESCHAR IV 17 ? ? ? 26 Ellisina brittanica (Brydone). . MEMB I 10 X X X 27 Crassicellepora voigti Berthelsen. . MEMB V or VI 9 X X 28 Membraniporidra declivis (Marsson) . . MEMB -f* ESCHAR II 12 X X X 29 Tricephalopora circumvallata (Levinsen) . . MEMB III 13 X X 30 Psilosecos angustidens (Levinsen) . . MEMB V or VI 12 X X 31 Tricephalopora cerberus Lang . . ESCHAR III 3 X X X 32 Allantopora stomatoporoides Lang . . MEMB II 10 X 33 Anornithopora minuta Voigt . . MEMB III 13 X X 34 Anornithopora polygona Voigt . . MEMB III 4 X X 35 “Herpetopora” danica Lang . . MEMB I 7 X 36 Monoceratopora quadrisulcata (Hennig) . . MEMB III 11 X X X 37 Gargantua parvicella (Voigt) . . MEMB IV 14 X X 38 Membranipora? johnstrupi Berthelsen . . MEMB II 4 X X 39 Aplousina? oedumi Berthelsen . . MEMB I 4 X X X 40 Onychocella poulseni Berthelsen . . MEMB IV 5 X X X 41 Pelmatopora? daniensis Voigt . . MEMB III 2 X 42 Cryptostomella pectinata Berthelsen . . MEMB V or VI 2 X X 43 Fissuricella fissa (Voigt) . . MEMB II 1 X X X 44 Pithodella? pristis (Levinsen) . . ESCHAR II 2 X X X 45 Tricephalopora robusta Berthelsen . . MEMB III 3 X 46 Systenostoma pontiferum Berthelsen . . ESCHAR V or VI 2 X X 47 Pachydera fissa Berthelsen . . MEMB III 3 X 48 Membranipora? sp. . MEMB I 1 ? ? ? 49 Phractoporella cordiformis (Levinsen) . . MEMB III 2 X X 50 Pliophloea vincularioides Voigt . . VINC III 1 X X 51 Leptocheilopora laticostata Berthelsen? . . MEMB III 1 X X Total number of species in samples 24 42 36 equivalent percentages are based on counts of fragments and weights of each size grade for each sample. Assemblages are catalogued USNM 169554—169570. Denmark occurences from Berthelsen, 1962. Sample occurrence in i weight-equivalent percent oj identifiable specimens in coarse fraction (>500 y.) Rank 1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 17 18 0. 26 2. 91 _ 1. 42 0. 23 6. 78 12. 99 0. 20 1.06 9. 62 15. 31 18. 17 28. 24 34. 12 1 8. 28 15. 18 21. 71 0. 14 20. 53 10. 01 - - 0. 08 - - - 0. 37 0. 29 0. 43 2. 83 3. 76 2 0. 12 0. 89 0. 19 22. 06 - 3. 51 6. 28 13. 21 17. 46 16. 96 17. 23 23.48 9. 22 11. 02 5. 48 2.64 1.02 3 3. 91 6. 62 20. 62 0. 14 7. 54 1. 63 - - - - - - 0. 62 0. 32 - 0. 22 0. 46 4 0. 08 - - 2. 24 0. 07 2. 14 6. 40 8. 64 20. 76 15. 79 13. 62 11.00 8. 92 4. 57 5. 22 1. 17 0. 51 5 11. 20 12. 64 16. 10 1. 05 5. 38 8. 40 1.05 - - - - 0. 51 0. 70 3. 07 9. 53 13. 28 10. 25 6 0. 09 0. 72 - 7. 76 0. 07 4. 96 4. 71 17. 82 15. 53 11.46 9. 72 9. 13 6. 65 7. 78 5. 13 2. 87 0. 70 7 12. 78 12. 53 6. 26 4. 60 5. 78 12. 63 7. 02 - 0. 08 - 0. 88 1. 52 0. 72 5. 88 5. 25 14. 79 13. 60 8 13. 36 5. 55 8. 22 14. 63 8. 87 6. 28 13. 84 0. 28 — 3. 08 9. 71 9. 94 11. 85 3.60 9. 76 11. 14 10. 06 9 0. 22 - 0. 38 3. 62 - 3. 06 8. 28 15. 66 6. 93 15.44 9. 60 5.41 6. 30 3. 43 1. 44 0. 92 0. 28 10 9. 47 11. 94 5. 57 0. 40 6. 42 1. 90 — - 0. 27 0. 18 - 0. 65 0. 13 0. 42 0. 67 0. 74 1. 40 11 7. 27 3. 72 2. 49 0. 14 10. 15 8. 07 - - - - - 0. 22 - 0. 11 1. 02 0. 50 0. 64 12 0. 35 1. 39 1. 02 2. 37 0. 43 3. 27 4. 51 1.45 0. 27 0. 96 1. 21 2. 79 11. 07 9. 90 5. 25 6. 97 7. 24 13 9. 21 4. 07 3. 13 - 8. 88 4. 06 0. 12 - 0. 38 — - - - 0. 03 0. 34 0. 45 - 14 6. 38 8. 42 2. 43 1. 59 4. 31 5. 96 0.27 - — - - — - 0. 07 1.86 2. 56 1.82 15 2. 40 1. 33 1. 65 - 7. 54 0. 28 0. 26 - - - - - 0. 41 0. 11 0. 82 0. 09 0. 18 16 4. 38 1. 15 3. 20 6. 07 1. 38 3. 24 3. 28 0. 56 0. 20 3. 14 1. 21 7. 96 7. 92 5. 61 4. 18 1. 37 1. 46 17 5. 92 3. 90 4. 79 0. 28 6. 93 1. 35 0. 26 0. 39 0. 20 0. 18 1. 56 0. 65 0. 50 0. 74 0. 67 1. 55 2. 26 18 - 0. 05 - 6. 07 - 0. 81 5. 99 8. 88 8. 80 6. 57 9. 43 5. 11 4. 52 4. 19 2. 26 0. 51 0. 13 19 - 0. 26 - - - - 0. 79 - - - - - - 0. 20 7. 68 1. 31 3. 99 20 2. 63 6. 16 1. 65 - 0. 65 1. 63 - - - - - - 0. 13 0. 07 0. 29 0. 30 1.06 21 0. 06 - - 3. 55 - 1. 62 2. 85 2. 90 8. 01 1. 32 1.02 1. 31 1. 11 1. 23 0. 67 0. 16 0. 13 22 - - - 1. 84 - 0. 54 0. 94 7. 45 2. 28 3. 55 2. 92 1. 39 2. 74 1. 97 1. 26 0. 37 0. 82 23 - - - 0. 93 0. 22 - 1. 20 4. 41 3. 78 2. 96 7. 08 6. 22 2. 18 0. 85 1. 31 0. 59 0. 18 24 0. 46 0. 21 0. 51 4. 48 0. 87 3. 37 3. 90 4. 41 2. 36 5. 32 3. 07 2. 98 2. 29 2. 70 1. 39 0. 68 0. 33 25 - 0. 15 - 5. 13 - 0. 26 1. 20 1. 05 0. 76 0. 41 - 2. 90 0. 45 0. 30 - - - 26 - - - - - 0. 91 4. 55 0. 08 0. 28 7. 70 - 2. 09 1. 19 3. 74 3.46 - - 27 0. 61 - - 3. 30 2. 20 - 2. 71 1. 45 - 0. 55 0. 34 - 2. 12 3. 60 1. 12 1. 02 1. 02 28 - - - 0. 75 - 0. 26 1. 47 3. 13 4. 60 0. 18 1. 36 0. 43 1. 09 1. 27 0. 97 0. 32 0. 10 29 - - - 0. 80 - 1. 35 2. 81 2. 92 0. 46 0. 94 4. 46 - 1. 09 1. 22 1. 21 0. 33 0. 36 30 - - - - - - 0. 14 - - - - - - 1. 32 - - 2. 06 31 - - - 0. 26 - - 0. 41 1. 17 0. 54 0. 55 2. 10 - 1. 27 1. 69 0. 38 0. 28 - 32 0. 06 - - 1. 32 0. 07 0. 54 - 1. 13 0. 92 0. 88 1. 22 2. 47 0. 78 1. 24 0. 58 0. 34 - 33 - - - 1. 81 0. 07 - - - 0. 08 - - 0. 51 - - - - - 34 - - - - - - 0. 50 - 1. 82 0. 26 1. 22 0. 43 0. 51 - - 0. 12 - 35 - - - 0. 53 - 0. 54 0. 41 0. 36 1. 50 0. 18 1.02 0. 22 0. 99 0. 99 0. 30 - - 36 0. 12 0. 05 0. 07 0. 40 0. 30 - 0. 52 1. 06 0. 18 0. 22 - 0. 65 1. 38 0. 17 0. 19 0. 08 - 37 - - - - - - - 1. 34 0. 78 - - - - 0. 14 0. 24 - - 38 - - - - - - 0. 26 0. 22 0. 38 - - - — — - 1. 06 - 39 0. 26 - - 0. 26 0. 07 - - - - - — - — 0. 83 0. 44 - - 40 0. 06 - - - 0. 80 - - — - - — - — — - - - 41 - - - - - - — - — — - - - - 0. 48 0. 62 - 42 - - - - - 0. 52 — — — — — - - - - - - 43 - - - - - - — — — — - — 0. 13 — 0. 34 - - 44 - — - - - - — — - — - - 0. 22 0. 03 0. 14 - - 45 - 0. 17 - - — - - - — — - - 1.01 - - - - 46 - - - - - - 0. 12 - - 0. 14 - - - - 0. 05 - - 47 - - - - 0. 14 - — — — — — - - - - - - 48 - - - 0. 14 - - - - - - - - - - - 0. 08 - 49 - - - - 0. 07 - - - - - - - - - - - - 50 0. 02 ' " " 51 27 23 18 32 27 30 32 24 29 26 21 25 34 38 38 35 28 22 although Berthelsen (1962:16) considered the zooecia lacking opesiules in such specimens to have lost them in fossilization. Morphotypes v and vi (Plates 4:1; 10-12) : Zooecia of types v and vi have not been separated in this study. There are at present no known criteria in zooecial morphology that consistently distinguish between fully developed zooids of these two morphotypes. Even though the frontal shield and ascus develop in different relationship to the frontal membrane, as discussed above, the membrane bounding the frontal surface, the frontal shield, and the underlying ascus are arranged similarly in fully developed zooids of the two types (Figure 7). Harmer (1957:645, 662) assumed that zooids in which the proximal part of the oper¬ culum is continuous with the membranous floor of the ascus developed by the umbonuloid pattern (morpho- type v), whereas those which developed in the leprali- oid manner (morphotype vi) should have the proximal margin of the operculum contiguous with the portion of the frontal shield forming the proximal lip of the primary orifice. As pointed out by Banta (1970:52), however, the relation of the membranous floor of the ascus to the operculum is identical in the two modes of development. Therefore, the structure of the zooe¬ cial orifice appears not to be a criterion for separating the two morphotypes. It has been assumed that ascophorans having frontal shields developed by the two major patterns represent two or more distinct phyletic lineages (see, for ex¬ ample, Ryland, 1967a:348). Because the two types cannot now be distinguished in fossils, however, the phylogenetic significance of the difference in develop¬ ment is unknown (Boardman and Cheetham, 1969: 229). The difference in development does not seem to involve functional differences in the ability to calcify frontally. Both kinds of frontal shields appear to be susceptible of long-sustained ontogenetic thickening and concomitant merging across the frontal edges of lateral walls of contiguous zooids. Ascophoran zooecia have convex, usually heavily calcified frontal lamellae commonly lacking zooecial boundaries in zoarial fragments inferred to represent later stages of zooid ontogeny. The frontal shield dis¬ plays perforations along its margins (Plate 10:4) or scattered over its surface (Plate 11:2-4) or on its midline proximal to the orifice (Plates 10:2-4; 11: 2-4). Except by similarity of position in living as¬ cophorans, there is no known way to distinguish be¬ tween perforations which in life were covered with SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY membrane (areolae, frontal pores) and those which represent true openings (ascopores, spiramina). In transverse and longitudinal sections (Plate 12), the frontal shield appears convex and much thicker than the basal and lateral walls. In most species, the trans¬ verse wall is also appreciably thinner than the frontal shield, but in Columnotheca cribrosa Marsson (Plate 12:1-2), it is about as thick as the frontal shield and is perforated with similar tube-like pores. In most species, distinct boundaries are discernible between lateral walls but disappear frontally, suggesting merg¬ ing of the epifrontal membranes of adjacent zooids. Adventitious avicularia are commonly present on the frontal shield, especially in the vicinity of the orifice. The distal portion of the shield is generally raised around the orifice to form a mucro (Plate 4:1) or a peristome (Plates 11-12). Taxonomic Significance of Morphologic Variation Even though the objective of this study has been to understand the functional morphology and distribution of Danian cheilostomes rather than their taxonomy, it has of necessity involved the recognition of operational taxonomic units. The basic units of this study are groups of fragmentary specimens inferred to have been parts of approximately contemporaneous, sympatric, morphologically similar colonies. The morphologic dif¬ ferences between any two of these groups of specimens (phena of Mayr, 1969) thus might reflect either intra¬ population variation or taxonomic differences. For nomenclatural convenience and in accordance with taxonomic practice in most previous studies of Danian cheilostomes, these operational units, or phena, are re¬ garded as separate species (Table 1) except where there is evidence of intergradation among coeval as¬ semblages. Intergrading phena are here united in single operational taxa, rather than being regarded as conspecific subspecies as they have in some previous studies. In the absence of evidence for the positions of these operational species in evolutionary lineages, it would be presumptuous to propose formal revisions of the existing taxonomy on the basis of the morpho¬ logic variation observed in this material. Within each of the operational species, those zooecia inferred to represent autozooids belong to a single morphotype. At the distal end of a zoarial fragment preserving the growing extremity of a colony, zooecia may lack or have only incomplete frontal structures NUMBER 6 23 (Plate 8:2). In those few zoarial fragments found to preserve the primary zone of astogenetic change (Plates 1:4, 6:1), all zooecia, including the ancestrula, appear to belong to the same morphotype. Variations in details of morphotype structure, however, are dis¬ cernible among the zooecia in many zoarial fragments studied. For example, Floridina gothica (Plate 7:2-3) has, within the same zoarial fragment, some zooecia with separate opesiules and others with opesiular in¬ dentations confluent with the opesia. In APicropora henningiana the opesiules are small, subcircular per¬ forations near the opesia on some specimens (Plate 6:1—2) and more elongate slits placed nearer midlength on others (Plate 6:3), but intermediates between these two extremes are common. A similar gradational series characterizes Pithodella cincta (Plate 2:2-4) in which the relative development of spines, cryptocyst, and avicularium are highly variable but overlapping in different specimens. In still other operational taxa, variations in details of zooecial morphology are less distinctly gradational or overlapping, and a number of specific separations might be justified in each. How¬ ever, the material studied seemed inadequate for making distinctions in the following complexes (Table 1): Callopora spp., Floridina spp., Pliophloea spp., and “Cellepora” damensis Voigt + spp. Zoarial form also appears to be variable within some operational species, but the evidence is less definite than that for zooecial variation. The small size of the fragments might account for the lack of specimens clearly showing more than one colony form. Specimens which show differences in zoarial form but which lack correlated differences in zooecial structure have been included in the same operational species. In this cate¬ gory fall the following (Table 1) : Puncturiella sculpta (d’Orbigny), Semiescharinella complanata (d’Orbi- gny) ; Micropora hennigiana + spp., Membraniporidra declivis (Marsson), and Psilosecos angustidens (Lev- insen). The last species seems to occur in only one form (membraniporiform) in the material studied here, but has previously been found only in another (eschariform). Paleoenvironments of Danian Cheilostomes Bryozoan Mounds Although Bryozoa occur in southern Scandinavia in several kinds of limestone of Danian age, they are typically associated with mound-like sedimentary Figure 10.—Numbers of species of cheilostome Bryozoa in mound and non-mound facies of Danian (Zones B—D) ; data from Berthelsen (1962:242-255); zonal classification after 0dum (1926:217). For comparability with other localities, the number of species in samples from Limhamn Mound II-Ni has been adjusted to the same morphologic basis, i.e., each dis¬ tinct zooecial or zoarial variant has been counted as a separate species. Therefore, the number of species is greater here than in Table 1 or the analyses (Figures 20—29) based on it. structures variously referred to in the literature as reefs, bioherms, or banks. Among the localities in Den¬ mark from which Danian cheilostomes were studied by Berthelsen (1962), those having evident mound structure on the average yielded more species than those of the same age showing only evenly bedded limestone (Figure 10). The abundance of cheilostomes at these localities, however, is highly variable but aver¬ ages about the same in mound and non-mound sedi¬ ments (Berthelsen 1962, Tables 3, 4). The diversity of mound faunas therefore suggests multiple adapta¬ tions within the geographic confines of a single mound. Typical mounds having well-defined boundaries oc¬ cur in the lower and middle parts of the Danian (Fig¬ ure 11) and are best exposed along Stevns Klint, Denmark, and in the quarry at Limhamn, Sweden. The Stevns Klint structures and larger mounds with less distinct limits at Fakse, Denmark, have been de¬ scribed and illustrated in a number of papers sum¬ marized by Rosenkrantz and Rasmussen (1960). Those at Limhamn were described and illustrated by Brotzen (1959). Stratigraphic correlations between the mound sedi¬ ments of Denmark and Sweden have not been com- 24 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY STEVNS KLINT, FAKSE, & HERF0LGE , DENMARK LIMHAMN, SWEDEN EVENLY BEDDED LIMESTONE WITH CHERT IN NODULES AND BEDS; BRYOZOA COMMON ; DOMINANT CHEILOSTOMES GROUPS 1-3. FAKSE MOUND W, N— (section reversed to conform with directions in diagram) (top of exposure) mound hn 3 ^—w— (local unconformity within danian) o LIMESTONE MOUNDS: - CHERT AND NODULAR BEDDED LS. SAMPLING POINT EH' CORE FACIES i IN SOME 0IOHERMS HAVING COLONIAL CORALS IN GROWTH POSITION, BEOOING 8 CHERT RARE, OCTOCORALS USUALLY PRESENT, CORALS COMMON TO ABUNDANT, BRYOZOA RARE TO COMMON, DOMINANT CHEILOSTOMES GROUP 4o I £ TRANSITIONAL FACIES , BEDDING a CHERT IRREGULAR, OCTOCORALS ABUNDANT; ' 2 CORALS RARE OR ABSENT, BRYOZOA ABUNDANT, CHEILOSTOMES OF MIXED GROUPS 1-4 I q FLANK FACIES; BEDDING a CHERT REGULAR , ABUNDANT; OCTOCORALS RARE , 1 D CORALS ABSENT; BRYOZOA ABUNDANT, DOMINANT CHEILOSTOMES GROUPS 1-3 MOUND UN, ^—w— BOESDAL MOUND E, -NW— QUATERNARY LIMESTONE IN UNDULATING BEDS WITH APPARENT INITIAL DIPS <10°; CHERT NODULES ABUNDANT;CORALS ABSENT; OCTOCORALS RARE ; BRYOZOA ABUNDANT; DOMINANT CHEILOSTOMES GROUPS 1-3. CERITHIUM LIMESTONE FISH CLAY HARD LIMESTONE HORIZONTAL AND VERTICAL SCALE 0 5 10 I 1 ■ ■ 1 I 1 1 1 ■ I H0JERUP MOUND M, 500/i >74/i Total 1 8. 7 100. 0 36. 6 77.4 236.5 2 5. 6 49. 2 33. 8 82. 1 250. 9 3 6. 5 13. 0 40. 5 96. 2 293. 2 4 12. 3 22. 0 61. 4 112. 7 308. 2 5 5. 9 23. 2 29. 9 74. 2 170. 7 6 6. 7 25. 3 39. 3 102. 8 329. 8 7 19. 1 50. 0 55. 9 127. 5 436. 3 9 13. 6 100. 0 12. 2 58. 8 281. 3 10 38. 2 100. 0 33. 4 86. 5 140. 8 11 25.6 100. 0 29. 4 87. 2 391. 8 12 12. 5 24. 8 82. 6 128. 5 288. 2 13 9. 1 25. 1 59. 5 119. 1 298. 6 14 28. 0 52. 6 64. 3 110. 8 358. 8 15 30. 9 44. 9 63. 7 109. 5 379. 2 16 12. 3 41. 8 81. 2 146. 2 354. 2 17 15. 1 37. 1 85. 5 142. 6 354. 2 18 12. 2 51. 7 66. 1 115. 1 322. 6 Table 4 .—Major biotic constituents in biofacies analysis of Mound II-Nj. Symbol Category Description CHEILO. .. . Cheilostome bryozoans. . . (Data analyzed separately.) CYCLO. Cyclostome bryozoans. . . . Flat, crust-like; slender, twig-like; and massive, mushroom-like colony fragments. OCTO. Octocorals. Smooth to tuberculate, rod-like fragments of axes, with or without calycal pits; small, fusiform, tuberculate spicules. SCLER.Scleractinian corals. Slender, branching colony fragments and cup-shaped solitary forms; plate-like fragments of septa, with or without perforations. ECHIN. Echinoderms. Mostly disarticulated ossicles of echinoids, asteroids, ophiuroids, and crinoids. SERP. Serpulids. Fragments of irregularly coiled, cylindrical tubes, many showing attachment scars or still adherent to other particles. RETI. Calcareous sponges. Irregular crusts to sub-spheroidal masses, all with tuberculate surfaces; detached or adherent to other particles. OTH.Other constituents (in Brachiopod valves, mostly fragmentary, usually punctate and plicate. minor amounts). Mollusk shell fragments, including oyster-like and scallop-like forms. Calcareous foraminiferal tests. Bairdiid ostracode valves. Lepadomorph barnacle valves. NUMBER 6 27 Figure 12.—Lithology and biofacies of Mound II-Ni, middle Daman, in the quarry of Skanska Cement A. B., Limhamn, Sweden. Location on north wall of quarry is shown on Plate 13:1. a. Cross section of mound, constructed in the field by tape, hand-level, and plumbline survey. Since the mound was measured and sampled in 1964-65, quarrying has removed it completely. Location of samples 1-18 is shown by crosses. Biofacies Ai, A 2 , and B were delineated by cluster and principal components analysis (see Figures 18-19) of abundances of eight major biotic con¬ stituents. Nodular bedding is shown by dashed lines. Chert, indicated in black, was not mapped in the upper third of the mound, but presumably was distributed as shown in Figure 11. b, Grain size (as dry-weight percentage of sample) obtained by wet-sieving through 74-/* (200- mesh) screen and dry-sieving the residue through a set of 2000, 1000, 500, 250, and 125 p screens after the larger lumps which would not disaggregate were picked out by hand and subtracted from the original weight. Disaggregation was effected by washing in water, supple¬ mented, where necessary, by alternately freezing and thawing. c , Moisture loss (as dry-weight percentage of sample) on heating for 48 hours at 102° C (treatment empirically determined to give constant weight). On the assumption that the samples (all from below sea level) were water-saturated when collected, water content is a kind of measure of effective porosity. Sample 8, having a low water content, was indurated. d. Non-carbonate content (as dry-weight percentage of sample) of eight samples (1, 5, 9-11, 14, 18) determined by digestion of 2 g of powered sample (with no visible chert) in 1: 1 hydrochloric acid. position suggested by him to be between the lower and middle Danian of Denmark, as mentioned above. The species of cheilostome Bryozoa identified from it (Table 1) compare closely with those listed by Berthel- sen (1962) from the middle Danian of eastern Den¬ mark (Zone C). One species in the Mound, Pelmato- pora? daniensis Voigt, suggests a younger age, but none has previously been found only in older rocks. The number of species in the Mound also compares favorably with the diversity of other middle Danian localities, if adjusted as in Figure 10. This diversity is higher, in general, than those of lower Danian local¬ ities. Therefore, the fauna of the Mound can be con¬ sidered to be consistent with a middle Danian age, and there is no suggestion that it might be older than the middle Danian faunas of Denmark. Mound II-Nj is slightly asymmetrical, with higher apparent" dips (measured up to 24°) on its eastern flank and with the massive core limestone nearer its eastern extremity. The core limestone reaches a maxi¬ mum thickness of at least 6 m (the lower boundary of the mound was covered with talus at the time measurements were made) and is gradationally over- lain by about 7 m of bedded limestone. Colonial scleractinians were found only as fragments out of growth position, although coral masses in growth posi¬ tion might have lain outside the exposed cross section. Nodules and beds of chert were numerous on the 28 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 100 n MAESTRICHTIAN 80 60 40 20 i - f ■ V I 2 0.5 0.12 LOWER DANIAN Figure 13. —Grain size (as dry-weight percentage of sample) of limestone samples from Maestrichtian and Danian of Sweden and Denmark (size ranges were determined as explained in Figure 12b) : a, Uppermost of three similar samples from upper 20 m of Maestrichtian, north wall of Limhamn Quarry, b, Middle of three similar samples from upper 1 m of lower Danian in north wall of Limhamn Quarry, c, Southeast part of core, Mound Ei, Boesdal. d, Southeast flank of Mound Ei. e, Sample 10, core of Mound II-Ni, Limhamn. f, Sample 9, core of Mound II-Ni. g, Sample 17, flank of Mound II-Ni. flanks of Mound II-N 1} and chert nodules also occurred in the core. For the most part, the chert appeared devoid of fossils, but in some places still- calcareous echinoderm ossicles and other skeletal ma¬ terial could be observed. Some of these skeletal elements were observed to lie partly in the chert and partly in the adjacent limestone. In a few of the lime¬ stone samples taken some distance from the obvious chert nodules or beds, siliceous steinkems of Bryozoa, in some cases with the skeleton also silicified, were found. Most of the samples were chert-free and, with the exception of one (Sample 8, Figure 12), were readily disaggregated. The limestone composing Mound II-Nx has variable grain size, but the high proportion of fine particles shown by all samples (Figure 12 b) characterizes them as calcilutites (see Plates 16-17). In general, there is a higher percentage of material finer than 500 n in this mound than in a lower Danian mound in Boesdal Quarry on Stevns Klint (Figure 13; Berthelsen 1962, Table 2). However, the ranges of grain size in the Boesdal and Limhamn mounds overlap. Maestrichtian and lower Danian samples from Limhamn (Figure 13) are also calcilutites, but some middle and upper Danian sediments analyzed by Berthelsen (1962, Table 2) are calcarenites with as little as 7.4 percent material finer than 500 /*. The Danian calcilutites from Limhamn and Stevns Klint differ from those of Maestrichtian age at Lim¬ hamn in being coarser grained and in showing more than one modal class in the particle sizes analyzed (Figure 13). This composite character is also reflected in the cumulative particle-size distribution (Figure 14a) and results from the occurrence in each sample of several skeletal constituents each having its distinc¬ tive size-frequency distribution (Figure 15). The sedi¬ ment thus is very poorly sorted. The distribution of grain size across Mound II-Ni (Figures 12 b, 13) suggests that the bedded flank lime¬ stones are approximately uniform, whereas the core limestone may be either coarser or finer. This pattern appears to be the result of differences in the distribu¬ tions of the biotic constituents and thus not the product of physical sorting. Insoluble residues, consisting of what on visual in¬ spection appears to be chiefly clay, silt-size quartz, mica, and glauconite in the samples of essentially chert- free limestone, arc distributed in conformity with NUMBER 6 29 SIEVE SIZE U) Figure 14.—Distribution of grain size and unidentified par¬ ticles in Sample 17, Mound II-Ni (cf, coarse fraction; ff, fine fraction; mx, matrix) : a, Cumulative grain size by dry- weight percentage of sample, b, Unidentified and composite grains as dry-weight percentage of fraction of sample be¬ tween each pair of grain sizes. mound structure (Figure 12 d), that is, with higher per¬ centages on the flanks than in the core. Concentration of this material in quantities up to 4J4 percent of dry weight on the flanks suggests that, if it was available at the same rate over the area of the whole mound, the core limestone accumulated more rapidly than that on the flanks. Also, because the “effective porosity” of the samples (Figure 12c) appears to be correlated with gr ain size rather than strictly with position within the mound, the structure of the mound appears not to be the result of differential compaction. As the identifiable major biotic constituents of each sample are concentrated in the coarse fraction (>500 /a) of the limestone (Figure 15), biofacies analysis of Mound II-Ni was limited to this material. Only a few constituents, such as foraminifers and oc- SIEVE SIZE (ju) Figure 15.—Distribution of major biotic constituents with grain size in Sample 17, Mound II-Ni. Coarse fraction (cf) percentages were calculated from particle counts of aliquots of each of the size fractions. Unidentified and composite grains in each size fraction of Sample 17 are shown in Fig¬ ure 14. (ff, fine fraction; mx, matrix; other symbols explained in Table 4, on p. 26.) tocorals (spicules), could be identified in significant amounts in the finer fraction. The coarsest material (>2000 fi) was analyzed in its entirety, but, to reduce sorting time, smaller aliquots were used for the 1000- 2000 and the 500-1000 ^ material. Very small amounts of these size grades were found to give weight per¬ centages of the constituents which, by chi-square tests based on the number of particles in the aliquot, are significantly different from the whole-fraction values in an appreciable number of cases (Table 2). Therefore, aliquots of 10 g of the 1000-2000 /* fraction and 5 g of the 500-1000 ^ fraction were used to minimize this risk. The weight of material processed for each sample and the aliquot percent of that weight in which the biotic constituents were analyzed are shown in Table 3. The biotic constituents of Mound II-Ni are divisi- 419-995 0 - 71-3 30 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY mm mm* UNIDENTIFIED a COMPOSITE GRAINS TRAN-I I SI- I I TION TRANSI TION TRANSI FLANK FLANK FLANK OCTO OCTO SCLER Figure 16. —Distribution of major biotic constituents in coarse fraction (>500/t) of limestones from Mound II-Ni. In each diagram, the dry-weight percentages of coarse fraction add up to 100 percent. The lower diagram was constructed without the unidentified and composite grains which were included in the upper diagram. (Symbols explained in Table 4; biofacies classification shown in Figures 18 and 19; location of samples indicated in Figure 12). ble into eight major categories (Table 4), which can be identified in significant amounts in the coarse frac¬ tions of the samples. Most constituents occur generally in fragments (Plates 16-17), but indications of abra¬ sion (Plate 8:1) are rare. Of the constituents, cyclo- stome and cheilostome bryozoans and octocorals are dominant (Figure 16), together making up at least half of the coarse fraction (or more than three-fourths of the identifiable constituents) of every sample. In general, the two groups of bryozoans vary in abundance NUMBER 6 31 Figure 17.—R-mode (constituent-by-constituent) dendro¬ gram of relationships among eight biotic constituents of coarse fraction (>500/*) of the 17 samples from Mound II-Ni and influence of principal components on clustering. Weighted- pair-group-method clustering is based on correlation coeffi¬ cients calculated from weight percentages transformed by the arcsine-square-root method recommended by Seal (1964). Unidentified and composite grains were excluded in calcula¬ tion of percentages. Dashed lines indicate values of correla¬ tion coefficient at 10-percent significance level. Ci through Cu are the first five factors from Table 5. Identifying and uniting influences are indicated by shading; opposing influences are shown by arrows; abscissa is product-moment correlation co¬ efficient. (Symbols for biotic constituents explained in Ta¬ ble 4.) in the same way and in opposition to the octocorals. To the extent that the compositional data approximate a two-constituent, closed system, the correlation be¬ tween bryozoans and octocorals can be expected to ap¬ proach — 1 (Chayes 1960). To investigate this rela¬ tionship and others not apparent directly from the abundance variation (Figure 16), the weight- percentage abundances in the 17 samples from Mound II-Ni were submitted to multivariate analysis. To mini¬ mize the constant-sum constraint, percentages were transformed (Seal 1964). Both constituent-by-constitu¬ ent (R-mode) and sample-by-sample (Q-mode) analyses were made. Relationships among the eight biotic constituents were investigated by cluster analysis and principal com¬ ponents analysis of the array of product-moment cor¬ relation coefficients among all pairs of constituents, the R-mode correlation matrix. Cluster analysis (Sokal and Sneath 1963) arranges the eight constituents in three groups (Figure 17). Within each group, the con¬ stituents are positively correlated at about the 10- percent significance level or higher, indicating that their abundances vary together. Cheilostome and cyclostome bryozoans form the most tightly knit group (correlation 0.91), and echinoderms and calcareous sponges the least (correlation 0.39). The remaining four constituents are correlated with each other at values of 0.44 to 0.76. The relationships among the three main clusters of constituents are not as clear as those within them. The echinoderm-sponge group clusters with the group (coral cluster) composed of octocorals, scleractinians, serpulids, and the category “other constituents” at a level not significantly differ¬ ent from zero; the abundances of these two groups therefore appear to vary independently. The cheilo- stome-cyclostome group shows a negative relationship, significant at about the 10-percent level, to one or both of the other two; thus, bryozoan abundances vary in Table 5. —Principal components analysis of variation in eight major biotic constituents of coarse fraction (> 500f) of 17 samples from Mound II-Nj based on correlation coefficients calculated from arcsine-square-root transformed weight percentages. r—±0.41 significant at a=0.10. Symbols for constituents are explained in Table 4. Factor Eigen¬ value Variance Eigenvector (factor loadings ) % tot. Cum. CHEILO CYCLO OCTO SCLER ECHIN SERP RETI OTH c, 4. 5582 57.0 57.0 -.89 -.96 +. 89 +. 79 +. 25 +. 62 +.36 +. 92 C 2 1. 3630 17.0 74.0 +. 22 +.21 -. 31 +. 14 +.72 +. 22 +. 76 -.03 C 3 0. 8109 10. 1 84. 2 -f. 34 +.09 -. 18 +. 37 -. 55 +.43 +. 14 +.08 C 4 0. 6284 7. 9 92.0 +.06 +.02 -.02 -. 17 +. 26 +.58 -.42 -.06 C 5 0. 4003 5.0 97.0 +.08 -. 01 -.22 +.40 +. 19 -. 19 -. 30 +. 13 C 6 0. 1708 2. 1 99. 1 -.02 +. 10 -. 11 -. 16 -.02 -.01 0.00 +. 35 c 7 0. 0613 0. 8 99. 9 +. 19 -. 10 +. 08 -.04 +. 03 -.03 -.06 +. 05 C 8 0. 0071 0. I 100.0 +. 01 +. 06 +.05 +. 02 +.01 0. 00 +.01 +.01 Communality (Factors Cj to Cj) 0. 97 0. 97 0. 97 0. 97 0. 99 0. 99 1. 00 0. 88 32 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY opposition to those of echinoderms and sponges, or the other four constituents, or all six constituents com¬ bined. These relationships are brought out further by prin¬ cipal-components analysis (Seal 1964), which relates the eight intercorrelated constituents to eight uncor¬ related (orthogonal) factors (Table 5), of which the first five explain the principal influences on the cluster dendrogram. These factors are mathematical abstrac¬ tions which together account for nearly all (97 per¬ cent) of the total variation and also for nearly all of the variation in each constituent (as shown by the very high communalities calculated from the loadings on the first five factors). The nature of the first five factors and their resulting influence on clustering are sug¬ gested by their loadings (significant at about the 10 percent level) on the constituents. (Note that the sixth through eighth factors lack significant loadings as a result of the low eigenvalues attached to each of them.) From the loadings, three kinds of influences are distin¬ guishable: (1) “identifying influences” are shown by factors (e. g., C 5 ) having a significant loading (either positive or negative) on only one constituent; (2) “uniting influences” are shown by factors (e. g., C 2 ) having significant loadings of the same sign on two or more constituents; (3) “opposing influences” are shown by factors (e. g., C 4 ) having significant loadings of opposite sign on two or more constituents. In these data, the first factor, accounting for more than half the total variance, is both a uniting and an opposing influence and thus explains the general shape of the dendrogram. The significant opposition is between the cheilostome and cyclostome bryozoans on the one hand and the octocorals, scleractinians, seipulids, and the category “other constituents” (the coral cluster) on the other. Cluster analysis, which forces the echino¬ derms and calcareous sponges into the opposing rela¬ tionship with the bryozoans, is thus seen to be an over- generalization. The uniting influence of the first fac¬ tor explains the bryozoan cluster and the coral cluster of the dendrogram. The second factor, accounting for as much variance as the next two put together, unites echinoderms and calcareous sponges at a more signifi¬ cant level than suggested by the dendrogram. The third and fourth factors appear to be mainly a contrast be¬ tween the echinoderm-sponge cluster and the serpulids: The fifth factor indicates that scleractinians may show some independence of variation, although the signifi¬ cance of the loading is borderline. Figure 18.—Q-mode (sample-by-sample) dendrograms of relationships among 17 samples from Mound II-Ni on the basis of eight biotic constituents. Weighted-pair-group-method clustering is based on correlation coefficients (right) and distance function (left) calculated from arcsine-square-root-transformed weight percentages. The transformed data were standardized by rows (Sokal and Sneath 1963) in the calculation of both correlation coefficients and distances. The biofacies are constituted as shown in Figure 19. NUMBER 6 33 Relationships among samples from Mound II-Nx were investigated by cluster analysis and by ordination. The cluster analysis, based on two different Q-mode matrices of association coefficients, resulted in slightly differing dendrograms (Figure 18), which are alike, however, in arranging the samples in two major clus¬ ters. The dendrogram based on the correlation coef¬ ficient groups together samples from both flanks (17 and 18 from the east flank; 1-3, 5, and 6 from the west flank) and, within the cluster consisting of samples principally from the core region (4, 7-16), distin¬ guishes two subclusters. The dendrogram based on the distance function (Sokal and Sneath 1963:147) places samples from the east flank (17 and 18) and one from the west flank (6) with those in the core region. In¬ clusion of these samples in the core cluster seems to loosen the clustering of samples 9 and 10 with the others in the core and thus to imply more complex relationships. Consequently, biofacies suggested by the correlation dendrogram are more readily mappable than those indicated by the distance dendrogram. A compromise between the two Q-mode dendro¬ grams in Figure 18 is suggested by ordination of the samples relative to the factors obtained by principal components analysis of the R-mode matrix. This tech¬ nique (Figure 19) plots the positions of the samples with respect to the orthogonal factor axes thus giving a geometric representation of the relationships among samples. For an almost undistorted representation, all five significant factors (Table 5) are required, but it is not possible to represent more than three factors in one diagram. This is not a serious limitation; the first two factors alone can be expected to approximate these relationships closely, for they account for 74 percent of the total variation and include the heaviest loadings on the three major constituent clusters—bryozoans, the coral cluster, and echinoderms-sponges (Figure 17). The ordination diagram (Figure 19) shows generally the same relationships that the dendrograms do; the flank samples differ distinctly from those in the core region, and the core samples form two broadly over¬ lapping subgroups. Further, it suggests that samples 6 and 17 are close to the flank cluster, whereas 18 is nearer one of the core subclusters. The facies assign¬ ments shown in Figure 19 are used in the following analysis of cheilostome species abundances. The nature of the biofacies suggested in Figure 19 is brought out in Table 6, which combines the R-mode and Q-mode results. The greatest contrast in composi¬ tion is between Biofacies B, in general occupying the flanks of Mound II-N 1} and Biofacies A x occurring in the core (Figure 12). Biofacies A 2 is transitional be¬ tween the two but seemingly has more in common with Ax. Biofacies Ax and A 2 are characterized by predomi¬ nance of the coral group (octocorals, scleractinians, serpulids, and the category “other constituents”) over bryozoans (cyclostomes and cheilostomes). Octocorals are the single most abundant constituent of these re¬ lated facies, and bryozoans are second in abundance in both. Biofacies Ax shows the nearest approach to equal proportions of the constituents, and scleractinians (making up more than 3 percent of the coarse-fraction biota), serpulids, echinoderms, sponges, and the cate¬ gory “other constituents” have their maxima here. Bio¬ facies A 2 shows less than 3 percent scleractinians and intermediate proportions of bryozoans, serpulids, and _*OCTO- SCLER- + ! OTH-SERP Figure 19.—Ordination diagram: distribu¬ tion of samples from Mound II-Ni with re¬ spect to first two principal components (factors Ci and C 2 , Table 5) of major biotic constituents. Coordinates of samples were obtained by pre-multiplying by the eigen¬ vectors the matrix of arcsine-square-root- transformed abundances. Lines connecting samples are a compromise between the two dendrograms obtained by cluster analysis and shown in Figure 18. 34 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Table 6. —Distribution of biotic constituents (as mean weight percentages for samples clustered as in Figures 18 and 19) in biofacies of Mound II-Nj. Constituents clustered as in Figure 17. Q-mode biofacies clusters R-mode clusters of CF constituents CYCLO CHEILO OCTO SCLER OTH SERP ECHIN RETI fd-3,5) 66.0 23. 1 1. 6 0.4 1.0 0.6 7.2 0. 5 R B 1 [(6, 17) 41.2 12. 6 23.6 1. 0 1.4 1.4 18.2 0. 8 A [(4, 7, 12) 19.4 5. 5 64. 7 0. 9 2. 2 1.6 5. 6 0.3 Aj-i 1(15, 16, 18) 28.0 9.5 47. 1 1. 3 2. 1 2.8 8. 8 0.4 [(9) 24. 1 14. 7 31.5 9.6 2. 9 4. 9 10.8 1.4 Ai' (10) 9. 2 3.3 59. 5 15. 5 4.3 1. 7 5. 9 0.6 1(11, 13, 14) 20. 7 5. 8 44. 5 3.5 3.2 3.2 18. 3 0. 7 “other constituents.” Echinoderms and calcareous sponges tend to be at their minima. Biofacies B is characterized by cheilostomes and cy- clostomes which together make up more than 50 per¬ cent of the coarse-fraction biota. Constituents of the coral group have their minima in this facies, though octocorals are a significant constituent of some samples. Echinoderms and calcareous sponges show a second abundance peak in this facies. The distribution of these biofacies in Mound II-Ni (Figure 12) is asymmetrical and slightly more complex than the mound structure alone indicates. Biofacies A x is in the core, but slightly displaced eastward or toward the steeper flank relative to the lithology and the atti¬ tude of the bedding. Biofacies B occupies the principal parts of the flanks, proportionately more of the long, gentle western flank than the shorter, steeper eastern one. Biofacies A 2 is present on both sides of the mound between A x and B and also occurs as outliers within B on both flanks. The differences between biofacies are more distinct on the long western flank than on the short eastern one. Comparison with Other Mounds The general characteristics of the biofacies of Mound II-Ni at Limhamn were compared with visual esti¬ mates of biotic constituents of samples collected from other mounds in the Limhamn Quarry and in Den¬ mark. The suggested distributions of biofacies in these mounds (shown in Figure 11) are briefly described below. Mound Mi, at the top of the Maestrichtian exposed in the south end of the sea cliff just below Hojerup Church on Stevns Klint, Denmark, is a small structure apparently lacking a distinct core or transitional facies. Bryozoans appear to be slightly more common in the center of the mound than on its flanks, though nowhere are they as abundant as in the Danian. Octocorals and scleractinians were not observed. Mound Ei, in the lower Danian exposed in the northeast face of the Boesdal Quarry just behind the sea cliff on Stevns Klint, Denmark, is comparable in size, structure, and lithology to Mound II-Ni, but the core of Mound Ei appears to be comparable to the transitional facies (Biofacies A 2 ) of the Limhamn mound in its proportions of biotic constituents. Bryo¬ zoans are in general more abundant on both flanks and in the core than they are in Mound II-Ni. Octocorals are common, though not dominant, in samples from the mound center. Also in contrast to the Limhamn mound, cheilostomes are more abundant than cyclo- stomes in most samples. Mound II-N 3 , in the middle Danian exposed in the north wall of Limhamn Quarry above and to the west of Mound II-Ni, is also comparable to that Mound in size, structure, and lithology and in distribution of biofacies. The greatest difference is the more extensive core facies (Biofacies Ai), which includes numerous masses of colonial scleractinians in, or nearly in, growth position. As in Mounds II-Ni and E l3 how¬ ever, the core and transitional limestones grade upward into flank-type limestones. The whole upper part of Mound II-N 3 appears to have been truncated before the deposition of suprajacent mound sediments. Mound Wi, at about the middle of the Danian sec¬ tion exposed in the west end of the quarry at Fakse, Denmark, differs from all of the mounds described above in having the core facies (Biofacies A x ) domi¬ nant so that it appears to have overgrown not only its NUMBER 6 35 own flanks but also those of adjacent mounds. As a result, the boundaries of the mound are indefinite. Also, the core includes abundant colonial scleractinians nearly in growth position, and octocorals are abundant only in the much less extensive transitional biofacies. A richly diversified biota associated with the corals has been described at Fakse (summarized by Asgaard 1968:104), and it includes bryozoans, mollusks, brachiopods, serpulids, echinoderms, and burrows of sponges, crustaceans, and possibly algae (Asgaard 1968:117). The flank limestones are richer in bryo¬ zoans than those of mounds described above. Inferences About Mound Formation The biofacies study of Mound II-Ni at Limhamn and comparison with other mounds suggest some gen¬ eral relationships in the formation of these structures: (1) Differences in content of non-carbonate detritus and in character of bedding are distributed in con¬ formity with mound structure, whereas differences in porosity are less regularly distributed (correlated only with differences in grain size). The structure of the mounds therefore appears to be a primary depositional feature, rather than a product of post-depositional alteration. If this is so, the present structure of the mounds indicates their depositional configuration. (2) Although most biotic constituents of the mounds occur in all of the samples, differences in their abundances are distributed in conformity with mound structure, whereas differences in grain size are less regularly distributed. Therefore, as inferred by Hennig (1899:38, 39), the biotic constituents for the most part probably accumulated approximately where they grew, rather than having been mechanically sorted. If this is so, the biofacies of the mounds represent different paleoenvironments. (3) The proportions of fine-grained non-carbonate detritus and fine-grained carbonate sediment vary in¬ dependently of each other within a mound. The non¬ carbonate detritus must have been transported into the area of mound deposition, whereas the calcilutite could have been produced near the site of its accumu¬ lation. If the difference in distribution indicates a dif¬ ference in source, Hennig’s inference (1899:38, 39) that the calcilutite represents detritus from the activity of predation on the carbonate-producing benthos may be correct. (4) If the inferences in both (1) and (2) are cor¬ rect, then the paleoenvironmental differences between biofacies are correlated with depositional topography. In general, the mound core, having abundant octo¬ corals and subsidiary colonial scleractinians and bryozoans, probably stood higher than the flanks, domi¬ nated by bryozoans but in places including patches of octocorals. (5) Because the cores and flanks of each mound probably originated at about the same topographic level, the growth of some or all of the core- or transitional-facies constituents was probably initially responsible for any topographic differences between biofacies, rather than vice versa. If this is so, some or all of the core- or transitional-facies constituents are the essential element in mound formation. (6) Because both cheilostome and cyclostome Bryo- zoa have minimum abundances in the core facies, and because they are abundant in non-mound Danian limestones, Bryozoa were probably not the initiators of mounds in the Danian, although they may have had this role in the smaller Maestrichtian mounds. If this is so, octocorals or colonial scleractinians or combina¬ tions of these two constituents are the most likely pre¬ requisite for mound formation, and, as the proportion of scleractinians increases, the extent of the core facies and the ultimate size of the whole mound also increase. (7) Because the spacing of mounds, both laterally and in stratigraphic sequence, is not random, and be¬ cause octocorals may occur in moderate abundance on the mound flanks without appreciably altering mound structure, factors other than biotic appear to have influenced mound formation by determining the sites and rates of growth of the cores. (8) Because calcilutite is dominant throughout all the mounds studied, it seems unlikely that the sites of deposition of core and flank sediments differed sig¬ nificantly in resistance to erosion. Intermittent, local¬ ized erosion, however, seems to have removed both core and flank sediments from some mounds while adjacent, coeval mounds were the sites of virtually continuous deposition. (9) If the inferences in (5) to (7) are correct, abundances of cyclostome and cheilostome bryozoans in the Danian mounds in general vary because of fac¬ tors promoting or inhibiting growth of octocorals or colonial scleractinians or combinations of the two groups. This inference appears to be substantiated by the strong negative correlation in the abundances of 36 the two bryozoan groups with those of the two coral groups. (10) The distinct positive correlation between the abundances of serpulids and the category “other con¬ stituents” and those of the coral groups suggests a prin¬ cipally epizoic relationship with corals providing the substrate; the smaller size of the bryozoans may have made them a less attractive substrate for epizoans. The less significant positive correlation between the abundances of the corals and those of echinoderms and calcareous sponges suggests that the latter con¬ stituents represent a third element of the mound biota, less strongly facies-controlled than the other two. If all the preceding inferences are correct, the varia¬ tions in abundances of species of each biotic constitu¬ ent of the Danian mounds could be expected to reflect differences in adaptations to paleoenvironments repre¬ sented by the biofacies. The relationship is investigated for cheilostome Bryozoa in the following section (pages 37-47). These inferences add little to the already extensive interpretations of the physical conditions surrounding the formation of Danian mounds, summarized by Berthelsen (1962:235-245) and Asgaard (1968:116- 118). The range of interpretations is indicated by pre¬ vious usage of the term “reef’ for two disdnct concepts applied to the Danian mounds. On the one hand, the mounds have been considered to have formed as shoal- water, wave-resistant structures containing crusts pos¬ sibly produced by calcareous algae and including large amounts of skeletal detritus “detached by . . . breakers and rolled by waves and tide-water” (Hadding 1941: 120, 122). On the other hand, the mounds have been thought to represent accumulations “at a depth of more than 50 m and probably more than 100 m” and to have their closest Recent analogues “in coral beds at a depth of 100-300 m” (Hadding 1941:124, 126). Evidence indicating a shallow-water origin for the Danian mounds has been chiefly from mollusks, bra- chiopods, some octocorals, and “penetrations of thal- lophytes . . . probably algal” (Asgaard 1968:117). These interpretations, for the most part, place the upper limit for the depths at which mounds were formed at 40 to 50 m, rather than the near-surface environments suggested by the first of the concepts mentioned above. Evidence indicating a deeper-water origin has been chiefly from scleractinian corals and some octocorals. The Danian scleractinians include both solitary and SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY colonial forms, all ahermatypic (Floris 1967). Teichert (1958:1087) considered the Danian colonial coral association to be closely analogous with the mod¬ ern Norwegian deep-water coral banks, and therefore thought that the Danian mounds accumulated at depths greater than 100 m. Floris (1967:151), making the same comparison, concluded that an average depth of 75 m was likely. Both shallow- and deep-water interpretations have been made for the extensive, coral-dominated mounds at Fakse, in which Floris (1967:151) suggested that depositional relief may have approached 50 m where submarine talus of rounded coral limestone fragments is present. It is unlikely, however, that relief ap¬ proached this magnitude among the middle Danian mounds at Limhamn or the lower Danian ones on Stevns Klint. Squires (1964), in consideration of the structure and faunal composition of Recent and fossil deep¬ water coral mounds, recognized three developmental stages (single colony, thicket, and coppice) leading to the formation of coral banks. Of the Danian mounds, only the largest of those at Fakse, in the middle Dan¬ ian, probably formed by continued accretion of debris from a living cap of coral together with impaction of mud and thus represent the full bank stage (Squires 1964:905). The largest Fakse mounds thus might be comparable to Recent banks in Norway (Dons 1944, Burdon-Jones and Tambs-Lyche 1960) as suggested by Teichert (1958). The more typical mounds, such as those in the middle Danian at Limhamn, probably represent less-advanced, smaller stages of bank devel¬ opment, that is, thickets or coppices (Squires 1964: 905), and thus could be compared with Recent coral mounds on the edge of the European Atlantic shelf from the British Isles to Portugal (LeDanois 1948: 161-191). The small amount of recognizable debris of colonial scleractinians in the Limhamn mounds sug¬ gests that these structures had not advanced beyond the thicket stage, but the great concentrations of broken calcareous axes of octocorals indicates that an appreciable amount of time was involved in the ac¬ cumulation of the mound cores. Perhaps the octocorals contributed to mound formation through growth analogous to that of the colonial scleractinians, and in the early Danian mounds on Stevns Klint octocorals may even have been able to produce thickets or cop¬ pices without a significant scleractinian component. If the cores of the typical Danian mounds repre¬ sent octocoral-scleractinian thickets or coppices, then NUMBER 6 37 some process periodically terminating coral growth must have been responsible for the upward gradation of the core limestone into bedded limestone indistin¬ guishable from that on the mound flanks. Similar ter¬ mination of thickets in the Miocene and Pliocene in New Zealand was reported by Squires (1964:913-914) and was inferred by him to have been produced by re¬ newed or increased sedimentation which drowned the corals. The sediments enclosing the New Zealand fos¬ sil thickets and those impacted in them and in known modern deep-water coral mounds are chiefly terrig¬ enous muds and sands (Squires 1964:909), in con¬ trast to the overwhelmingly carbonate composition of all of the fine-grained sediment associated with Danian mounds. If the calcilutite in the Danian mounds was produced largely in situ through fragmentation of skeletons of benthos by the action of other organisms, as suggested above, this material may have been win¬ nowed from the coarser debris and periodically rede¬ posited over the cores, or it might simply have been produced in greater quantity in the core region. The local unconformities suggest periodic movement of sediment, whereas the distribution of non-carbonate sediment suggests more rapid accumulation of car¬ bonate mud in the cores. Thus both processes may have been involved in the drowning of coral growth. The Danian mounds contrast with all known mod¬ em deep-water coral mounds in the abundance and diversity of their associated bryozoan faunas. From the Norwegian and Celtic regions 55 species of cheilostomes have been reported by LeDanois (1948), Burdon- Jones and Tambs-Lyche (1960), and Ryland (1963), though the number of species within a given area is much smaller (Table 7). The greater number of these species is restricted, according to ranges given by the authors cited, to water more than 50 m deep. With regard to colony form, the Danian and Recent cheilo¬ stomes associated with coral mounds are similar, ex¬ cept that the Recent fauna includes a number of species having specialized forms apparently not rep¬ resented in the Danian, and the Danian includes a proportionately higher number of eschariform species. With respect to zooid morphotype, the two faunas are very different, with a considerable increase in the Re¬ cent fauna of species having more complex structure; this change is presumably a continuation of the gen¬ eral evolutionary trend in zooid structure discussed earlier. This change could account for the fact that the two faunas have no genera in common, even if they represent the same environment. Table 7. —Numbers of species of cheilostomes associated with coral mounds in Danian and Recent North Atlantic. Data on Recent faunas are from LeDanois {1948), Burdon-Jones and Tambs-Lyche {I960), and Ryland {1963). Danian Lim- hamn II-N X Recent Nor¬ way European Atlantic Total Total associated with coral mounds: Restricted to water 51 + 32 39 55 >50 m deep Restricted to water 11 20 26 <50 m deep 0 3 3 Predominant colony form: Membraniporiform 31 22 23 35 Eschariform 15 1 1 2 Vinculariiform Cellariiform, retepori- 5 4 6 8 form, or flustriform 0 5 9 10 Zooid morphotype: I 4 3 3 3 II 9 4 3 5 III 12 0 0 0 IV 15 5 4 6 V+VI 11 20 29 41 The Danian paleoenvironment prevalent in southern Scandinavia thus appears generally to have been high¬ ly favorable to cheilostome diversity and density. The development of a coral facies within this environment then appears to have provided a secondary increase in cheilostome diversity rather than the primary im¬ petus for it. Biofacies Distribution of Danian Cheilostomes Distribution in Mound II-Ni If the distribution of cheilostome species in the Danian mounds was controlled by the same factors that produced the biofacies, then the samples from Mound II-Nx should cluster on the basis of their cheilostome abundances in conformity with the Q- mode dendrogram based on the major biotic constitu¬ ents (Figure 18). A slight modification of the an¬ alytical procedure was required for the cheilostome abundances. The small numbers of minute fragments 38 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NO. SPECIMENS/APPROX. 30G NO. SPECIMENS / APPROX. IOG FRACTION 500-1000 ju 0 40 80 120 160 NO. SPECIMENS/APPROX. 3G TOTAL COARSE FRACTION (>500//) BO I2P 160 200 240 WEIGHT-EQUIVALENT PERCENT Figure 20.— Abundances of cheilostome species and numbers of species present in three size grades and total coarse fraction of Sample 17, Mound II-Ni. Size grades differ from each other and from the total proportions. (Species numbered as in Table 1.) Table 8. —Number of cheilostome specimens counted in coarse-fraction size grades of samples from Mound II-Ni {total, 18,705). Counts of identifiable specimens were used to calculate weight-equivalent percentages of species listed in Table 1 (id, identifiable; un, unidentifiable). Size grade Sample 1 2 3 4 5 6 7 9 10 11 12 13 14 15 16 17 18 >2000 m / id 2 9 23 10 12 20 38 11 10 4 12 5 40 58 32 81 23 i un 0 3 0 0 0 1 6 2 0 0 0 2 8 1 11 7 1 1000-2000 m / id 1277 598 223 110 615 114 353 201 197 313 22 96 291 405 419 678 410 1 un 82 42 27 12 20 47 29 50 58 57 11 38 28 24 9 60 50 500-1000/z / id 1127 639 699 261 762 228 286 771 500 516 194 184 318 662 515 599 342 i un 112 48 191 144 24 93 124 285 334 237 111 165 116 63 319 191 147 CF (>500 m ) id 2406 1246 945 381 1389 362 677 983 707 833 228 285 649 1125 966 1358 775 1 un 194 93 218 156 44 141 159 337 392 294 122 205 152 88 339 258 198 NUMBER 6 39 recovered for some of the 51 operational species iden¬ tified from Mound II-Ni (Table 1) made it impractical to obtain direct weight percentages. Therefore, the particles within each of the coarse-fraction size grades (containing virtually all of the identifiable cheilo- stomes, as shown in Figure 15) were assumed to be of equal weight, and weight-equivalent percentages were calculated from the particle counts (Table 8). The percentages of species in each of the three size grades are significantly different (Figure 20); as a result the weight-equivalent percentages for all three size grades were combined proportionally. All cluster analyses discussed below are thus based upon weight- equivalent percentages within the entire coarse frac¬ tion of each sample. Some species (numbered 38-51 in Table 1) were found to occur in so few samples and in such small quantities that they were omitted from the comparison. Thus, only the 37 species having at least one abundance of 1.5 percent or greater or occurring in at least six samples were included in the analyses. In Figure 21 is illustrated the general similarity of the Q-mode cluster analysis based on cheilostome abundances to that based on the major biotic con¬ stituents (from Figure 18). The cheilostome com¬ position of the core facies (A x ) is distinct from that of the flank facies (B), and that of the transitional facies (A 2 ) is more like that of the core. The distri¬ butions of the cheilostome species conform even more closely to the structure of the mound than do the major biotic constituents. The samples from each flank (1-3, 5, 6 on the western flank; 16-18 on the eastern one) form distinct subclusters within the flank cluster. The samples from the structural center of the mound (9— 12) form a distinct subcluster within the core-transi¬ tional cluster. The high degree of conformity of cheilo¬ stome species distributions to mound structure suggests an even greater sensitivity to topography-related en¬ vironmental differences than that displayed by the biota as a whole. The abundances of individual cheilostome species vary markedly across the mound (Figure 22), but the 37 CHEILOSTOME SPECIES 8 MAJOR BIOTIC CONSTITUENTS z 4 .6 .8 i.o in h ■*> » SIMILARITY (R) SIMILARITY (R) Figure 21.—Q-mode dendrograms of relationships among 17 samples from Mound II-Ni showing general distribution of 37 cheilostome species {left) in biofacies based on eight major biotic constituents {right). Each dendrogram was formed by weighted-pair-group-method clus¬ tering of correlation coefficients calculated from arcsine-square-root-transformed weight per¬ centages, standardized by rows. Biofacies assignments are as shown in Figure 19. Q-mode dendrogram of cheilostome species based upon distance function (not shown) is almost indentical to that based on correlation coefficient. UNIDENTIFIED FRAGMENTS Figure 22.—Variation in abundances (as weight-equivalent percentages) of species of cheilo- stome bryozoans in coarse fraction of samples from biofacies in Mound II-Ni. Species are numbered as in Table 1 in order of maximum abundance in any single sample. Biofacies as in Figure 12. Species 28 to 51, having abundances too small to diagram, have been lumped in the portion of the diagram shaded black. Species groups are in accordance with the R-mode dendrogram of Figure 25 and explained in the text (pages 41-44). Unidentified fragments are represented by the unshaded portion at the bottom of the diagram. NUMBER 6 41 differences in species composition of flank, core, and transitional assemblages are not apparent from the abundances alone. It is not clear, for example, whether there are three distinct assemblages, two distinct as¬ semblages with the third an intermixture, or a more or less continuous gradation from first to third. No species is restricted to any single biofacies, and only two (20 and 31) are absent from even one biofacies (AO • On the other hand, four species (13, 17, 18, and 25) occur in every sample. The relationships among assemblages should be reflected in both species diver¬ sity and density. The total number of species present in each sample varies from 18 to 38 without apparent pattern relative to the structure of the mound (Table 1). Number of species, however, is probably not an effective meas¬ ure of species diversity, because it may fluctuate with sample size, and rare species may be present or absent practically at random. The information function H(S) = — lo g e pu i = 1 where S is the number of species and pi is the pro¬ portion of the ith species, is used in ecology as a more effective measure of diversity (Buzas and Gibson 1969, and references therein). The information function for the 17 samples from Mound II-Ni (Figure 23) shows pronounced diversity highs in the transitional facies DISTANCE IN METERS FROM EAST END OF MOUND UN, Figure 23. —Variation in species diversity (below) and dom¬ inance (above) of cheilostome assemblages in the 17 samples from Mound II-Ni. Ai, Aj, and B indicate approximate posi¬ tions of biofacies based on eight major biotic constituents (Figure 19). Numbers 1-18 mark sample locations along horizontal axis of mound exposure. (A 2 ) separating a diversity low in the core (Ai) from lows on the flanks (B). Together with the Q-mode dendrogram based on the cheilostome species, the diversity pattern suggests that Biofacies A 2 contains a mixture of two distinctive assemblages which domi¬ nate Biofacies A x and B. This relationship appears to be borne out by the variation in species dominance across the mound as measured by the function (Buzas and Gibson 1969) where e is the base of natural logarithms. The low diversities of the core and flank assemblages corre¬ spond to high dominance, and the transitional assem¬ blage to low dominance, although the negative corre¬ lation is not perfect (Figure 23). If the cheilostome assemblage of Biofacies A 2 rep¬ resents an extension of that of Biofacies A x by admixture of species of Biofacies B, as the Q-mode dendrogram (Figure 21) and the diversity and domi¬ nance profiles (Figure 23) suggest, then the increase in diversity from Biofacies A x to Biofacies A 2 should be accompanied by a logarithmic increase in density (Odum et al. 1960). Most of the samples in Biofacies Ai and A 2 appear to conform to the expected relation¬ ship (Figure 24). In contrast, the assemblage in Bio¬ facies B has a much higher density than would be ex¬ pected from its low diversity and lies distinctly off the trend of the Aj. and A 2 samples; Biofacies B thus rep¬ resents a more distinctive group of samples. To identify which species of the 37 characterize each assemblage, the abundance data were submitted to R-mode cluster analysis (Figure 25). Slightly differing results were obtained with unweighted and weighted pair-group clustering (dendrograms on left and right X 2.8 > H 2.6 (f) OC £ 2.4 5 2.2 Figure 24.-— Relation between diversity and density of cheilo¬ stome assemblages of the 17 samples from Mound II-Ni. Di¬ versity is shown as in Figure 23. Density is dry-weight per¬ centage of cheilostomes in whole sediment sample. 14 6 IK © 15 16 7 © © - o 4 © A M 2 o 9 O o 12 ©' 3 '©o 5 o _ o J_ o _> B o 17 18 1 1 III i i i i 2 3 4 5 6 7 8 9 10 DENSITY AS WT.% OF SEDIMENT Figure 25.—R-mode relationships among 37 cheilostome species in Mound II-Ni and their distribution in biofacies. Clustering is based upon correlation coefficients calculated from arcsine-square-root-transformed weight-equivalent percentages for each species in the coarse fractions of the 17 samples. Unidentified fragments of cheilostomes were excluded from calcula¬ tion. Species are numbered as in Table 1. Left, Dendrogram based on unweighted-pair-group-method (UPGM) clustering (Sokal and Sneath 1963:309) and trellis diagram (shaded similarity matrix; Sokal and Sneath 1963:176) of individual correlations between all pairs of species. Clustering method emphasizes individual correlations at all levels of clustering; thus groups 2 and 3 cluster with group 4, despite several high correlations with group 1 (inset, at top). Homogeneity of groups 1 and 4a contrasts with heterogeneity of groups 2, 3, 4b, and 4c, each of which shows conspicuous overlaps with two or more other groups. Correlation coefficients in trellis diagram have been grouped in six equal classes shaded so that density indicates closeness of correlation. MAXIMUM % IN BIOFACIES CLUSTER BIOFACIES B a 2 A, ^ 5 ^ 17^2 1 15,16,/ rL 9 / 10./'ll ■ • • • • 0 ■ • • 0 0 • • • • 0 • 0 • • # 0 • 0 0 mm • • 0 jijy • • O • • 0 • • 0 • is) • • • • • 0 • 1 # £3) • • # • • • • 3 §) • • O • 0 • • • 0 • • • • O • • • ■ • ■ O • • • • • O IZJO SI • 0 • • • • • © O • • • • • • • 0 • 0 0 0 • • • • 0 • • • • • • • • 0 — 0 0 0 0 • 0 • I# o| • O O • • • 0 0 0 § • 0 • 0 O • • • • • 1 • • ol 0 0 • 1 • • • 0 • • • • 0 • 0 0.2 0.4 0.6 r(WPGM) Right, Dendrogram based on weighted-pair-group-method (WPGM) of clustering (Sokal and Sneath, 1963:309) and biofacies distribution of species in each group. Clustering method emphasizes correlations among groups regardless of how many species compose each group; thus groups 1 and 2 have equal weight at the level at which they join, even though they include very different numbers of species. The WPGM dendrogram becomes less like that clustered by UPGM at lower levels of correlation. However, the only significant difference between the two is that 2 and 3 join 1 rather than 4. The membership of each of the six clusters is the same in both dendrograms, and, with the exception of species 33, the species are arranged in identical order. The biofacies distribution of each species is indicated by its maximum weight-equivalent percentage within each of the Q-mode sample clusters based on the major biotic constituents (Figures 18, 19). Within each biofacies, the clusters are arranged from west to east. The abundances are grouped in six classes increasing geometrically and shaded in approximate proportion to density. The biofacies distributions of groups 2 and 3 are slightly more like that of 1 than that of 4. 44 of Figure 25), but both methods indicate that more than two-thirds (26) of the species belong to one or the other of two contrasting clusters: group 1 with 11 species and group 4a with 15. The correlations between pairs of species within each of these two clusters are high (trellis diagram, left side of Figure 25), and, in this sense, these clusters are homogeneous compared to the other four, each of which includes only five species or fewer. The abundances of the two main clusters, 1 and 4a, in the biofacies based on major biotic constituents are markedly different (right side of Figure 25). Every species of group 1 has its maximum abundance in Bio¬ facies B, whereas each one in group 4a is most abundant either in Biofacies Aj (7 species), in Bio¬ facies A 2 (2 species), or subequally in Biofacies Ai and A 2 (6 species). Regardless of which of the three dis¬ tributions a species in group 4a has, it is relatively more abundant in Biofacies A 1 than in Biofacies B. Therefore, group 1 appears to characterize the flanks and group 4a the core of the mound. It is noteworthy, however, that every one of the species in both groups occurs in all three biofacies; it is the abundance of species of groups 1 and 4a and not their presence or absence that conforms to the biofacies. The dominance of group 1 in Biofacies B and of group 4a in Biofacies Ai then accounts for the low diversities and high domi¬ nances shown by samples from these parts of the mound (Figure 23). The opposing abundance gradients of species groups 1 and 4a cross in the transitional areas (Biofacies A 2 ) between core and flanks. The intermediate abundances of both major clusters of species in Biofacies A 2 thus appear to explain the higher diversity and lower domi¬ nance of samples from this facies. Furthermore, the tendency for species of group 4a to maintain higher abundances in this facies than those of group 1 seems to account for its closer resemblance to Biofacies Ai than to B (Figures 18, 19). The transitional facies is additionally characterized by species groups 2, 3, 4b, and 4c, some species of which are rare or even absent (species 20 and 31) in other facies. The diversity, dominance, and composition of the transitional as¬ semblage thus seem to be an ecological edge effect, and the assemblage probably should be considered an ecotone (Odum 1959:278). If correct, this interpreta¬ tion would further argue against appreciable post¬ mortem transportation of the cheilostome fauna. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Morphologic Basis of Distribution The preceding analysis shows that Danian cheilostome species in Mound II-Ni are distributed in general concordance with the biofacies that are based upon the major biotic constituents of the sediment, and it sug¬ gests that the morphologic differences among the cheilostomes could be expected to reflect differences in adaptation to two principal environments represented by the flank and core facies. Furthermore, because the distributional differences among these species are ex¬ pressed by variation in abundance rather than by presence and absence, the underlying differences in adaptation to the two principal environments are prob¬ ably gradational, with the transition facies between the core and the flanks calling for some combination of the adaptive characteristics of the other two. The two major groups of cheilostome species having the most contrasting patterns of distribution (groups 1 and 4a in Figure 25) also show morphologic differ¬ ences with respect to inferred colony forms and zooid morphotypes (Figure 26). All the 11 species clustered in group 1 may be inferred to have possessed rigidly erect colonies (7 species eschariform, 4 species vincu- lariiform), and they include all the inferred vinculari- iform species identified from Mound II-Ni except one rare species omitted from the R-mode analysis (species 50 in Table 1, found in only one sample; as this occur- VINC (/) # o cn o u_ > ESCHAR Z 3 % f3l 0 o o MEMB • 4cY- 9 I ir am m ~yoR~vr ZOOID MORPHOTYPES Figure 26. —Inferred colony forms (see Table 1) and zooid morphotypes of R-mode groups of cheilostome species in Mound II-Ni. Species belonging to each cluster are identi¬ fied by number in Figure 25, and their morphology is indi¬ cated in Table 1. Circles are approximately proportional to summed maximum abundances of species having each com¬ bination of zooid morphotype and colony form, and are di¬ vided in approximate proportion to the groups represented. Largest circle, about 80%; next to largest, about 40%; next to smallest, about 20%; smallest, about 10%. NUMBER 6 45 rence is in Biofacies B, this species presumably would sort with group 1). This group of species, as indicated above, dominates the flanks of the mound, where the cheilostome assemblage shows an unexpectedly high density relative to its diversity. All the 15 species clustered in group 4a appear to have had closely encrusting (membraniporiform) colonies. As indicated above, these species have maxi¬ mum abundances in the core or transitional facies, and the cheilostome assemblages in these biofacies show the expected direct logarithmic relationship between den¬ sity and diversity. The 11 species belonging to the other four groups (2, 3, 4b, and 4c), which characteristically show high abundances in the transitional facies, appear to have had encrusting or eschariform, but not vinculariiform, colonies. With respect to their potential density relative to the substrate, these species may be considered inter¬ mediate in adaptation between groups 1 and 4a. Five species (listed in Table 9) were interpreted to have unstable colony form, including both erect and encrusting phases. Among these species, there is a tendency for the encrusting phase to be replaced by the erect habit from the core to the flanks of the mound, even though the ratio between the two forms differs from one species to another. The same tendency is apparent in the genus Floridina (Table 10), which has a vinculariiform species dominating on the flanks, a membraniporiform species dominating in the core, and intermediate proportions of membraniporiform and eschariform species in the transitional facies. To test the apparent correlation between biofacies distribution and inferred colony form, the abundance data were regrouped from species to colony forms (up¬ per part of Figure 27). These three categories were then substituted for cheilostomes in the data array on Table 9. —Changes in inferred colony forms {see Table 1) of unstable cheilostome species in biofacies of Mound II-Nj. Ratios are between average weight-equivalent percentages of inferred erect and encrusting phases. Species are numbered as in Table 1; samples are grouped as in Figures 18 and 19. Species group Unstable species Colony forms Biofacies B A 2 A i 1 16 vinc/memb 100/0 75/25 33/67 18 eschar/memb 99/1 54/46 26/74 4a 25 ESCHAR/mEMB 32/69 0/100 0/100 5 eschar/memb 1/99 0/100 0/100 3 28 eschar/memb 1/99 0/100 0/100 Table 10. —Changes in abundance proportions in biofacies of Mound II-Nj of three species of Floridina having different inferred colony forms. Species are numbered as in Table 1; samples are grouped as in Figures 18 and 19. Species group Species Colony form Biofacies B ^■2 A\ 1 2 VINC 89 9 1 3 20 ESCHAR 1 24 0 4a 7 MEMB 10 67 99 which R-mode cluster and principal-component anal¬ yses were made. The result (Figure 28) is similar to the original dendrogram, with the erect groups (vincu¬ lariiform and eschariform) clustering at a high level and occupying the cheilostome position in the bryozoan cluster of the original dendrogram. The encrusting group (membraniporiform), on the other hand, shifted to a new position far removed from the other cheilo¬ stomes, into the coral cluster. This pattern of clustering confirms that the major difference in biofacies is ex¬ pressed morphologically in colony form. If the adaptation expressed in colony form is based upon the structure of the zooids—the hypothesis pro¬ posed in the theoretical consideration of cheilostome structure above—then the abundance data regrouped by inferred zooid morphotypes (lower part of Figure 27) should also correlate with the biofacies. The dis¬ tribution patterns in Figure 27 suggest that this cor¬ relation is much less conspicuous than that between colony form and biofacies, but the morphotypes cluster with the major biotic constituents (Figure 29) in much the same way that the colony forms do. Mor¬ photypes iv-vi cluster at a high level and occupy the cheilostome position in the bryozoan cluster of the original dendrogram, as do the erect colony forms shown in Figure 28. Morphotype m, like the encrusting colony form, joins the coral cluster in a position far removed from morphotypes rv-vi. The remaining mor¬ photypes occupy somewhat looser positions, morpho¬ type n joining the bryozoan cluster at a low level of correlation and morphotype i joining the non-bryozoan groups at a still lower level. The patterns of changes in abundance of these two morphotypes, however, con¬ form in general with that of morphotype m (Figure 27), that is, with peaks in the core and transition facies. The reason for the looseness of correlation of bio¬ facies with zooid morphotypes, as compared to that with colony form, is suggested in Figure 26. With the 419-995 0 - 71-4 46 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY xvxxxvxxxvx .v.'.v.v.vXv.v!v/J- •XvMvIvXvXv'- ■•!! XvXvXvX-x^: ~- XvX X;Xv: ix*x*x*x*x‘x*xvx*x*x*xvxvxvx* SSSS xxxxxx xxxx* ■■■■■■■■■■■■>V : : : x : x : : : : x : : : x : v.v.v.v. ‘XvXvX ••XvXv , v.v.vX X 4 *XvX •.v.v.v.v.*.*. x : x : x : : : X;XvXyXtX xxxxxxxxxxxxxxxxx • • xxxxxxxxxxxxxxxxx.. wmmmk,,,, wmmmmm wmmmmmmmmmmmmmz*. . V.V.V.V.V.*.V.V.V.V.V.*.V.V.V.V.,V.V.V.V.V Mm. . l»**ii*i*»s iMWW !HBB XxXXXXXXXX.viiHSiiiiiii:::: ... . smmmMM mmmmm \mmMM fmmmw mm 1181 vXvX’/XvXvX'X’X’ ioilll* >XX".”X;: XvXvX :::::::: Figure 27.—Variation in abundances (as weight-equivalent percentages) of morphologic groups of cheilostome bryozoans in coarse fractions of samples from Mound II-Ni. Abundances shown in Figure 22 for 51 species have been redistributed by morphology, with unidentified cheilostomes omitted. Abundances in each diagram add up to 100 percent. In the distribution of inferred colony forms {above) abundances of stable species showing same form have been summed and abundances of unstable species have been divided according to proportions of forms exhibited and the portions added to the appropriate categories. In the distribution of inferred zooid morphotypes {below) abundances have been summed for species having the same morphotype. NUMBER 6 47 Figure 28. —R-mode dendrogram of relationships among major biotic constituents and inferred colony forms of cheilo- stomes in Mound II-Ni and influence of principal components on clustering. Dendrogram is similar to that shown in Figure 17, except that cheilostome abundances have been propor¬ tionately divided among three colony forms as indicated in Figure 27. Abscissa is product-moment correlation coefficient. exception of group 2, comprising two species, and 4b, consisting of just one, each of the R-mode groups of species includes a range of inferred zooid morphotypes. In the two major groups of species (1 and 4a) domi¬ nating contrasting biofacies, however, the ranges of morphotypes differ. A majority of group 1 species ex¬ hibit morphotypes iv-vi, whereas those of group 4a are distributed among all morphotypes with a concen¬ tration in groups ii-iv. Group 3 shows an intermediate range of morphotypes, and group 4c is like 4a but ex¬ cludes morphotypes v-vi. These relations between the Danian species groups and the two sets of morphologic characteristics therefore suggest that the more com¬ plex zooid morphotypes are associated with the erect colony habit, whereas the encrusting form of growth may be virtually independent of morphotype. Conclusions The marked increase in numbers of cheilostome taxa from Early Cretaceous time to an apparent pause dur¬ ing Danian time reflects morphologic diversification at both colonial and zooidal levels of organization. The diversified and abundant fauna associated with Danian mounds in southern Scandinavia thus represents the culmination of primarily divergent evolutionary trends in cheilostome morphology. The abundances and biofaces distributions of cheilostomes in the Dan¬ ian mounds are consistent with the hypothesis that, in general, the form of the colony depends upon the structure of the zooids. This is not to say that any given colony form was restricted to any particular zooid morphotype, but rather that the attainment of a colony form permitting increased zooid densities was functionally more probable for some zooid mor- photoypes than for others. The correlation between such a functionally specified structural series and ob¬ served abundances, of course, establishes only an in¬ ferential relationship, and it is with this limitation that the following adaptive relationships are suggested: (1) The ability of cheilostomes to assume an erect growth habit appears to represent an evolutionary adaptation that vastly increased the potential for zooid density relative to the amount of substrate occupied. The erect colony, compared to the encrusting one, should be much less dependent upon the availability of surfaces suitable for direct adherence, less affected by sedimentation, and exposed more fully to the sur¬ rounding water. The advantage of erect growth is sug¬ gested in the Danian cheilostome fauna by the over¬ whelming abundance of forms inferred to have had rigidly erect colonies, even though they account for fewer than half the species present. Furthermore, their negative correlation in abundance with larger constit¬ uents of the sediments, which might have provided the surfaces of adherence for encrusting colonies, is con¬ sistent with a reduced sensitivity to limitation of sub¬ strate. 1.0 .8 .6 4 .2 0 -2 - .4 Figure 29.—R-mode dendrogram of relationships among major biotic constituents and inferred zooid morphotypes of cheilostomes in Mound II-Ni. Dendrogram is similar to those in Figures 17 and 28, except that cheilostome abundances have been proportionately divided among five categories of zooid morphotypes as indicated in Figure 27. Abscissa is product- moment correlation coefficient. 48 (2) Although some modem erect forms having flexi¬ ble (flustriform) colonies may have originated through reduction of zooecial calcification, the primary evolu¬ tionary attainment of erect growth appears to have been made possible by reinforcement of zooid walls beyond that present in the earliest cheilostomes known. Calcified basal walls and at least partly calcified walls on the frontal side of the zooid appear to have been the minimal advancements necessary as a prospective adaptation for the structural support of a rigidly erect colony. Among Danian and earlier cheilostomes which can be inferred to have grown erect, zooecia have at least the morphotype n degree of complexity—that is, they possess an extensive gymnocyst, cryptocyst, or frontal shield. (3) For a growing, rigidly erect, non-fenestrate cheilostome colony, the most efficient means of taking up stresses—due to vertical loading, bending, and twisting—appears to be that of concentrating further calcification, beyond the minimal requirement for erect growth, near the frontal surfaces of the zooids and to do this in increasing proportion toward the proximal end of the colony. To the extent that their joint calcification approaches a laterally merging, con¬ tinuously thickening, distally tapering skeletal mass analogous to the outer walls of an enlarging cantilever beam, zooids of morphotypes n-vi appear to form a graded series of increasingly efficient building blocks for larger colonies and hence denser populations. Ex¬ cept for the unexpectedly low abundance of morpho¬ type m (cribrimorph), the abundances of morpho¬ types among Danian species inferred to have grown erect are proportional to their postulated functional efficiencies. (4) If a zooid structure at least as complex as morphotype n was required for attainment of a rigidly erect growth form, whatever adaptive significance frontal calcification may have had originally must be associated with a function other than colony support, such as protection of the lophophore and associated organs. Some or all of the morphotypes beyond type ii could also represent, with respect to rigidly erect growth, prospective adaptations which served a differ¬ ent original function or functions. Such a series of prospective adaptations would account for the pres¬ ence of the more complex morphotypes in encrusting species among Danian and earlier faunas and in the majority of modern cheilostomes of every colony form except some that are non-rigidly erect (flustriform). SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY The rarity of inferred erect cheilostomes of morpho¬ type in in the Danian fauna suggests that this morpho¬ type arose as a prospective adaptation. The correlation between abundances of inferred erect species and mor¬ photypes iv—vi in the Danian fauna, however, suggests that evolution beyond morphotype in consisted of direct adaptive improvements for rigidly erect growth. Otherwise, each evolutionary step in adaptation for the other function or functions would have required a parallel increase in prospective adaptation for colony support. There is evidence that some lineages having complex morphotypes started with erect colonies and evolved an encrusting form with concomitant thinning of the frontal shield (Cheetham 1968:10—11). Whether this kind of trend might account for many other en¬ crusting species with complex morphotypes can be determined only by detailed studies of many other lineages. (5) If the rigidly erect form represents an adaptive type attained during cheilostome evolution, then the ecologic niche or group of niches for which it was suited must have been either vacant or occupied by organisms for some reason competitively inferior to cheilostomes. Among the other constituents of the biota associated with Danian mounds, the cyclostome bryo- zoans display zoarial forms which can be inferred to approximate the cheilostome rigidly erect (vinculari- iform and eschariform) growth most closely. Cyclo- stomes having apparently erect colonies are known from strata throughout the Cretaceous, and some are coeval and sympatric with the earliest known cheilo¬ stomes. During Late Cretaceous time, the cyclostomes were progressively overtaken in numbers of genera and species by the cheilostomes (Voigt 1959a: 702), which apparently have had a large margin of numeri¬ cal dominance at these categorical levels throughout the Cenozoic. Whether the relationship between diver¬ sities in the two groups is one of negative correlation, suggesting competitive replacement, has not been deter¬ mined because of the many uncertainties in the ranges of cyclostome taxa (Larwood et al. 1967:385). Literature Cited Asgaard, U. 1968. Brachiopod Palaeoecology in Middle Danian Lime¬ stones at Fakse, Denmark. Lethaia, 1:103-121, 7 figures. Askren, L. T., Jr. 1968. Bryozoan Paleoecology from the Tertiary of Ala¬ bama. Southeastern Geology, 9:157-163, 3 figures. NUMBER 6 49 Banta, W. C. 1968. The Body Wall of Cheilostome Bryozoa, I. The Ectocyst of Watersipora nigra (Canu and Bassler). Journal of Morphology, 125:497-508, 1 plate, 6 figures. 1970. The Body Wall of Cheilostome Bryozoa, III. The Frontal Wall of Watersipora arcuata Banta, with a Revision of the Cryptocystidea. Journal of Mor¬ phology, 131: 37-56, 26 figures. Berthelsen, O. 1948. Studies on the Bryozoan Species Coscinopleura elegans and Coscinopleura angusta n. sp. from the Senonian and Danian Deposits of Denmark. Dan- marks geologiske Undersegelse, 3(3): 1-15, 4 figures. 1962. Cheilostome Bryozoa in the Danian Deposits of East Denmark. Danmarks geologiske Under- sogelse, 83:1—290, 28 plates, 31 figures. Boardman, R. S., and A. H. Cheetham 1969. Skeletal Growth, Intracolony Variation, and Evolu¬ tion in Bryozoa: a Review. Journal of Paleontol¬ ogy, 43:205-233, plates 27-30, 8 figures. Boardman, R. S., A. H. Cheetham, and P. L. Cook 1970. Intracolony Variation and the Genus Concept in Bryozoa. Proceedings of the North American Paleontological Convention, pages 294-320, 12 figures. Bock, W. J., and G. von Wahlert 1965. Adaptation and the Form-Function Complex. Evolution, 9:269-299, 1 figure. Bowman, R. I. 1961. Morphological Differentiation and Adaptation in the Galapagos Finches. 326 pages, 22 plates, 74 figures. Berkeley and Los Angeles, University of California Press. Brotzen, F. 1959. On Tylocidaris Species (Echinoidea) and the Stratigraphy of the Danian of Sweden with a Bibli¬ ography of the Danian and the Paleocene. Sveriges geologiska Undersokning, Avhandlingar, series C, 54(2) : 1-81, 3 plates, 19 figures. Burdon-Jones, C., and H. Tambs-Lyche 1960. Observations on the Fauna of the North Bratthol- men Stone-Coral Reef near Bergen. Arbok for Unicersitetet i Bergen, matematisk-naturvidensk, Serie 4, 29:1-24, 1 figure. Buzas, M. A., and T. G. Gibson 1969. Species Diversity: Benthonic Foraminifera in Western North Atlantic. Science, 163:72-75, 2 figures. Canu, F., and R. S. Bassler 1920. North American Early Tertiary Bryozoa. United States National Museum Bulletin, number 106, 879 pages, 162 plates, 279 figures. 1926. Studies on the Cyclostomatous Bryozoa. Proceed¬ ings of the United States National Museum, volume 67, number 7593, 124 pages, 31 plates, 46 figures. Casey, R. 1964. The Cretaceous Period. In W. B. Harland, A. G. Smith, and B. Wilcock, editors, The Phanerozoic Time-Scale. Quarterly Journal of the Geological Society of London, 120(S) : 193—202. Chayes, F. 1960. On correlation between Variables of Constant Sum. Journal of Geophysical Research, 65:4185-4193. Cheetham, A. H. 1963. Late Eocene Zoogeography of the Eastern Gulf Coast Region. Geological Society of America Memoir, number 91, 113 pages, 3 plates, 34 figures. 1966. Cheilostomatous Polyzoa from the Upper Brackle- sham Beds (Eocene) of Sussex. Bulletin of the British Museum (Natural History), Geology, 13:3- 115, 81 figures. 1968. Morphology and Systematics of the Bryozoan Genus Metrarabdotos. Smithsonian Miscellaneous Collec¬ tions, volume 153, number 1, 121 pages, 24 figures. Cook, P. L. 1964. Polyzoa from West Africa. I. Notes on the Stegano- porellidae, Thalamoporellidae and Onychocellidae (Anasca, Coilostega). Resultats scientifiques des campagnes de la Calypso: lies du Cap Vert. An- nales de Vlnstitut oceanographique, Monaco, 41:43-78, 1 plate, 13 figures. 1965. Notes on the Cupuladriidae (Polyzoa, Anasca). Bulletin of the British Museum (Natural History), Zoology, 13:151-187, 3 plates, 6 figures. 1968a. Polyzoa from West Africa, the Malacostega, Part I. Bulletin of the British Museum (Natural History), Zoology, 16:115-160, 3 plates, 20 figures. 1968b. Bryozoa (Polyzoa) from the Coasts of Tropical West Africa. Atlantide Report No. 10, pages 115-262, plates 8-11, 2 figures. Dobzhansky, T. 1956. What Is an Adaptive Trait? American Naturalist, 90:337-347. Dons, C. 1944. Norges Koralrev. Trondheims biologisk Station, Meddelelser, 135:37—82, 15 figures. Dullemeijer, P. 1958. The Mutual Structural Influence of the Elements in a Pattern. Archives Neerlandaises de Zoologie, 13 (supplement 1): 74-88. Floris, S. 1967. Scleractinian Coral Faunas of Denmark and West Greenland in Senonian, Danian and Paleocene. Meddelelser fra Dansk geologisk Forening, Kjo- benhavn, 17:150-152. Funnell, B. M. 1964. The Tertiary Period. In W. B. Harland, A. G. Smith, and B. Wilcock, editors, The Phanerozoic Time-Scale. Quarterly Journal of the Geological Society of London, 120(S) : 179—191. Gautier, Y. V. 1962. Recherches ecologiques sur les bryozoaires cheilo- stomes en Mediterrannee occidentale. Recueil des travaux de la Station marine d’Endoume, 38(24) : 5-434, 91 figures. 50 Greeley, R. 1967. Natural Orientation of Lunulitiform Bryozoans. Bulletin of the Geological Society of America, 78:1179-1182, 2 plates, 2 figures. Hadding, A. 1941. The Pre-Quaternary Sedimentary Rocks of Sweden VI. Reef Limestones. Lunds Universitets Arssknft, 37:116-127. Harmer, S. F. 1902. On the Morphology of the Cheilostomata. The Quarterly Journal of Microscopical Science, Lon¬ don, new series, 46:263-350, plates 15-18. 1926. The Polyzoa of the Siboga Expedition, Part II, Cheilostomata Anasca. Siboga Expeditie, Leyden, volume 28b, pages 181—501, plates 13-34, 23 figures. 1930. Polyzoa, Presidential Address Delivered at the An¬ niversary Meeting of the Linnean Society of Lon¬ don on the 24th of May, 1929. Proceedings of the Linnean Society of London, 141:68-118, 1 plate, 11 figures. 1957. The Polyzoa of the Siboga Expedition, Part IV, Cheilostomata Ascophora II. Siboga Expeditie, Leyden, volume 28d, pages 641—1147, plates 42—74, figures 49-118. Hennig, A.H. 1892. Studier ofver Bryozoerna i Sveriges Kritsystem, I. Cheilostomata. Lunds Universitets Arsskrift, 28: 1-51, plates 1-2. 1894. Studier ofver Bryozoem i Sveriges Kritsystem, II. Cyclostomata. Lunds Universitets Arsskrift, 30:1- 46, plates 1-2. 1899 Studier ofver den Baltiska Yngre Kritans Bildning- shistoria. Geologiska foreningens i Stockholm for- handlingar, Stockholm, 21:19-82. Jullien, J. 1881. Note sur une nouvelle division des Bryozoaires cheilostomiens. Bulletin de la Societe zoologique de France, Paris, 6:271-285, 4 figures. Jiirgensen, T. 1968. Cheilostome Bryozoa from the Danian Deposits at Vallensbaek, Denmark. Meddelelser fra Dansk Geologisk Forening, 18(2) : 187-204, 1 plate, 4 figures. Kaufmann, K. W., Jr. 1968. The Biological Role of Bugula —Type Avicularia (Bryozoa), (Preliminary Report). In E. Annos- cia, editor, Proceedings of the First International Conference on Bryozoa, Atti della Societd Italiana diScienze Naturali, 108:54-60, 1 figure. Labracherie, M., and J. Prud’homme 1967. Observations pakioecologiques sur les Bryozoaires des formations classiques du Blayais. Actes de la Society Linneenne de Bordeaux, serie B, 104:3-9, 2 plates. Lagaaij, R., and Y. V. Gautier 1965. Bryozoan Assemblages from Marine Sediments of the Rhone Delta, France. Micropaleontology, 11:39-58, 34 figures. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY Lang, W. D. 1914. Some New Genera and Species of Cretaceous Cheilostome Bryozoa. Geological Magazine, de¬ cade 6, 1:436—444, plate 34. 1916. Calcium Carbonate and Evolution in the Polyzoa. Geological Magazine, decade 6, 3: 73-77. 1919. The Pelmatoporinae, an Essay on the Evolution of a Group of Cretaceous Polyzoa. Philosophical Trans¬ actions of the Royal Society of London, series B, 209:191-278, 72 figures. 1921. Catalogue of the Fossil Bryozoa {Polyzoa) in the Department of Geology, British Museum (Natural History), The Cretaceous Bryozoa {Polyzoa), Part I. The Gribrimorphs. Volume 3, 267 pages, 8 plates, 115 figures. London, British Museum (Natural History). 1922. Catalogue of the Fossil Bryozoa {Polyzoa) in the Department of Geology, British Museum {Natural History), The Cretaceous Bryozoa {Polyzoa), Part II. The Cribrimorphs. Volume 4, 403 pages, 8 plates, 125 figures. London, British Museum (Natural History). Larwood, G. P. 1962. The Morphology and Systematics of Some Creta¬ ceous Cribrimorph Polyzoa (Pelmatoporinae). Bulletin of the British Museum {Natural History), Geology, 6:1-285, 23 plates, 132 figures. 1969. Frontal Calcification and Its Function in some Cretaceous and Recent Cribrimorph and Other Cheilostome Bryozoa. Bulletin of the British Museum {Natural History), Zoology, 18:171-182, 10 figures. Larwood, G. P., A. W. Medd, D. E. Owens, and R.Tavener- Smith 1967. Bryozoa. Pages 379-395, 3 figures, in The Fossil Record, Geological Society of London. Laughbaum, L. R. 1960. A Paleoecologic Study of the Upper Denton Forma¬ tion, Tarrant, Denton, and Cooke Counties, Texas. Journal of Paleontology, 34:1183—1197, 3 figures. Levinsen, G. M. R. 1925. Undersogelser over Bryozoerne i den danske Kridt- formation. Det kongelige Danske Videnskabernes Selskabs Skrifter, naturvidenskabelig og mathema- tisk Afdeling, 8 Raekke, 7:283—445, 8 plates, 28 figures. Lutaud, G. 1961. Contribution a l’etude du bourgeonnement et de la croissance des colonies chez Membranipora mem- branacea (Linne), bryozoaire cheilostome. An- nales de la Societe royale zoologique de Belgique, 91:157-299, 8 plates, 28 figures. LeDanois, E. 1948. Les Profondeurs de la mer; trente ans de recherches sur la faune sous-marine au large des cotes de France. 303 pages, 8 plates, 56 figures. Paris, Payot. Marcus, E. 1926. Bryozoa. In G. Grimpe and E. Wagler, editors, Die Tierwelt der Nord- und Ostsee. Number 4, NUMBER 6 51 Part 7C, 100 pages, 168 figures. Leipzig, Akade- mische Verlagsgesellschaft. Maryanska, T. 1969. Bryozoa from the Uppermost Maastrichtian and Palaeocene Deposits of the Middle Vistula Gorge near Pulawy. Prace Muzeum Ziemi, Warszawa, 14:85-129, 13 plates. Maxwell, W. G. H. 1968. Atlas of the Great Barrier Reef. 258 pages, 166 figures. New York, Elsevier. Mayr, E. 1969. Principles of Systematic Zoology. 428 pages. New York, McGraw-Hill. Nye, W. L., D. Hamilton, and J. P. Eames 1940. Procedure Handbook for Aircraft Stress Analysis. 334 pages, 241 figures. San Francisco, Aviation Press. Odum, E. P. 1959. Fundamentals of Ecology. 546 pages, 158 figures. Philadelphia and London, W. B. Saunders. Odum, H. T., J. E. Cantlon, and L. S. Kornicker 1960. An Organizational Hierarchy Postulate for the In¬ terpretation of Species—Individual Distributions, Species Entropy, Ecosystem Evolution, and the Meaning of a Species—Variety Index. Ecology, 41:395—399, 3 figures. Odum, H. 1926. Studier over Daniet i Jylland og pa Fyn. Danmarks geologiske Undersogelse, series 2, 45:5-306, 3 plates, 29 figures. Oliver, W. A., Jr. 1968. Some Aspects of Colony Development in Corals. In Paleobiological Aspects of Growth and Devel¬ opment, A Symposium. Paleontological Society Memoir, Number 2, Supplement to Journal of Paleontology, 42(5): 16-34, 6 figures. Ostroumov, A. 1903. Sur le developpement du cryptocyst et de la cham- bre de compensation. Zoologischer Anzeiger, Deutsche Zoologische Gesellschaft, 27:96-97. Pergens, E., and A. Meunier 1886. La faune des Bryozoaires garumniens de Faxe. Annales de la Societe royale malacologique de Belgique, 21: 187-242, plates 9-13. Powell, N. A., and P. L. Cook 1966. Conditions Inducing Polymorphism in Thalamo- porella rozieri (Audouin) (Polyzoa, Anasca). _ Cahiers de Biologie Marine, 7:53-59, 1 figure. Rosenkrantz, A. 1937. Bemaerkninger om det ostsjaellandske Daniens Stratigrafi og Tektonik. Meddelelser fra Dansk geologisk Forening, Kjobenhavn, 9:199-212, 6 figures. Rosenkrantz, A., and H. W. Rasmussen 1960. South-eastern Sjaelland and Mon, Denmark, Guide to Excursions Nos. A 42 and C 37, Part 1. Inter¬ national Geological Congress, XXI Session, 17 pages, 11 figures. Rudwick, M. J. S. 1968. Some Analytic Methods in the Study of Ontogeny in Fossils with Accretionary Skeletons. In Paleo¬ biological Aspects of Growth and Development, a Symposium. Paleontological Society Memoir, Number 2, Supplement to Journal of Paleontology, 42(5) : 35-49, 17 figures. Ryland, J. S. 1963. Systematic and Biological Studies on Polyzoa (Bry¬ ozoa) from Western Norway. Sarsia, 14:1-59, 14 figures. 1967a. Polyzoa. Oceanography and Marine Biology, An Annual Review, 5:343-369, 8 figures. 1967b. Respiration in Polyzoa (Ectoprocta). Nature, 216:1040-1041. Schopf, T. J. M. 1969. Paleoecology of Ectoprocts (Bryozoans). Journal of Paleontology, 43:234-244, 5 figures. Seal, H. L. 1964. Multivariate Statistical Analysis for Biologists. 207 pages. London, Methuen. Silen, L. 1938. Zur Kenntnis des Polymorphismus der Bryozoen. Die Avicularien der Cheilostomata Anasca. Zo- ologiska Bidrag fran Uppsala, 17:149-366, 18 plates, 80 figures. 1941. Cheilostomata Anasca (Bryozoa) collected by Dr. Sixten Bock’s Expedition to Japan and the Bonin Islands, 1914. Arkiv for Zoologi, Stockholm, 33A: 1-130, 9 plates, 183 figures. 1942. Origin and Development of the Cheilo-Ctenostoma- tous Stem of Bryozoa. Zoologiska bidrag fran Uppsala, 22:1-59, 64 figures. 1944. On the Formation of the Interzoidal Communica¬ tions of the Bryozoa. Zoologiska Bidrag fran Uppsala, 22:433-488, 1 plate, 59 figures. Simpson, G. G. 1953. The Major Features of Evolution. 434 pages. New York, Columbia University Press. 1958. Behavior and Evolution. Pages 507-535 in A. Roe and G. G. Simpson, editors, Behavior and Evolution. New Haven, Yale University Press. Sokal, R. R., and P. H. A. Sneath 1963. Principles of Numerical Taxonomy. 359 pages. San Francisco, W. H. Freeman. Squires, D. F. 1964. Fossil Coral Thickets in Wairarapa, New Zealand. Journal of Paleontology, 38:904-915, plates 147- MS. Stach, L. W. 1935. Growth Variation in Bryozoa Cheilostomata. Annals and Magazine of Natural History, series 10, 16:645-647. 1936. Correlation of Zoarial Form with Habitat. Jour¬ nal of Geology, 44:60-66, 1 figure. 1937. The Application of Bryozoa in Cainozoic Stratig¬ raphy. Australian and New Zealand Association for the Advancement of Science, Report of the twenty-third meeting, pages 80-83. 52 Teichert, C. 1958. Gold- and Deep-Water Coral Banks. Bulletin of the American Association of Petroleum Geologists, 42:1064-1082. Thompson, D. W. 1942. On Growth and Form. Edition 2, 116 pages. New York: MacMillan Company. Veenstra, H. J. 1963. Microscopic Studies of Boulder Clays. State Uni¬ versity of Groningen, Thesis, 211 pages, 13 plates, figures. Voigt, E. 1923. Ober einige neue und wenig bekannte Bryozoen der Gattung Floridina aus dem Danien von Faxe. Meddelelser fra Dansk geologisk Forening, 6:1-8, 1 plate. 1924. Uber neue Bryozoen aus Daniengeschieben Anhalts. Palaeontologische Zeitschrift, 6:3-13, 1 plate, 1 figure. 1925. Neue cribrimorphe Bryozoen aus der Familie der Pelmatoporidae in Kreidegeschieben Anhalts. Zeitschrift fiir Geschiebeforschung, 1(3): 97—104, plate 3, 4 figures. 1928. Neue artikulierte cheilostome Bryozoen aus einem Kreidegeschiebe obersenonen Alters von Cothen in Anhalt. Zeitschrift fiir Geschiebeforschung, 4(3): 105-114, 5 figures. 1930. Morphologische und stratigraphische Untersuchun- gen fiber die Bryozoen fauna der oberen Kreide. I. Teil. Die cheilostomen Bryozoen der jfingeren Oberkreide in Nordwestdeutschland, in Baltikum und in Holland. Leopoldina (Berichte der kaiser- lich leopoldinischen deutschen Akademie der Natur- forscher zu Halle), 6:379-579, 39 plates. SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 1939. tJber die Dornen Spezialisation bei cheilostomen Bryozoen und die Nichtumkehrbarkeit der Entwick- lung. Palaeontologische Zeitschrift, 21.87—107, 19 figures. 1956. Untersuchungen fiber Coscinopleura Marsson (Bryoz. foss.) und verwandte Gattungen. Mit- teilungen aus dem Geologischen Staatsinstitut in Hamburg, 25:27-75, plates 1-12, 7 figures. 1957. Bryozoen aus dem Kreidetuff von St. Symphorien bei Ciply (Ob. Maastrichtien). Bulletin, de I’ln- stitut royal des Sciences naturelles de Belgique, 33(43) : 1-48, 12 plates, 1 figure. 1959a. La signification stratigraphique des Bryozoaires dans le Cretace superieur. 84 a Congres des Societis savantes, pages 701—707. 1959b. t)ber Fissuricella n. g. (Bryozoa foss.). Neues Jahrbuch fiir Geologie und Palaontologie Ab- handlungen, 108:268—269, plates 25—26. 1964. A Bryozoan Fauna of Dano-Montian Age from Boryszew and Sochaczew in Central Poland. Acta Palaeontologica Polonica, 9:419-480, plates 1-16, 2 figures. 1967. Okologische und stratigraphische Untersuchungen an Bryozoen der oberen Kreide. Berichte deutsche Gesellschaft fiir geologische Wissenschaften, Reihe A, Geologie and Palaontologie, 12:479-491. 1968a. Zwei homoeomorphe Arten der Bryozoengattung Columnotheca Marsson. Lethaia, 1:382—401, 9 figures. 1968b. On the Cretaceous Age of the So-called Jurassic Cheilostomatous Polyzoa (Bryozoa), A Contribution to the Knowledge of the Polyzoa-Fauna of the Maastrichtian in the Cotentin (Manche). Bul¬ letin of the British Museum (Natural History), Geology, 17:1-45, 8 plates, 2 figures. PLATES 1-17 54 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 1 (All figures X55; specimens coated with NH»C1) 1. —Ellisina brittanica (Brydone) : Frontal view of part of zoarial fragment encrusting basal side of cyclostome zoarium; zooecia have basal and frontal walls apparently uncalcified except for narrow gymnocysts; a small interzooecial avicularium is distal to each zoo- ecium; USNM 169491, Sample 9, Limhamn Mound II-Ni. 2. — Callopora sp.: Basal view of unilaminate fragment in part encrusting an echinoderm ossicle; basal and frontal walls of zooecia apparently calcified only peripherally so that extensive gymnocysts and small, rim-like cryptocysts (arrow) on proximal ends of frontal sides of zooecia are visible; USNM 169492, Sample 9, Limhamn Mound II-Ni. 3. —" Herpetopora” danica Lang: Frontal view of part of zoarium encrusting frontal side of cyclostome zoarium; basal and frontal walls of zooecia apparently uncalcified except for narrow gymnocyst; USNM 169493, Sample 14, Limhamn Mound II-Ni. 4. — Allantopora stomatoporoides Lang: Frontal view of small zoarium with ancestrula at lower center, slightly broken, encrusting smooth interior of gooseneck barnacle valve; basal walls of zooecia apparently uncalcified; frontal side of zooecia having extensive proximal gymnocyst supporting ring of spine bases (arrow) on mural rim; narrow, shelf¬ like cryptocyst within mural rim; USNM 169494, Sample 15, Limhamn Mound II-Ni. 5. — Fissuricella fissa (Voigt) : Frontal view of zoarial fragment encrusting frontal side of unidentified cheilostome; calcified frontal structure interpreted as a gymnocyst (Voigt 1959b: 260) shows extreme development; USNM 169495, Sample 6, Limhamn Mound II-Nl i Vi NUMBER 6 57 PLATE 2 (All figures X55; specimens coated with NPLCl) 1 .—Membraniporidra declivis (Marsson) : Zooecia on margin of narrow bilaminate fragment showing wide cryptocysts and small interzooecial avicularium with mandibular bar; USNM 196496, Sample 5, Limhamn Mound II-Ni. 2-4. — Pithodella cincta Marsson: 2, Frontal view of bilaminate fragment expanding distally from a subcylindrical proximal end; zooecia have broad cryptocysts and large adventitious avicularia on proximal gymnocysts; spine bases lacking; USNM 169497, Sample 1, Limhamn Mound II-Ni. 3, Frontal view of bilaminate fragment showing zooecia having gymnocysts with complete or partial rings of spine bases; cryptocysts narrow; avicularia lacking; USNM 169498, Sample 5, Limhamn Mound II-Ni. 4, Frontal view of uni¬ laminate fragment showing zooecia having scattered, isolated gymnocystal spine bases, narrow cryptocysts, and small avicularia; USNM 169499, Sample 17, Limhamn Mound II-Nl 58 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 3 (All figures X55) 1-5 .—Pithodella cincta Marsson: 1, 4, Longitudinal views of bilaminate fragments having zooecia with extensive gymnocysts of approximately same thickness as other zooecial walls; zooecium on right in 1 is ovicelled; USNM 169500, 169501, Sample 5, Limhamn Mound II-Ni. 2, 3, 5, Transverse views of bilaminate to subcylindrical fragments showing zooecia with continuous basal and lateral boundaries and gymnocysts continuous with lateral walls; USNM 169502-169504, Sample 17, Limhamn Mound II-Ni. 6, 7 .—Floridina gothica (d’Orbigny) : 6, Longitudinal view of subcylindrical fragment hav¬ ing zooecia with extensive, concave cryptocysts curved frontally at their distal ends; 7, transverse view of subcylindrical fragment showing radially arranged zooecia all reaching the zoarial axis and their basal walls forming a small, triangular hollow (arrow); lateral and basal boundaries between zooecia are continuous, and cryptocyst is continuous with lateral walls; USNM 169521, 169522, Sample 1, Limhamn Mound II-Ni. 8, 9.— Smittipora? prismatica (Hagenow) : 8, Transverse view of subcylindrical fragment showing radially arranged zooecia lacking basal walls; zooecia are excluded from zoarial axis in their proximal parts (indicated by thick cryptocysts) by widening of zooecia in adjacent rows; recurved distal end of cryptocyst (arrow) is shown in zooecium on left; 9, longitudinal view of subcylindrical fragment showing zooecia with cryptocysts descending steeply into distal part of cavity where their free ends are recurved proximally (arrow); exclusion of promixal ends of zooecia from zoarial axis is shown near bottom of view; USNM 169517, 169518, Sample 5, Limhamn Mound II-Ni . ^ fevl hi ' r F' Jk. . Rr 'J '■-iJj -. ,.Vi V - *^* " - V .^B ^4- - lUflp ''.; •’ w?i i ■ ’i-Jf*^' •'"■*• ■£%' s Q SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 6 61 PLATE 4 (All figures X55; specimens coated with NFLC1) 1. —Balantiostoma pallata Maryanska: Frontal view of unilaminate fragment having zooecia with convex, marginally perforate frontal shields which project over the orifice as mucrones; paired distal-oral spine bases are discernible on non-ovicelled zooecia; adventitious avicu- laria are present on some zooecia; USNM 169505, Sample 14, Limhamn Mound II-Ni. 2, 3.— Anornithopora minuta Voigt: 2, Frontal view of unilaminate fragment having zooecia with costal shields margined proximally by gymnocysts of varying widths; costae joined by numerous lateral fusions and fused medially in a narrow area; orifices with two to four distal spine bases; several interzooecial avicularia with spatulate rostra are present; USNM 169506, Sample 14, Limhamn Mound II-Ni. 3, Frontal view of unilaminate fragment having zooecia without gymnocysts; orifices with four or five distal spine bases; and an'avicularium; USNM 169507, Sample 9, Limhamn Mound II-Ni. 4. — Anornithopora polygona Voigt: Frontal view of zooecia encrusting a fragment of smooth mollusk shell and in turn encrusted by a cyclostome; costal shields lack gymnocysts and have narrow median areas of fusion; zooecium at center has broken ovicell; USNM 169508, Sample 4, Limhamn Mound II-Ni. 5. — Tricephalopora circumvallata (Levinsen): Frontal view of unilaminate fragment hav¬ ing zooecia with costal shields margined peripherally by a ridge-like tertiary frontal wall; costae joined only medially; remnants of peristomes bearing adventitious avicularia are present; zooecium on right is ovicelled; USNM 169509, Sample 9, Limhamn Mound II-Ni. 62 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 5 (All figures X55; specimens coated with NHiCl) la, b .—Aechmella pindborgi Berthelsen: la, Frontal view of two growing margins of same colony converging around encrusted, slightly grooved octocoral axis; zooecia and inter- zooecial avicularia (arrow) have incomplete cryptocysts limited to proximal and lateral margins; lb, frontal view of other side of same zoarium showing zooecia and avicularia with completely developed, concave cryptocysts and opesiae with proximolateral inden¬ tations; USNM 169510 Sample 9, Limhamn Mound II-Ni. 2, 3 .—Onychocella ravni Berthelsen: 2, Frontal view of dome-shaped zoarium reflecting the shape of the encrusted calcareous sponge; zooecia have concave cryptocysts and opesiae with proximolateral indentations; at the bifurcations of lineal series are curved, vicarious avicularia also with cryptocysts; 3, curved zoarium reflecting the shape of the corallite of an encrusted colonial scleractinian coral; USNM 169511, 169512, Sample 11, Limhamn Mound II-Ni. IWffil 1 3 m 9 LJ , J ^P'‘‘‘ ■»- I. n» .'IV Lft # MK v i 1 J • I P r ^ liH ■# B; - ■ j vjvv h // 4 I i a v x^i5i i 1 Em f K^PKTljWi^ hf\ j r ^ jrjs, 1 I 4f I "< - » B 4 1 _4 %hj_ 64 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY 3 4 NUMBER 6 65 PLATE 6 (All figures X55; specimens coated with NHiCl) 1-3 .—Micropora hennigiana Berthelsen: 1, Frontal view of small zoarium encrusting smooth interior of same barnacle valve as specimen shown on Plate 1:4; zooecia regularly arranged in triads around ancestrula (arrow), all having concave cryptocysts with opesiules proxi- molateral to opesiae; broken zooecium at bottom of view shows apparent lack of calcified basal wall and presence of dietellae; USNM 169513, Sample 15, Limhamn Mound II-Ni. 2, Frontal view of zoarial fragment encrusting basal side of Callopora sp., over¬ lapping its broken edge at upper right ; zooecia irregularly arranged on irregularly concave surface, one zooecium on upper left having its morphologic axis reversed; zooecium at extreme right has entozooecial ovicell (arrow); USNM 169514, Sample 11, Limhamn Mound II-Ni. 3, Frontal view of part of a large zoarium encrusting a fragment of ribbed brachiopod shell; ancestrula just below center (arrow); zooecia show regular growth parallel to underlying topographic pattern and irregular growth across it; USNM 169515, Sample 7, Limhamn Mound II-Ni. 4 .—Floridina sp.: Frontal view of part of a large zoarium encrusting a mollusk shell fragment; zooecia and large vicarious avicularia (at left center and upper right) have concave cryptocysts which, in avicularia, continue around distal end of opesiae; zooecial opesiae have broad proximolateral indentations; zooecia at upper right and at right center show entozooecial ovicells (arrows) ; a small, closed kenozooecium is present just below center; USNM 169516, Sample 11, Limhamn Mound II-Ni. 66 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 7 (All figures X55; specimens coated with NH 4 C1) 1, 5.— Smittipora? prismatica (Hagenow) : 1, Frontal view of proximal portion of subcyclin- drical fragment showing proximal zooecium with frontal side closed by a calcareous lamella except for a median slit and U-shaped opercular scar (arrow) with deep proximolateral pits: more distal zooecia show thick, but concave cryptocysts separated by distinct mural rims; 5, frontal view of less heavily calcified zooecia with more concave cryptocysts descending steeply toward opesiae; opesial indentations are apparently not differentiated from opesiae; USNM 169519, 169520, Sample 5, Limhamn Mound II-Ni. 2, 3.— Floridina gothica (d’Orbigny) : 2, Frontal view of subcylindrical fragment showing zooecia with concave cryptocysts and opesiules separated from opesiae by cryptocystal bars; USNM 169523, Sample 3, Limhamn Mound II-Ni. 3, Frontal view of subcylindrical fragment showing zooecia with opesiular indentations confluent with opesiae; three zooecia have entozooecial ovicells (arrows) ; USNM 169524, Sample 5, Limhamn Mound II-Ni. 4, 7.-— Onychocella? columella Berthelsen: 4, Frontal view of narrow bilaminate fragment showing zooecia having concave cryptocysts and opesiae with notched proximolateral corners; interzooecial avicalarium shows distinct mandibular bar; 7, frontal view of subcylindrical fragment showing zooecia (two of which have entozooecial ovicells), avicu- laria, and a kenozooecium (arrow) closed except for a central perforation; USNM 169525, 169526, Sample 5, Limhamn Mound II-Ni. 6.— Puncturiella sculpta (d’Orbigny): Frontal view of subcylindrical fragment showing zoo¬ ecia with concave, perforate cryptocysts, opesiules (arrow) proximal to opesiae } and small, interzooecial avicularia in lineal series with zooecia; USNM 169527, Sample 5, Limhamn Mound II-Ni. V i ■ |,-»u 1 Sk^v jfe. ,-^B <1* 1 'IU .1,1 |, | ^ WZfi. . PLATE 8 (All figures X55; specimens coated with NHiCl) 1-3. —Coscinopleura angusta Berthelsen: la, Frontal view of lateral margin of large, bilami¬ nate fragment showing zooecia having concave cryptocysts continuous around their opesiae, the proximolateral corners of which are notched; large, vicarious vibraculiform zooecia (coscinozooecia) having convex, perforated, probably gymnocystal proximal covers and asymmetrical distal openings are in a series on lateral margin of zoarium; grooves on zooecia at right may have resulted from predation on the soft membranes which presumably covered each zooecium; lb, frontal view of central part of same fragment showing grooves and two zooecia with entozooecial ovicells; USNM 169528, Sample 7, Limhamn Mound II-Ni. 2, Frontal view of small bilaminate fragment tapering toward proximal end and preserving growing margin at distal end; distalmost zooecium has incomplete cryptocyst; USNM 169529, Sample 14, Limhamn Mound II-Ni. 3, Frontal view of bilaminate frag¬ ment apparently from proximal end of large colony showing zooecia with greatly thickened cryptocysts but retaining distinct zooecial boundaries; USNM 169530, Sample 17, Lim¬ hamn Mound II-Ni. 70 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 9 (All figures X55) 1-3 .—Coscinopleura angusta Berthelsen: 1, 2, Transverse and longitudinal views of wide, bilaminate fragments apparently from distal parts of large colonies showing zooecia having moderately thick, concave cryptocysts; boundaries between zooecia for the most part are not discernible, as a result of alteration or of complete calcification of cuticles; minute open spaces in both views scattered randomly through zooecial walls have no external expression and seem to be products of alteration; 3, transverse view of narrow bilaminate fragment apparently from proximal part of large colony showing zooecia having enormously thickened cryptocysts; zooecial boundaries not discernible, but crude, coarse lamination near frontal surface may be remnant of original skeletal structure; shaft-like opesiae are shown in two zooecia on left, suggesting that some lophophores remained functional. USNM 169531- 169533, Sample, 17, Limhamn Mound II-Ni. NUMBER 6 71 2 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 6 73 PLATE 10 (All figures X55; specimens coated with NH 4 C1) 1 .—Pachythecella lundgreni (Pergens & Meunier): Frontal view of subcylindrical proximal portion of bilaminate fragment showing zooecia lacking boundaries and having thick, apparently imperforate frontal shields upturned around orifice as peristome; zooecium at lower left bears a. suboral frontal avicularium (arrow) ; USNM 169534, Sample 1, Limhamn Mound II-Ni. 2-4.— Porina salebrosa Marsson: 2, Frontal view of bilaminate fragment apparently from proximal part of colony, showing zooecia lacking boundaries and having thick frontal walls perforated by a large, circular ascopore (arrow) at about midlength and by smaller, scattered pores; portion of shield raised around orifice as a peristome bears numerous small, and rarer large avicularia; USNM 169535, Sample 14, Limhamn Mound II-Ni. 3, Frontal view of subcylindrical fragment showing zooecia with thick frontal shields which have apparently been abraded; 4, frontal view of bilaminate fragment preserving growing margin and showing zooecia having thin frontal shields with prominent ascopores, marginal pores at distinct interzooecial boundaries, and peristomial avicularia at varying stages of development; USNM 169536, 169537, Sample 1, Limhamn Mound II-Ni. 74 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 11 (All figures X55; specimens coated with NHtCl) 1-4. —Columnotheca cribrosa Marsson: 1, End view of growing tip of subcylindrical fragment showing radially arranged zooecia having transverse walls perforated by distal ends of interzooecial canals; USNM 169540, Sample 5, Limhamn Mound II-Ni. 2, Frontal view of branched fragment from distal part of zoarium preserving one growing tip showing verticillate arrangement of zooecia; zooecia have frontal shields lacking interzooecial boun¬ daries and bearing small scattered pores; near base of peristome are slightly larger spiramina, which in some zooecia in most fragments examined are paired (arrows), as typical for this species (Voigt 1968a:385); large, adventitious avicularia are near peristomes of few zooecia; USNM 169541, Boesdal Mound Ei. 3, Frontal view of branched fragment apparently from proximal part of zoarium; zooecia have few perforations except for spira¬ mina; avicularia are common; scattered, dash-shaped, shallow grooves arranged in branching uniserial patterns possibly are burrows of ctenostome bryozoans (J. D. Soule, personal communication, 1969) ; USNM 169542, Sample 1, Limhamn Mound II-Ni. 4, Frontal view of fragment from distal part of zoarium showing zooecia near growing tip with numerous small frontal pores and more proximal zooecia with progressively fewer pores; zooecium just below center on right has paired spiramina (arrows); dash-shaped grooves are common; USNM 169543, Boesdal Mound Ei. NUMBER 6 77 PLATE 12 (All figures X55) 1-4 .—Columnotheca cribrosa Marsson: 1, Longitudinal view of subcylindrical fragment showing zooecia having thick, convex frontal shields perforated by canal-like pores similar to those passing through thick transverse walls; middle zooecium on right has hyperstomial ovicell (arrow) completely hidden externally by development of peristome; USNM 169544, Boesdal Mound Ei. 2, Longitudinal view of ovicelled zooecium showing spiramen (arrow) perforating peristomial wall and opening internally distal to orifice; USNM 169545, Sample 1, Limhamn Mound II-Ni. 3, 4, Transverse views of subcylindrical fragments showing thin lateral zooecial walls with distinct interzooecial boundaries which disappear in the thick frontal walls; the section in one figure (4) passes in part through transverse walls; USNM 169546, 169547, Boesdal Mound El 5, 6 .—Porina salebrosa Marsson: Transverse views of bilaminate fragments showing zooecia having thin basal and lateral walls with distinct interzooecial boundaries and thick frontal walls lacking them; USNM 169538, 169539, Sample 14, Limhamn Mound II-Ni. 78 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 13 1, 2. —North wall of limestone quarry at Limhamn, Malmo, Sweden: 1, View from —40-m level, September 1964; dark vertical stripes are modern algal growths; from —60-m level to bottom of quarry, evenly bedded chalk with chert nodules is exposed; these sediments (belonging to Maestrichtian Stage) are marked at top (A) by major disconformity (Brot- zen 1959:15—17); above this level to about —50 m (B), bryozoan-rich limestones with numerous chert nodules form lower zones of Danian Stage as undulating beds which in places are truncated; from top of these undulating beds (Bioherm Group I of Brotzen) nearly to ground level, middle Danian bryozoan-rich limestones with well-developed mounds are exposed; these sediments include Brotzen’s Bioherm Groups II and III, the contact between which lies approximately at C; evenly bedded upper Danian limestones exposed above Bioherm Group III in the south wall of the quarry (Brotzen 1959:29) are not shown in this view; Mound Ni, which was sampled in detail, is exposed in lower half of Bioherm Group II and is cut by an access road at the —40-m level and another between the —40- and —60-m levels; Mound N 3 is in upper part of Bioherm Group II, above the —40-m level to left of view. 2, Contact between sediments transitional to core (A) and composing the east flank (B) of Mound II-Ni, middle Danian; transitional sedi¬ ments are poorly bedded, chalky limestones with sparse chert nodules, some of which appear as dark blobs near 16; flank limestones are light gray, have more abundant chert, and show finely nodular bedding; numbers 16 to 18 mark positions of samples (see Figure 12). SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY NUMBER 6 81 PLATE 14 1, 2.—North wall of limestone quarry at Limhamn, Malmo, Sweden: 1, Contact between core sediments (A) and sediments transitional to the east flank (B) of Mound II-N3, middle Danian. The core facies consists of indurated white lime-mud filling the space in and around colonial scleractinian corals nearly in position of growth; chert is lacking. The coral mass shown is about 4 m across. The original content of scleractinian skeletons, now either replaced by calcite or represented by molds (Brotzen 1959:21), is much less than in the coral limestone at Fakse, Denmark (Berthelsen 1962:23). Dark material at lower left is modem algal growth. 2, Contact between sediments composing west flank (A) and those transitional to core of Mound II-N 3 ; flank facies consists of gray limestone with abundant chert; transitional sediments show irregular, discontinuous bands of finely nodular bedding; thickness of internal shown, about 2 m. 3. —West wall of Limhamn quarry, just above —40-m level, showing middle Danian sediments of upper part of Bioherm Group II; bedded gray limestones with chert nodules and inter¬ beds intersect in pattern suggesting local unconformities; bedding surface A-B truncates beds on south flank of mound on lower right and appears to be concordant with beds on north flank of partly contemporaneous mound on left; zones of induration, such as have been reported at the contacts between overlapping mounds in the lower Danian of Stevns Klint, Denmark (Rosenkrantz and Rasmussen 1960:6, fig. 5), were not observed here. 82 SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY PLATE 15 1, 2.—Sediments of Mound II-Ni, middle Danian, Limhamn: 1, Flank-facies: gray, bryo- zoan-rich limestone with finely nodular bedding, west flank; limestone layer, only partly shown here, is about 1 m thick and occurs between two layers of chert nodules; Sample 3 was taken from this location (see Figure 12). 2, Transitional-facies: white, massive limestone with a thin, finely nodular bed containing numerous octocorals, just west of center of mound; Sample 7 was taken from this location (see Figure 12). 3.—North wall of Limhamn quarry, just above — 40-m level, showing coral-rich core facies (X) of one mound overlying flank limestones (Y) of another mound and grading vertically upward into bedded limestones (Z) like those on its own flanks; middle Danian, Bioherm Group II. f, v'qB'«<►.*Sl**<*. '^r’^wjl IMnWBiti > nnn ■jWEyasHc-aKtW,^''- *3*r:’ 5( v r v< t O r, v; i -y^fctV -:-«m. A',.f? - ~>cr.'-? •'I»£sir*w ■•* ' jPi j2gK|8®jigj vISPE^; i .>«ss «HE§S PwW* 1 ' in^Kl. ' Arf 9^P-* V t» W -^f *-. ‘rfc !_■».,£0’., -■ r ^-~' .. -vTA^Jfe .» * j|S?^gfr^£cyTftFJCBj^tfl/^ ' -- ••■--.■'■ ' ■' - - /,.■-■■ ' . v._;__ 1 r " mi i j jjf*—«* v '••*£ ill! V r~ /,y- l -"y H 'u/_ 'ji ■ ~^jS H»)p^SSS