‘— -‘ L.‘ l’ARY Mn ‘ ”gran State Uruvcrsity THE INTERRELATIONSHIP OF EARLY COLONY DEVELOPMENT, MONTICULES, AND BRANCHES IN PALEOZOIC BRYOZOANS BY Mark Edward Podell A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geology 1978 ABSTRACT THE INTERRELATIONSHIP OF EARLY COLONY DEVELOPMENT, MONTICULES, AND BRANCHES IN PALEOZOIC BRYOZOANS BY Mark Edward Podell The early colony development of Ordovician trepostome bryozoan initially consists of a triangular stage, the protoecial cone, followed by a reversal of budding direc- tion that produces a circular colony, the ancestrular disk, in which vestiges of the protoecial cone may be ob— served. Monticules, which are polymorphic clusters reg- ularly positioned over colony surfaces, reproduce the structure of the zone of early development, including a replicate of the ancestrula, the monarchozooid. A second type of ancestrular replicate, the basilozooid, is found within the axial zones (endozones) of colony branches. A damaged colonly, in which several monticules were de- stroyed by borings, shows disruptions of normal growth patterns in the areas affected. Morphogenetic gradients are developed around both the ancestrula and monticules and within branch axes. The early development of a colony has substantial effects on all later stages of growth. Any regulatory activity carried out by the ancestrula could have been carried out by the ancestrular replicates at multiple locations in all stages of colony growth. To Amy and my Parents ii ACKNOWLEDGEMENTS I would like to thank Dr. Robert L. Anstey for the many hours of assistance during this project, and to Dr. Gary R. Hooper, and Dr. J. Alan Holman for their sugges- tions on manuscript improvements. Special thanks to the entire staff at the Michigan State University Center of Electron Optics for their instruction on current scanning electorn microscope, and dark room techniques so useful in this study. iii TABLE OF CONTENTS LIST OF TEXT-FIGURES LIST OF PLATES INTRODUCTION MATERIALS AND METHODS. EARLY COLONY DEVELOPMENT DEVELOPMENT OF MONTICULES. DEVELOPMENT OF BRANCHES. CONCLUSIONS. REFERENCES iv Page vi 18 26 33 35 LIST OF TEXT-FIGURES Text-Figure 1 Comparison of the ancestrula, monarchozooid, and basilozooid Dominance effects in the early development of Prasopora conoidea Ulrich Development of monticules in the early astogeny of Prasopora conoidea Ulrich. . . . . . Subcolony development in Prasopora simulatix Ulrich. Development of colony branches in the astogeny of Hallopora spp. Page 11 23 25 31 LIST OF PLATES Plate 1 Scanning electron micrographs development 2 Scanning electron micrographs development 3 Scanning electron micrographs development A Scanning electron micrographs and branch axes of of of of 5 Light micrographs of the initial monticules and branch axes. vi early colony early colony early colony monticules region, Page 13 16 2O 27 INTRODUCTION The work of Cumings (1904; 1905; 1912) on early stages of colony growth firmly established the zoological af- finities of the major Paleozoic bryozoan groups, many of which had previously been considered corals. This subject received no further attention, except for a brief reference by Borg (1965) in a posthumous publication, until the re- cent papers by Corneliussen and Perry (1973), Boardman and McKinney (1976), and McKinney (1977; 1978), who placed early colony development clearly within the context of the overall analytical and functional morphology of several genera. In addition, recently renewed interest in the doc- umentation of the evolutionary patterns of paedomorphosis and recapitulation (Gould 1977) makes it necessary that complete development sequences be available for phylogene- tic analysis. Furthermore the recognition of morphore- gulatory variation in some graptolite lineages (Urbanek 1960; 1973), suggests the possiblity that bryozoans, which have analogous colony development, might display similar phylogenetic patterns. The common implication of all the above is that early colony development must not only be better known, but must be related both to later develop- mental stages and to phylogeny. This paper seeks to ac- complish the first two of these three tasks. 1 Cuming's earlier papers (190“; 1905) compared the larval development and early astogeny of Recent bryozoans with those of the Paleozoic families Fenestellidae and Palescharidae. In his study of the early development of several genera (1912), he settled the issue of the system- atic position of the Trepostomata. Borg (1965) briefly compared the early colony development of the Paleozoic genus Prasopora to that of modern cyclostomes. Cornelius- sen and Perry (1973) compared the initial region of the Silurian species Hallopora elegantula to that of the Ordovician H. dalei figured by Cumings. Boardman and McKinney (1976) provided a detailed comparison of the of the Paleozoic genus Rhombotrypa with that of Recent lichenoporid cyclostomes described by Harmer (1896) and Borg (1926; 1933). In addition, McKinney (1977a; 1977b; 1978) has incorporated early colony development into his overall studies of the functional morphology of Paleozoic lyre-shaped and paraboloid-based bryozoan colonies. This paper will provide additional details of early colonly de- velopment in a variety of Ordovician trepostomes, including scanning electron micrographs of important developmental stages, and will also attempt to show that the early de— velopmental stages, in slightly modified form, are repro- duced repeatedly in all later stages of colony growth. Urbanek's work (1960; 1973; reviewed by Gould 1977), on morphogenetic gradients in graptolite colonies and their phylogentic modifications, provides an intriguing example of howtflmaeffects ofa growth regulator, analogous to auxins in plants, can be documented in the fossil record by means of "natural experiments" on damaged and regener— ated colonies. This paper will illustrate an analogous situation in a damaged bryozoan colony in which normal growth patterns have been disrupted in the damaged area. In reference to Urbanek's work, however, Boardman and Cheetham (1973) thought it highly unlikely that a morpho- genetic substance produced by the primary zooids of a bryozoan colony could be continuously diffused throughout all of colonly growth, primarily because of the limited size and apparently brief duration of a colony's initial region. The present study will illustrate, however, that the colony'sfounder zooid is regularly reproduced at mul- tiple locations throughout all stages of colony growth, and that any morphogenetic activity carried out by the founder zooid was likewise reproduced by all of these secondary founders, or monarchic zooids. Morphogenetic gradients in an Ordovician bryozoan colony were mapped by Anstey et a1. (1976), who inferred from them the activity of a morphogentic substance anal- ogous to that described by Urbanek. Similarly located physiologic gradients were measured in living cheilostome bryozoans by Bronstein (1939). In addition, Dzik (1975) believed that trends in cheilostome phylogeny reflected the modification of morphoregulatory substances. This study will illustrate the details, using scanning electron microscopy, of the monarchic zooids found at the origin of morphogenetic gradients in later stages of the same colonies in which early development was also analyzed. Some preliminary details of this work have been presented in abstract and letter form by Anstey and Pachut (1977) and Anstey et al. (1978). MATERIALS AND METHODS This study will emphasize the use of scanning elec- tron microscopy (using an 181 Super III) to illustrate the details of all stages of colony development, based on both external colony surfaces and etched serial polished sec- tions. Scanning electron micrographs provide better re- solution of both skeletal wall structure and external morphology, especially of minute colonies, than that pos- sible with light microscopy. Specimen materials include 120 zoaria or brachiopod shells on which multiple zoaria were encrusted, of Middle and Late Ordovician age from Minnesota and the Ohio Valley respectively. All specimens were ultrasonically cleaned and "sputter" coated with gold prior to microscopy. Polished sections were briefly etched with formic acid (5-10 seconds). All micrographs were prepared at a stan— dard beam orientation of 90 degrees to the microscope stage, thereby eliminating distortion due to foreshorting. EARLY COLONY DEVELOPMENT The growth of a stenolaemate bryozoan colony begins with the settlement of the founder zooid, the ancestrula. Initially the founder zooid grows parellel to the substrate, forming a tubular chamber constructed of external simple skeleton (Boardman and McKinney 1976) termed the protoecium (the zooid producing the protoecium is termed by Ryland (1970) the proancestrula). The proancestrula initially grows distally (along the substrate), but eventually turns up- ward (anteriorly) to become the ancestrula proper (orien- tation terminology from Gautier 1970). The protoecium may be separated from the ancesturla by a slight constriction, as in the genus Prasopora (Cumings 1912), or by a diaphragm, as in the genus Rhombotrypa (Boardman and McKinney 1976). In most Ordovician trepostomes there is no demarcation be- tween the two, as in the genus Hallopora (P1. 1, fig. A). The external simple skeleton of the proancestrula appears to differ crystallographically from the skeleton of subsequently formed parts of the colony. In what are most likely trepostome protoecia encrusting on Rafinesqunina valves, the external surface of the skeleton consists of a fanlike cluster of proximally radiating elongate crys- tal units with sutured margins between crystals (Pl. 1, figs. 1-3; P1. 2, fig. 1). As seen in etched sections, however (Pl. 1, figs. 4—6), the protoecial wall, although thinner, is constructed of multiple laminations identical to those in the wall of subsequently budded zooids. 5 Fig. Pig. Fig. Fig. Fig. Fig. EXPLANATION OF PLATE 1 l. Trepostome protoecium. iichigan State Uni~ versity, 220314-0002Aa. Scanning electorn micro- graph, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 360. 2. Trepostome protoecium. Michigan State Uni- versity, 220314-0002Aa. Enlargment of upper right- hand portion of Fig. 1, rotated 900 to the left, illustrating elongate crystal units with sutured margins, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 2400. 3. Trepostome protoecia. Michigan State Uni- versity, 220314-00024b. Scanning electron micro- graph, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 180. A. Hallopora sp. Michigan State University, 220315400090. Scanning electron micrograph of a longitudinal section through the protoecium, ancestrula and associated thickened backwall, Madison, Indiana, Dillsboro Formation (Late Ordo- vician), X 200. 5. Amplexopora ? sp. Michigan State University, 2203IH-00027a. Scanning electron micrograph of a transverse section through the colony base il- lustrating the early stages of backbudding and skeletal wall structure, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 60. 6. Trepostome ancestrula. Michigan State Univer- sity, 220314-0002Ac. Scanning electron micrograph of an etched and polished section through the protoecium, ancestrula, and primary zooidcfi‘thesame colony as illustrated in P1. 2 fig. 6 showing de- tails of wall structure, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X A00. PLATE 1 7 The first generation of additional zooids is produced by asexual budding from the distal side of the ancestrula near the substrate (Pl. 1, fig. A). The primary zooids are those in contact with the distal wall of the ances- trula. Rhombotrypa has four primary zooids (P1. 2, figs. 3, 5), whereas Hallopora (Pl. 5, fig. 2), AmpleXOpora ? (P1. 1, fig. 5; Pl. 3, fig. 2), Homotrypa ? (text-fig. 1a; Pl. 3, figs. 3, 5),Prasopora (text-fig. 2), and 13 unidenti- fiable early colonies (Pl. 2, fig. 6; P1. 3, figs. 1, A) have three. The second generation of zooids (secondary) buds distally from the primaries, and the third generation buds likewise from the secondaries, producing an initially triangular colony (Pl. 2, fig. 6) with the protoecium at its apex, termed the protoecial cone by Boardman and McKinney (1976). The shape of the protoecial cone is gov- erned by the number of newly budded zooids produced per generation. Colonies with only a small increase in each generation form long isosceles triangles (Pl. 3, fig. 1), whereas an accelerated budding rate quickly enlarges the distal margin to produce subcircular colonies (Pl. 1, fig. 5). The latter type of development leads to the proximal displacement of the lateral sides of the growing margin, so that after only a few generations, zooids have filled in the region behind the ancestrula and form small circular colonies (P1. 3, fig. A). The proximal addition of later generation zooids that fill in the region behind the TEXT—FIG. 1. Comparison of the ancestrula, monarchozooid, and basilozooid. A, An ancestrular cluster with the re— gion of the protoecial flange (shaded), ancestrula (label— ed A), primary zooids (labeled 1—3), secondary zooids (labeled a-d), and backbudded zooids (labedled e and f), enlarged from P1.3, fig. 3. B, A monticule including the monarchozooid (labeled M), triangular central cluster of mesopores (shaded), and primary, secondary and backbudded zooids (labeled as in A), from P1. A, fig. 3. C, A branch axis (from P1. u, fig. 1, rotated 180°) including the basilozooid (labeled B) and the ring of zooids (shaded) that originated from it; this basilozooid is histological- 1y continuous with the ancestrula; original primary and secondary zooids labeled as in A and B, all X 18. lO ancestrula will be referred to below as backbudding. In the development of the protoecial cone, some taxa produce more buds on one side of the growing margin than the other, producing somewhat spiral protoecial cones that may be described as righthanded (P1. 2, figs. 3, A) or lefthanded colonies (P1. 5, fig. 2). At some early stage in the development of the protoe- cial cone, the proximal wall of the ancestrula and its laterally adjacent primary zooids becomes abnormally thick- ened, forming a V—shaped "flange" around the apex of the triangular colony (Pl. 2, fig. 6), the lateral ends of which extend as processes protruding from the colony mar- gin (Pl. 2, fig. 5). The apical flange does not overlap the protoecuim, which still remains visable at this stage of development, projecting proximally form the center of the flange. After backbudding has filled in the space behind the ancestrula, the flange remains observable as a V-shaped region of thickened wall (text-fig. la; Pl. 3, figs. 3, 4) within the early colony, and its lateral edges may project into some zooecia (Pl. 2, fig. 5). This struc- ture is particularly useful in identifying the ancestrula and the primary zooids in both early and somewhat later stages of colony growth (Pl. 3, figs, 2-5). This thick- ened wall region was observed in the early development of four trepostome genera by Cumings (1912) and in two addi— tional genera by Boardman and McKinney (1976) and McKinney (1977b). Scanning electron micrographs of this thickened .- a: _=) .3 ‘-e§“f¢9hor’a? Vb“ «ye-"$421“: if.“ C: ) ‘- 4: 5 : '- . I r ". 1. o. . a; ‘3: ‘9. .0 *2 I. '0 TEXT-FIG. 2. Dominance effects in the early development of Prasopora conoidea Ulrich.r Michigan State University 220323-00001. A, Transverse section cut at the colony base illustrating the radial disposition of cystiphragms and negative morphogenetic gradients centered on the ancestrula (shaded dark grey). Monarchozooids are shaded light grey, and zooids in which the cystiphragms point away from the ancestrula are colored solid black. B, Transverse section approximately 0.5 mm above the colony base showing the increase in the number of cystiphragms that have become reoriented around their respective mon- ticules and the increase in the number of small zooids within the ancestrular and monticular clusters. C, Trans- verse section approximately 1.0 mm above the substrate showing the continuation of cystiphragm reorientation and the distal spread and enlargment of the clusters of small zooids, all X 18. ll 12 region viewed longitudinally (Pl. 2, fig. 2) and trans- versely (Pl. 2, figs, A, 5) indicate that the small crystals making up the wall laminae change from a vertical orienta- tion on its distalmargin to horizontal in the center to vertical again on its proximal margin, demonstrating that the wall laminae are folded over in a proximally developed flexure. Boardman and McKinney (1976) attributed an iden- tical development (a double layer of external simple skeleton on the proximal side of the ancestrula) in Rhombotrypa to a proximal flexure of the external colony wall down to the substrate. The thickened flange and its lateral projections could have been of functional impor- tance to the early colony, by creating turbulence in its wake and there by improving the filter—feeding ability of the early zooids. Following the development of the protoecial flange, backbudding is accelerated at the expense of distal budding, so that the colonies become subcircular with the ancestrula centrally located (Pl. 3, figs. 2-5). This stage of de- velopment is termed herein the ancestrular disk. The loci and rates of subsequent budding determine the ultimate colony growth habit: peripheral budding produces sheet— like colonies (and variations thereof), whereas distal ex- tension of the internal zooecia and internal budding pro- duces the series mound to hemisphere to pillar to branching colony forms. The development of morphogenetic gradients within the protoecial cone is variable. In both Rhombotrypa and EXPLANATION OF PLATE 2 Fig. l. Trepostome protoecium. Michigan State Univer- sity, 2203lA—0002Ab. Enlargement of protoecium on left in Pl. 1, fig. 3, illustrating the radial ori- entation of crystal units, Versailles, Indiana, Dil- lsboro Formation (Late Ordovician), X 630. Fig. 2. Hallopora sp. Michigan State University, 22031A-00090. Enlargement of thickened region on the left side of the ancestrula and protoecium of Pl. 1, fig. A (reversed image), illustrating wall structure within the protoecial flange, Madison, Indiana, Dil— lsboro Formation (Late Ordovician), X 1000. Figs. 3-5. Rhombotrypa sp. Michigan State University, 2203lA-00030. Versailles, Indiana, Dillsboro For- mation (Late Ordovicain). 3, Scanning electron micro- graph of a tangential section through the colony base showing a positive morphogenetic gradient,X 20. A, Enlargement of initial region of fig. 3 displaying four primary zooids, ancestrula, and thickened wall, X 180. 5, Enlargement of lower right hand corner of fig. A, illustrating the flange of the thickened wall protruding into the right hand most primary zooid, X 360. Fig. 6. Trepostome protoecial cone. Michigan State University, 2203lA-0002Ac. Scanning electron micro- graph of early astogeny illustrating the external wall of the protoecium, thickened backwall, protrud- ing flange, and primary zooids, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 100. 13 PLATE 2 \i- new ’2 “ fi” \ “‘3, ’ v u 15 Hallopora (Pl. 2, fig. 3; Pl. 5, fig. 2), the ancestrula is the smallest zooid inthe early colony, and subsequent gen- erations increase in size away from it, producing a posi- tive morphogenetic gradient. In the earliest stage of Prasopora conoidea (text-fig. 2a),however, the ancestrula is the largest zooid, and subsequent generations decrease in size away from it, producing a negative morphogenetic gradient. Available data indicate, however, that all posi— tive gradients disappear by the stage of the development of the ancestrular disk. All available ancestrular disks display either no obvious gradients (as in Rhombotrpya), or well developed negative gradients leading away from the ancestrula (as in most of the taxa studied). This suggests that some colonies experienced a developmental change from positive allometry in the protoecial cone to negative allometry in the ancestrular disk. The dominance of the ancestrula within the ancestrular disk is illustrated by the orientation of cystiphragms in monticuliporid genera (Pl. 3, figs. 3, 5; text—fig. 2a). In addition to their radial development of morphogenetic gradients, the cystiphragms in each zooecium are radially aligned on the ancestrula. Identical radial alignment of cystiphragms is developed around the monticules of monti- culiporids (Boardman and Utgaard 1966). Furthermore, the ring of zooids immediately surround— ing the ancestrula in many colonies becomes differentiated from the other zooids of the disk to form the first EXPLANATION OF PLATE 3 Fig. 1. Trepostome protoecial cone. Michigan State University, 2203lA-000A6a. Scanning electron micro- graph of a colony with a large number of distal zooids; protoecium is in the upper center of the mi- crograph, Madison, Indiana, Dillsboro Formation (Late Ordovician), X 75. Fig. 2 Amplexopora ? sp. Michigan State University, 2203lA-00027b. Scanning electron micrograph of the ancestrular disk with its centrally located ancestrula and protoecial flange, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 35. Figs 3, 5. Homotrypa sp. Versailles, Indiana, Dills- boro Formation (Late Ordovician). 3, Michigan State University 2203lA-00028a. Scanning electron micro- graph of the ancestrular disk illustrating the ances- trula (distal to thickened wall area) surrounded by cystiphragms oriented towards it, X 35. 5, Michigan State University, 22031A—0028b. Scanning electron micrograph of a ancestrular disk illustrating similar features to those in fig. 3, X 35. Fig. A. Ancestrular disk. Michigan State University, 2203lA—00027c. Scanning electron micrograph illus- trating the thickened wall around the ancestrula and negative morphogenetic gradient in peripherally spreading zooids, Versailles, Indiana, Dillsboro Formation (Late Ordovician), X 25. 16 I JJ/thfl ‘ ‘5‘ J. . i :: 3%!" ._ 3, '4'" I I Y " , C l8 monticule (polymorphic cluster) of the early colony (text- fig. 2b). CommOle the primary and secondary zooids of the protoecial cone (still observable because of the thick- ened wall area of the apical flange) become as large as the ancestrula, and many small newly budded zooids pro- duced from the distal side of the ancestrula displace the large zooids away from the ancestrula, so that a ringlike structure is developed (text—fig.2b, c). Most trepostome monticules include a central cluster of abnormally small zooids (Pl. A, fig. 3, A). With the appearance of the first monticule, early colony development is complete. DEVELOPMENT OF MONTICULES The second stage of colony development involves the differentiation of additional monticules as the colony grows peripherally. New monticules appear at regular dis— tances from the ancestrular cluster (the original monti- cule) and from each other. In Prasopora conoidea (text-fig. 2a) new monticules arise at regular intervals of approx— imately one cm. The key developmental aspect of monticules is their nearly exact duplication of the ancestrular clus- ter, complete with a replicated ancestrula, primary and secondary zooids, and vestiges of the protoecial cone. In addition they display dominance effects over their region of the colony (which incorporates about 200 or so extra- monticular autozooids) identical to the dominance of the ancestrula over the ancestrular disk. Large sheetlike l9 colonies (or the exozonal surfaces of monticulated colonies having other growth forms) are in fact simple aggregates of hexagonally arranged monticular subcolonies that each replicate the ancestrular disk. Because each subcolony is developed within the spatial constraints imposed by neigh- boring subcolonies, their boundaries cannot be circular like that of the ancestrular disk, but become nearly hexagonal (Anstey et al. 1976). The dominance effects displayed by monticules are shown by their location at the center of radially developed morphogenetic gradients with- in subcolonies and in certain taxa, radial alignment of cystiphragms or lunaria. In the ancestrular cluster the backbudded zooids (e and f in text-fig. la) are separated from the ancestrula and the primary zooids (A, and 1, 2 and 3) by the thick- ened wall region that remains from the apical flange of the protoecial cone, thus preserving the triangular structure of an earlier stage of develOpment. In monti- cules the ancestrular replicate, the monarchozooid (Anstey and Pachut 1977; Anstey et al. 1978), is likewise at the apex of a triangular structure formed by the sec- ondary zooids and central cluster of small zooids (text- fig. lb; text-fig. 3; Pl. A, figs. 3, A), although in monticules there is no V-shaped region of thickened wall. In monticules, very small zooids (probably polymorphic) are budded from the distal side of the monarchozooid, and form a central cluster that is generally triangular EXPLANATION OF PLATE A Fig. l. Hallopora sp. Michigan State University, 2203lA-00052. Scanning electron micrograph of a transverse section through a pillar shaped colony displaying an enlarged, centrally located basilozooid surrounded by several new zooids (rotated 1800 from text-fig. 1c), Madison, Indiana, Dillsboro Formation (Late Ordovician), X A0. Fig. 2. Homotrypella hospitalis (Nicholson). Mich- igan State University, 22031A-00055. Scanning elec- tron micrograph of a transverse section through a branch illustrating the centrally located enlarged basilozooid encircled by newly budded zooids, Madison Indiana, Dillsboro Formation (Late Ordovician), X 30. Fig. 3. Hallopora sp. Michigan State University, 2203lA-00019. Scanning electron micrograph of a monticular polymorphic clusterillustrating the monarchozooid (below central cluster of mesopores) and median primary zooid (above central cluster of mesopores), West Harrison, Indiana, Eden Shale (Late Ordovician), X 35. Fig. A. Peronopora sp. Michigan State University, 2203lA-00017. Scanning electron micrograph of a monticule showing the monarchozooid (very bottom of micrograph) and the median primary zooid (center of micrograph at upper end of central cluster of meso- pores), West Harrison, Indiana, Eden Shale (Late Ordovician), X 35. 2O PLATE 4 .d A S. r195... , N974. Two 0.05. one «a a u e. . 6.. 3m 22 in shape (shaded region in text-fig. 3). In the ances- trular cluster, the small zooids displace the primary zooids distally, as Cumings (1912) illustrated in Prasopora con- oidea. In colonies with three primary zooids, the large zooid opposite the monarchozooid and separated from it by the small zooids in a monticule is a replicate of the median primary zooid of the early colony. The central cluster of small zooids commonly bifurcates around the median primary, thus forming a U-shaped structure on the distal side of the monticule and imparting to the mont- icule and additional element of bilateral symmetry. In many monticules the small central zooids increase in size from the monarchozooid to the median primary zooid (Pl.5, figs. 1, 3) suggesting that continued budding within the monticule leads to the formation of a succession of median primaries, with the older ones displaced into the mont- icular border. This size gradient within the monticule itself duplicated the positive morphogenetic gradient ob- served in the protoecial cones of Hallopora and Rhombotrypa (Pl. 5, fig. 2; P1. 2, fig. 3). In the genus Hallopora the median primary zooid of monticules is particularly distinctive because it is floored by a very shallow dia- phragm (Pl. A, fig. 3). In most taxa the large monticular polymorphic zooids form a complete ring around the central cluster of small zooids. The monarchozooid is usually located inside the outer ring of large polymorphs slightly into the small cluster of small zooids (text-fig. lb), TEXT-FIG. 3. Development of monticules in the early asto— geny of Prasopora conoidea Ulrich. Michigan State Univer- sity, 220323-00001. A three dimensional reconstruction of the central portion of the colony base (approximately 1.5 mm high) form serial acetate peels illustrating the budding of small zooids (shaded light grey) from the ancestrula (shaded dark grey) and from two monarchozooids (shaded medium grey), Cannon Falls, Minnesoa, Decorah Shale (Middle Ordovician), X 18. 23 2A whereas the median primary zooid is usually well within the central cluster, and may be completely surrounded by small zooids. Outside the monticule the nonpolymorphic autozooids gradually decrease in size in a negative grad- ient extending from the monticular border to the subcolony boundary. In the family Monticuliporidae cystiphragms are gen- erally radially arranged around the monticules (Boardman and Utgaard 1966). In some genera the cystiphragms point towards the monticule, as in Prasopora, but in others away from it, as in Monticulipora. These arrangements suggest the developmental dominance of the monarchozooid over its own subcolony. Subcolony maps can be prepared by drawing their boundaries at the places where cystiphragms reverse their orientation. The dominance of the monarchozooid over these patterns is confirmed by observations of areas damaged by boring organisms in Prasopora simulatrix (text- fig. A). In undamaged subcolonies the radial alignment of the cystiphragms is well developed, but in the adjacent regions affected by the borings, no monticules are obser— vable and the cystiphragms are randomly oriented. Pre- sumably the loss of the monarchozooids to the borers re- sulted in a local absence of developmental regulation. This example provides additional confirmatitw1of the hypo- thesisradvancedelsewhere (Anstey et a1. 1976, 1978; Pachut 1977) that bryozoan subcolonies were regulated by the diffusion of a morphogenetic substance released from the monarchozooid. TEXT-FIG. A. Subcolony development in Prasopora simula— trix Ulrich. Michigan State University, 220317—00009. Subcolony boundaries (solid white lines) based on the orientation of cystiphragms around individual monticular centers (shaded verylight grey) and monarchozooids (black dots). The apparently irregular nature of the subcolony boundaries is an artifact caused by locally random orientation of cystiphragms within the boundary region. Where monarchozooids have been removed by borers (areas labedled B), cystiphragms are randomly oriented, Frankfort, Kentucky, "Trenton" Formation (Middle Ordovician), X 18. 25 26 The post—ancestrular disk development of Prasopora conoidea documents the transfer of local dominance from the ancestrula to the newly differentiated monarchozooids in the peripheral regions of the colony. Serial sections taken from the lowest level within the colony show that the cystiphragms are not initally radially aligned on their local monticules, but are all pointing towards the ances- trula. Higher sections show that the zooids nearest each monticule are the first to turn away from the ancestrula and towards the local monarchozooid, and that the realign- ment progresses as "waves" that spread radially from the monticular centers. At the height of 1mm above the sub— strate, all of the zooids have become incorporated into local subcolonies, each dominated by a monarchozooid and the only zooids still aligned with the ancestrula are those in its local subcolony (text-fig. 2c). This evidence strongly suggests that zooid alignment is regulated by a morphogenetic substance produced initially by the ances— trula and diffused outward into the colony. Subsequently newly differentiated monarchozooids begin to produce the same substance which gradually increases in concentration radially away from each, thereby spreading the zones of local dominance until the early colony is completely sub- divided into such zones. DEVELOPMENT OF BRANCHES Cumings (1912) suggested that the development of colony branches could be related to that of monticules, EXPLANATION OF PLATE 5 Fig. l. Amplexopora septosa (Ulrich). Indiana Univer- sity, 8979—17005. Light micrograph of a tangential section through a monticule showing the monarchozooid (centrally located in lower half of micrograph) and enlargement in new zooids as they are displaced dis— tally away from it, Miamitown, Ohio, Eden Shale (Late Ordovician), X A0. Fig. 2. Hallopora dalei (Milne-Edwards & Haime). Indiana University, 9106-23. Light micrograph of a transverse section through the colony base illustra— ting the positive morphogentic gradient radiating away from the centrally located ancestrula, near Guilford, Indiana, Dillsboro Formation (Late Ordovi- cian), X 35. Fig. 3. Heterotrypa ulrichi (Nicholson). Indiana University,g8976-25007. Light micrograph of a tan- gential section through a monticule illustrating the distal spread and enlargement of new zooids from the monarchozooid, Miamitown, Ohio, Eden Shale (Late Ordovician), X A0. Fig. A. Hallopora nodulosa (Nicholson). Indiana University, 897A-15005. Light micrograph of a trans— verse section through a branch showing the enlarged, centrally located basilozooid and negative morpho- genetic gradient radiating away from it, Gallatin County, Kentucky, Eden Shale (Late Ordovician), X 35. Fig. 5. Hallopora ramosa (D'Orbigny). Indiana University, 900A-l3. Light micrograph of a trans- verse section taken through a branch displaying the centrally located basilozooid and negative morpho- genetic gradient radiating away from it, near Guilford, Indiana, Dillsboro Formation (Late Ordovician), X 35. Fig. 6 Eridotrypa simulatrix (Ulrich). Indiana University, 9135-IA. Light micrograph of a transverse section cut through a branch illustrating centrally located, enlarged basilozooid, near Guilford, Indiana, Dillsboro Formation (Late Ordovician), X 35. 27 29 and Blake (1976) effectively illustrated the developmental interrelationship of branches and monticules in the genus Rhabdomeson. Branching colonies, however, commonly have two developmental zones: an inner (axial) zone of long, vertically oriented thin—walled undifferentiated zooecia, termed the endozone, and an outer (cortical) zone of short, thick-walled zooecia with a variety of intrazooecial and extrazooecial structures, the exozone. Even massive and hemispherical colonies may have cyclic repetitions of laminar growth zones that vary from endozonal to exozonal characteristics. Monticules are developed only within the exozone, and the regulation of exozonal characteristics is related to the morphogenetic gradients associated with monarchozooids. This study seeks to demonstrate that a second type of ancestrular replicate, the basilozooid (Anstey and Pachut 1977; Anstey et a1. 1978) is found with in the axial zones of colony branches, and may be involved in the regulation of endozonal development. After the stage of the ancestrular disk, some colonies grow upward from the substrate by vertical extension of the zooecial tubes instead of lateral extension of the colony by marginal budding. Such colonies initially become a hemispherical mound two or three mm high, which subsequent- ly becomes attenuated into a pillar. The region of upward growth is centered on the ancestrula, which continues up the branch axis as a centrally located zooid (text-fig. 5). This zooid, the basilozooid, is histologically continuous 30 with the ancestrula, is centered within the endozone, and exozonal differentiation occurs at uniform distances from it. In the genus Hallopora (text-fig. 5), new buds devel- op on all sides of the basilozooid, and gradually displace the original primary and secondary zooids outwards by new- er buds so that the new buds collectively form an inverted axial cone centered on the basilozooid that expands in diameter up the branch (text-fig. lc, 5a). A hemispherical colony of Homotrypella hospitalis (Pl. A, fig. 2) produced a small pillarlike protuberance on the margin of the colony. Within this structure, which resembles an early stage in branch development, a large axial zooid is not only centrally located, but the cysti- phragms of all the surrounding zooids are radially aligned with it. Large axial zooids are present in the endozones of many Paleozoic branching colonies (P1. A, figs. 1, 2; Pl. 5, figs, 3-5), which are probably basilozooids. Unusually large axial zooids are also present in a number of post- Paleozoic cyclostomes (Nye 1976). In addition to large size, central or near central location, vertical continuity within the branch axis, and location at the origin of morphogenetic gradients in size, shape, zooecial structure, and budding, are general characteristics of basilozooids. The processes leading to branch bifurcation are incom- pletely known, but new branches may arise from an expansion of the axial endozone and the differentiation of a second TEXT-FIG. 5. Development of colony branches in the astogeny of Hallopora spp. longitudinal section re- constructed from serial acetate peels. A, Michigan State University 2203lA-00089. A colony branch il— lustrating the centrally located basilozooid (shaded dark grey), zooids originating directly from it (shaded light grey), and zooids originating from other parts of the endozone (shaded medium grey), West Harrison, Indiana, Eden Shale (Late Ordovician). B, Michigan State University, 2203lA-00052. Pillar- like colony illustrating the continuation of the ancestrula as a basilozooid (shaded dark grey), and budding of new zooids (shaded light grey) from the basilozooid. Transverse, sections El and B illus- trate the basilozooid (dark grey) and buds origina- ting from it (light grey) at two levels within the colony, Madison, Indiana, Dillsboro Formation, (Late Ordovician). C, Michigan State University, 22031A- 00067. Longitudinal view of pillarlike colony with similar additions of new zooids to that of A and B. Transverse section Cl illustrates a very early stage in the budding of new zooids from the basilozooid, Miamitown, Ohio, Eden Shale, (Late Ordovician), all X 17. 31 ”my—Faw- :- . _~ am an; n “u v I ____‘in£' .afikn. .. rift“. .- .- -.. ..,. - hf‘y‘ 5Q" TEXT—FIGURE 5 32 / / . / 33 basilozooid, just as new monticules develop as a distance effect with respect to previous ones. A transverse section of Leptotrypella pellucida (not illustrated) has two sepa- rate basilozooids in an expanded endozone, suggesting that branch bifurcation is a result of theduplication of the axial monarch. Some branching and frondescent colonies have small warty protuberances capped by a monticule, suggesting that some branches may have developed directly from a mon— ticule. Whether or not the development of a branch can cause absorption of previously exisiting exozonal skeleton is presently unkown. More complex developmental processes than those described above must have been involved in the growth of frondescent and anastomosing colonies. CONCLUSIONS Early colony development, as observed in Ordovician trepostome bryozoans, consists initially of distal budding from the ancestrula along the substrate, forming a flat triangular colony in which a positive morphogenetic grad- ient may be present. The completion of backbudding forms a circular colony, the ancestrular disk, characterized by negative morphogenetic gradients leading away from the ancestrula. Subsequent growth may be marginal, by means of budding along the substrate, and/or upward, by vertical extension of the colony center. Colony surfaces deve10p regularly positioned polymorphic clusters, or monticules, in which the monarchozooid replicates the ancestrula and 3A maintains negative morphogenetic gradients to the margin of the monticular subcolony. Branch axes also contain an ancestrular replicate, the basilozooid, which may maintain morphogenetic gradients within the axial endozone. From these observations it may be concluded that the early develOpment of a colony has a substantial effect on all later stages of colony growth. The common development of morphogenetic gradients originating from ancestrular re- plicates indicates that this phylum might well be a good one in which to investigate heterochronous patterns of evolution caused by variation in morphoregulatory factors. LIST OF REFERENCES REFERENCES Anstey, R. L. and Pachut, J. F. 1977. Recognition of new polymorphs in Paleozoic bryozoans: specialized mor- phoregulatory zooids. Geol. Soc. Am. Abstr. w. Prog. 9, 236—237. Anstey,R. L., Pachut, J. F., and Prezbindowski, D. R. 1976. Morphogenetic gradients in Paleozoic bryozoan colonies. Paleobiology 2, l3l-lA6. Anstey, R. L., Podell, M. E., and Pachut, J. F. 1978. Mor- phoregulatory monarchic zooics in Paleozoic bryozoan colonies. (In press). Blake, D. B. 1976. Functional morphology and taxonomy of branch dimorphism in the Paleozoic bryozoan genus Rhabdomeson. Lethaia 9, 169-178. Boardman, R. S. and Cheetham, A. H. 1973. Degrees of colony dominance in stenolaemate and gymnolaemate bryozoa. pp. 121-220. In: Boardman, R. S., Cheetham, A. H., and Oliver, W. A., eds. Animal Colonies. 603 pp. Dowden, Hutchinson and Ross; StroudSburg, PA. Boardman, R. S. and McKinney, F. K. 1976. Skeletal arch- itecture and preserved organs of four-sided zooids in covergent genera of Paleozoic Trepostomata (Bryozoa). J. Paleont. 50, 25-78. Boardman, R. S. and Utgaard, J. 1966. A revision of the Ordovician bryozoan genera Monticulopora, Peronopora, Heterotrypa, and Dekayia. J. Paleont. A0, 1082—1108. Borg, F. 1926. Studies on Recent cyclostomatous Bryozoa. Zool. Bidrag Uppsala 10, 181—507. Borg, F. 1933. A revision of the Recent Heteroporidae (Bryozoa). Zool. Bidrag Uppsala 1A, 253-39A. Borg, F. 1965. A comparative and phyletic study on fossil and Recent Bryozoa of the suborders Cyclostomata and Trepostomata. Arkiv Zool. 17 (1), 91p. 35 36 Bronstein, G. 1939. Sur les gradients physiologiques dans une colonie de Bryozoaires. g. R. hedb. Seance Acad. Sci., Paris. 209, 602-603. Corneliussen, E. F. and Perry, T. G. 1973. Monotrypa, Hallopora, Amplexopora, and Hennigopora (Ectoprocta) from the Brownsport Formation (Niagaran), western Tennessee. J. Paleont. A6, 151-220. Cumings, E. R. 190A. Development of some Paleozoic Bryozoa. Amer. J. Sci., Ser. A, 17, A9—78. Cumings, E. R. 1905. Development of Fenestella. Amer. J. Sci. Ser. A, 20, 169-177. Cumings, E. R. 1912. Development and systematic position of the monticuliporids. Geol. Soc. Amer., Bull. 26, 3A9-37A. Dzik, J. 1975. The origin and early phylogeny of the cheilostomatous Bryozoa. Acta Paleontol. Polon. 20, 395-A23- Gautier, 'T. G. 1970. Interpretive morphology and taxon- omy of bryozoan genus Tabulipora. Univ. Kansas Paleont. Contrib. A8, 21p. Gould, S. J. 1977. Ontogeny and Phylogeny. 501 pp. The Belknap Press of Harvard Univ. Press: Cambridge, Massachusetts, London, England. Harmer, S. F. 1896. On the development of Lichenopora verrucaria. Fabr. Microscop. Sci. Quart. Jour. 39, 7l—lAA. McKinney, F. K. 1977a. Functional interpretation of lyre- shaped Bryozoa. Paleobiology. 3, 90-97. McKinney, F. K. 1977b. Paraboloid colony bases in Paleo— zoic stenolaemate bryozoans. Lethaia. 10. 209-217. McKinney, F. K. 1978. Astogeny of the lyre-shaped Carbon- iferous fenestrate bryozoan Lyroporella. J. Paleont. 52, 83-90. Nye, O. B. 1976. Generic revision and skeletal morphology of some ceroporid cyclostomes. Bull. Amer. Paleont. 69. 1—222. Pachut J. F. 1977. Environmental stability and morpho- genetic relaxation in bryozoan colonies from the Eden Shale (Ordovician, Ohio Valley): A developmental ex- plaination of stability-diversity-variation hypotheses. Thesis, Michigan State Univ. 37 Ryland, J. S. 1970. Bryozoans. Hutchinson and Company, Ltd., London. 175p. Urbanek, A. 1960. An attempt at biological interpretation of evolutionary changes in graptolite colonies. Acta Palaeont. Polon. 5 (2), l27-23A. Urbanek, A. 1973. Organization and evolution of graptolite colonies. pp. AAl-51A. In: Boardman, R. S., Chee- tham, A. H., Oliver, W. A., eds. Animal Colonies. 603 pp. Dowden, Hutchinson and Ross; Stroudsburg, PA.