Enamel microstructure in lemuridae (mammalia primates) Assessment of variability.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 95221-241 (1994) Enamel Microstructure in Lemuridae (Mammalia, Primates): Assessment of Variability MARY c. MAAS Department of Biological Anthropology and Anatomy, Duke University Medical Center, Durham, North Carolina 27710 KEY WORDS bands Strepsirhini, Enamel prisms, Hunter-Schreger ABSTRACT This study describes the molar enamel microstructure of seven lemurid primates: Hapalemur griseus, Varecia variegata, Lemur catta, Lemur macaco, Lemur fulvus rufus, Lemur fulvus fulvus, and Lemur fulvus albifrons. Contrary to earlier accounts, which reported little or no prism decussation in lemurid enamel, both Lemur and Varecia molars contain a prominent inner layer of decussating prisms (Hunter-Schreger bands), in addition to an outer radial prism layer, and a thin, nonprismatic enamel surface layer. In contrast, Hapalemur enamel consists entirely of radial and, near the surface, nonprismatic enamel. In addition, for all species, prism packing patterns differ according to depth from the tooth surface, and for all species but Varecia (which also has the thinnest enamel of any lemurid), average prism area increases from the enamel-dentine junction to the surface; this may be a de-dopmental solution to the problem of accommodating a . Arc a with enamel deposited from a fixed number of cells.’ + larger outer sur. Finally, con g some previous reports, Pattern 1 prisms predomiaate’a only in the rl -*t ;.iperficial prismatic enamel. In the deeper enamel, prism cross-sections include both closed (Pattern 1) and arc-shaped (Pattern 2 or, most commonly, Pattern 3). This sequence of depth-related pattern change is repeated in all taxa. It should also be emphasized that all taxa can exhibit all three prism patterns in their mature enamel. The high degree of quantitative and qualitative variation in prism size, shape, and packing suggests that these features should be used cautiously in phylogenetic studies. Hapalemur is distinguished from the other lemurids by unique, medially constricted or rectangular prism cross-sections at an intermediate depth and the absence of prism decussation, but, without further assessment of character polarity, these differences do not clarify lemurid phylogenetic relations. Some characters of enamel microstructure may represent synapomorphies of Lemuridae, or of clades within Lemuridae, but homoplasy is likely to be common. Homoplasy of enamel characters may reflect functional constraints. 0 1994 Wiley-Liss, Inc. I The microstructure of primate dental enamel, like other aspects of tooth morphology, contains data about the phylogeny and functional morphology of primates. Researchers seeking phylogenetic information for the most part have focused on enamel prism cross-sectiona1 patterns (e.g*, Boyde and Martin, 1982, 1987; Carter, 1922; Dumont, 1993; Gantt, 1980, 1983, 1984a7b, 0 1994 WILEY-LISS, INC. 1986; Gantt et al., 1977; Martin et al., 1988; Shellis and Poole, 1977). Most have attempted to distinguish taxonomically diagnostic prism patterns that might elucidate Received January 19,1994; accepted May 30,1994. Address reprint requests to Mary C. Maas, Box 3170, Duke UniversityMedical Center, Durham, NC 27710. 222 M.C. MAAS phylogenetic relations of problematic fossils or clarify questions of origin or relationships of higher taxonomic groups. Problems with interpreting the results of these studies have arisen, however, because of variation in enamel microstructure, particularly within individuals. In investigations of primate enamel, some initial assessments that seemed to yield phylogenetically useful data (Gantt et al., 1977; Gantt, 1979, 1980) have been contradicted (Boyde and Martin, 1982, 1984b; Martin et al., 1988; Vrba and Grine, 1978). These latter studies demonstrated that variation in enamel microstructure is substantial and must be a major consideration in any attempts to use microstructural data in systematics. Most reports of strepsirhine enamel structure are based on single species (and in some cases single specimens) (Gantt, 1980; Martin et al., 1988; Shellis, 1984; Shellis and Poole, 1977). Not surprisingly, there are conflicting accounts of strepsirhine prism patterns and other microstructural characters. To date, there have been no comprehensive taxonomic surveys of enamel structure within this diverse group of primates, and there have been detailed studies of only two species, Daubentonia madagascarensis (Shellis and Poole, 1979; Koenigswald and Pfretzschner, 1987) and Otolemur crassicaudatus (Maas, 1993b). A fuller documentation of strepsirhine enamel structure has relevance not only for understanding phylogenetic relations within the group, but for studies of phylogenetic relations of early Tertiary prosimian-grade primates. However, before enamel structure can be used to address these issues, the nature of variation within individuals, within taxa, and between taxa must be clarified and conflicting reports of strepsirhine enamel patterns resolved. For example, we need to know whether different descriptions of strepsirhine enamel represent artifacts of preparation or intrinsic structural variation. Likewise, we need to consider how variation is expressed within teeth, among individuals, and among species and a t different, hierarchical levels of complexity. Finally, we need to determine which aspects of primate enamel structure are most appropriate for addressing different aspects of biology, including phylogeny and function. This study documents enamel microstructural variation in molar teeth of one group of strepsirhines, the Lemuridae. Lemurids are among the most taxonomically and ecologically diverse of the strepsirhines, and, although their taxonomy and systematics are not fully resolved’, they are the group on which most reports of strepsirhine microstructure are based. The aims of this study are to present an accurate description of lemurid molar enamel microstructure and to discuss the nature and implications of enamel variation within the group. ENAMEL MICROSTRUCTURE IN PRIMATES Two important properties of enamel are durability and structural complexity. Durability (which assures its preservation in the fossil record) is largely a function of its high mineral content (95% by weight). The structural complexity of enamel reflects its development. Enamel structure can be viewed hierarchically (Koenigswald and Clemens, 1992; Koenigswald et al., 1993). Thus, from smallest to largest units, enamel within a single tooth is comprised of crystallites, prisms, enamel types (arrangement and orientation of groups of prisms), and Schmelzmuster ‘Currently there is a lack of consensus in lemurid systematics. Independent studies have proposed generic distinction of Lemur catta and other Lemur species: Simons and Rumpler (1988) named the L. fuluus-L. macaeo-L. rubriuenter-L. mongoz-L. caronatus group =Eulemur“;Groves and Eaglen (1988) designated the same group “Petterus.”TattersalI and Koopman (1989) argued that both Petterus and Eulemur are junior synonyms of Prosimia Boddaert, 1785. Most recently, Tattersall and Schwartz (1991) have suggested that the genus Lemur includes all Lemuridae except Hapalemur, although currently most workers recognize the distinction between Varecia uariegata and Lemur spp. by separate generic designations (Petter et al., 1977). Lepilemur is currently considered distinct from other lemurs at the family level (e.g., Groves and Eaglen, 1988; Tattersall, 1982; Tattersall and Schwartz, 1991),but the position ofHapaZemur is less clear. Hapalemur is generally included in Lemuridae, but whereas some recognize a close relationship between Hapalemur and Lemur catta (Groves and Eaglen, 1988; Simons and Rumpler, 19881, others argue that Hapalemur is sister group to a clade consisting oflemur (includingthe Lemur fuluus group) and Varecia (Tattersall and Schwartz, 1991). Because of the lack of consensus, I follow a conservative approach here and use the generic designation Lemur for all lemurids but Varecia and Hapalemur. LEMUR ENAMEL MICROSTRUCTURE 223 (arrangement and organization of enamel venes between bodies of longitudinally types within a tooth). Crystallites, the aligned prisms (Boyde, 1964; Boyde and smallest structural units, may show regular Martin, 1984b; Gantt, 1986). Pattern 3 patterns of variation in orientation. These prisms, like Pattern 2 prisms, have cerviform the major regions of crystallite discon- cally incomplete boundaries, but, like Pattinuity known as prism boundaries, which tern 1prisms, are packed in horizontally offdemarcate enamel prisms and interpris- set rows. Boyde (1964; Boyde and Martin, matic enamel. Minor regions of crystallite 1984b) described three variants of Pattern discontinuity have been termed “enamel 3: 3A, 3B, and 3C. These differ primarily in seams” (Lester and Koenigswald, 1989) and the degree of cervical closure of prism occur within prism bodies (see arrow, Fig. boundaries and in the amount of interpris6 0 . They appear to be associated develop- matic enamel between adjacent prisms. mentally with a central ridge on the Tomes However, the variations on the three-patprocess (secretory process of the ameloblast) tern scheme have not been thoroughly docu(Lester and Koenigswald, 1989).Among pri- mented (Boyde, 1964; Boyde and Martin, mates, seams have been reported in human 1984b; Gantt, 1980, 1983, 1986; Shellis, deciduous teeth (Lester and Koenigswald, 1984). Moreover, intermediate forms of the 1989) and galago molars (Maas, 1993b). three basic patterns are common, and any The next hierarchical level of enamel or- tooth is likely to contain all three (Martin et ganization defined by Koenigswald and col- al., 1988). Enamel tubules are an additional feature leagues (Koenigswald and Clemens, 1992; Koenigswald et al., 1993), the prism, is com- of the prism level of organization. They are monly used in mammalian systematics. linear structures of controversial origin (see Most systematists follow Boyde’s (e.g., 1964, Boyde and Lester, 1984), which are gener1971) scheme of three major prism patterns, ally confined to prism bodies. In some but which is based on observations of developing not all cases, enamel tubules are continuous enamel but is typically applied to mature with dentine tubules, which are prominent (fully mineralized) enamel. The attributes histological features of dentine (Lester et al., used to classify prisms pertain to their mor- 1987). Tubules are common in marsupial phology viewed in cross-section (perpendicu- enamel, where they generally extend from lar to the prism long axis) and include com- the enamel-dentine junction (EDJ) to the pleteness of prism boundaries, packing tooth surface, but in primates they occur arrangement of adjacent prisms, and the less frequently and more typically end close relative proportions of prismatic and inter- to the EDJ (see arrow, Fig. 11). The next hierarchical level of enamel prismatic enamel. Pattern 1 prisms have complete, typically circular boundaries, are structure defined by Koenigswald and colpacked so that prisms in one row are offset leagues (Koenigswald and Clemens, 1992; from prisms in the adjacent rows, and usu- Koenigswald et al., 19931, the enamel type, ally are separated by relatively broad inter- concerns the arrangement and orientation prismatic regions. In Pattern 2 enamel, the of groups of prisms. Enamel types include prism boundaries are discontinuous to- radial enamel, where prism long axes are wards the cervix (root-crownjunction) of the more or less parallel to one another in their tooth, and prisms are arranged in longitudi- course from the EDJ to the tooth surface, nal (apicocervical) columns; thus, viewed in and decussating enamel, where prisms are cross-section, prisms in adjacent rows are arranged in bundles with different orientaaligned, and interprismatic enamel is apico- tions. A single plane of section thus intercervically continuous between adjacent cepts long axes of prisms in adjacent bunprisms (see arrows, Fig. 6B). In the variant dles at different angles. These adjacent of Pattern 2 enamel most commonly seen in decussation zones are designated parazones primates, interprismatic enamel is not re- and diazones, depending on whether prism stricted to inter-row sheets (as is typical of long axes are largely parallel or perpendicuungulates with Pattern 2 prisms) but inter- lar to the section plane. When sectioned 224 M.C. MAAS specimens are viewed under incident light, parazones and diazones appear as alternating light and dark bands, termed HunterSchreger bands (HSB). Decussating enamels vary according to several criteria: 1)the number of prisms contained in each decussation zone, 2) the amount of cuspal inclination of zones, from the EDJ to the tooth surface, 3) the angles between prisms in adjacent parazones and diazones, 4) whether boundaries between zones are abrupt or occur across a transitional zone several prisms wide, and 5) the complexity of zones, including patterns of zone bifurcation. Zone bifurcation occurs when some prisms within a zone abruptly change direction (Koenigswald and Clemens, 1992). A third enamel type is nonprismatic enamel, where C-axes of most crystallites are parallel to one another, without the regular changes in orientation that demarcate prism boundaries. The next hierarchical level, the Schmelzmuster, is the level of organization that describes the arrangement and distribution of enamel types within a tooth (Koenigswald and Clemens, 1992; Koenigswald et al., 1993). For example, a tooth may contain an inner zone of decussating enamel and an outer zone of radial enamel, or a tooth may consist entirely of a single enamel type. Likewise, the distribution of enamel types may be uniform or may differ among parts of the tooth. This study will examine lemurid microstructure at the levels of Schmelzmuster, enamel types, prisms, and crystallites. The focus is on variation within and between taxa a t each hierarchical level. MATERIALS AND METHODS Dental specimens used in this study consisted of upper and lower molars extracted from 16 lemurid cadavers. The specimens were obtained from the Duke University Primate Center (DUPC) and represent seven species and subspecies: Hapalemur griseus, Varecia variegata rufus, Lemur catta, Lemur macaco, Lemur fulvus rufus, Lemur fulvus fulvus, and Lemur fulvus albifrom (Table 1).The teeth were prepared, as outlined below, to document Schmelzmuster, enamel types, and prism morphology. TABLE 1. Lemurid saecimens included in this study Species Hapalemur grisew Hapalemur griseus Varecia uariegata rubra Varecia uariegata rubra Lemur catta Lemur catta Lemur catta Lemur macaco macaco Lemur fuluus rufus Lemur fuluus rufus Lemur fuluus rufus Lemur fuluus fuluus Lemur fuluus f u l u w Lemur fuluus futuw Lemur fuluus albifruns Lemur fuluus albifrons Specimen Age (years) DUPC’ 1326 DUPC 1317 DUPC 1127 DUPC 1363 DUPC 1183 DUPC 1242 DUPC 1020 DUPC 1268 DUPC 1192 DUPC 1021 DUPC 1239 DUPC 1202 DUPC 1266 D U E 1196 DUPC 1189 DUPC 1265 1 3 1 5 3 6 9 4 3 7 16 1 7 15 4 8 DUF’C, Duke University Primate Center. For study of Schmelzmuster and enamel types, two pairs of occluding upper and lower molars from each individual were embedded in epoxy resin (Epo-Tek; Epoxy Technology, Inc., Billerica, MA) and sectioned, following procedures described previously (Maas, 1993b). One pair was sectioned mesiodistally, in two planes: 1) through the buccal cusps and (Fig. 1A,B, a-a’) and 2) through the lingual cusps (Fig. 1A,B, b-b’). The second pair was sectioned orthogonal to the first, in two buccolingual planes: 1) through the mesial cusps (Fig. lA,B, c-c‘) and 2) through the distal cusps (Fig. 1A,B, d-d’). In all cases, planes of section were oriented perpendicular to the cervical plane (root-crownjunction) of the tooth (see Fig. 1C). The sectioned specimens were mounted on scanning electron microscope (SEM) stubs, polished with 800 and 1,200 grit silicon carbide abrasive paper and 0.05 p,m alumina powder, lightly etched in 5% HC1 for 10 seconds, and examined in a JEOL 35-C SEM. Sections were photographed a t low magnification ( 2 0 0 ~for ) preparation of photomontages of entire sectioned teeth, and a t higher magnification (1,000-2,000~) for two-dimensional measurement of angles of orientation of HSB, enamel prisms, and crystallites. Descriptions and measurements are based on buccolingual and mesiodistal sections of both upper and lower molars. Orientations of radial prisms and decussation zones were measured as the two-di- LEMUR ENAMEL MICROSTRUCTURE a Mesial d- a' - d' b - \ Distal Mesial \ Surface-Tangential Sectior Lingual C I Root I Fig. 1. Oblique occlusal views of (A) upper and (B) lower right first molars of Lemur fulvus rufus, illustrating orientation of four planes of section: a-a', b b ' , c x ' , and d-d'. The gray box in B illustrates location of surface-tangential sections, which is enlarged in the inset to illustrate the relationship of estimated ameloblast secretory area (dashed hexagons) to prism cross-sectional area (solid arcades). C illustrates lower molar section d-d' and the location of major structural features relative to the plane of section. 225 mensional angle between their long axes and the EDJ. Degree of decussation was estimated as the two-dimensional angle between the preferred orientation of prism long axes in the centers of adjacent parazones and diazones. The distribution of enamel types was observed under incident light, where HSB and radial enamel can be distinguished by their different reflective properties. The thickness of each enamel type was measured perpendicular to the EDJ and recorded as a proportion of total enamel thickness a t the point of measurement. To assess depth-related variation in prism cross-sectional morphology, surface-tangential facets were prepared on the buccal surfaces of first or second lower molars (Fig. 1B). The facets were ground with 800 and 1,200 grit silicon carbide abrasive paper until a dentine window was exposed at the center of the facet. This allowed documentation of prism cross-sections at three depths, relative to the outer tooth surface: superficial, intermediate, and deep. Superficial enamel was represented around the outer one-third of the facet circumference, intermediate enamel around the middle one-third of the facet circumference, and deep enamel in the inner one-third, surrounding the central dentine window (this method introduces some sectioning artifact, because prisms a t different positions on the facet are unlikely to be sectioned precisely perpendicular to their long axes). The ground facets were polished with 0.05 pm alumina powder. The specimens then were cleaned and etched lightly with acid (5% HC1 for 10 seconds). The etched specimens were sputter-coated and then examined and photographed in the SEM. Two micrographs (2,OOOx) were made at each enamel depth. Prism areas (PA) and estimated ameloblast secretory areas (ASA) were measured for ten prisms selected randomly from each micrograph (see Maas, 1993b). PA is the area contained within a prism boundary (for Pattern 2 and Pattern 3 prisms, the cervical prism boundary was estimated by a line connecting the two sides of the prism boundary). Estimated ASA is the area of a hexagon drawn to enclose one prism body and onehalf of the surrounding interprismatic M.C. MAAS 226 enamel (see Boyde, 1964; Fortelius, 1985). PA and ASA were traced directly from micrographs onto acetate overlays and measured using a computerized digitizing program (Sigmascan; Jandel Scientific, San Rafael, CA). An index of prism area to estimated ameloblast secretory area (PAIASA x 100) was calculated for each of the prism and estimated ameloblast area measurements. This index is a measure of the relative amounts of prismatic and interprismatic enamel. The higher the index, the smaller the proportion of interprismatic enamel. Statistical analyses of prism parameters included single classification analysis of variance (ANOVA) and nonparametric Kruskal-Wallis tests (SAS, 1987; Sokal and Rohlf, 1981)for several levels ofvariation: 1) among depths for each individual, 2) among depths for each species (individuals pooled), 3) among individuals of each species, for each enamel depth (except L. macaco, where n = 11, and 4) among species (individuals pooled), for each enamel depth. Significant pairs were identified by multiple comparisons procedures, using Tukey’s studentized range test (SAS, 1987). The more conservative nonparametric Kruskal-Wallis test was used in addition to the parametric ANOVA because some variables did not meet all the criteria for analysis of variance. RESULTS Schmeizmuster and enamel types Figures 2 and 3 illustrate the molar Schmelzmuster for each of the seven taxa. The figures show nonocclusal surfaces (buccal slope of hypoconid or lingual slope of entoconid), where the full thickness of enamel is preserved. Hapalemur molar enamel consists primarily of radial enamel, with a thin Fig. 2. Micrographsof buccolinguallysectioned molars (sectiond-d’; see Fig. lB, C). EDJ at bottom in all micrographs. A Hapalemur griseus, DUPC 1317, lingual slope entoconid (occlusal at right), LM,.B: Vureciu variegatu, DUPC 1363, buccal slope hypoconid (occlusal at leR), RM,. C: Lemur cattu, D W C 1183, lingual slope entoconid (occlusal at right), LM,. D: Lemur macuco, DWC 1268, buccal slope hypoconid (occlusal at leR), RM,. Note orientations of decussating prisms in parazones (p) and diazones (d) in E D . Scale bar, 100 km. LEMLTR ENAMEL MICROSTRUCTURE Fig. 3. Micrographs of buccolingually sectioned molars. EDJ at bottom. A. Lemur fuluus rufus, DUPC 1192, lingual slope entoconid (occlusal at right), RM,. B Lemur fuluus fuluus, DUPC 1202, lingual slope entoconid (occlusal at left), LM,. C: Lemurfuluus albifmns, D W C 1265, lingual slope entoconid (occlusal at left), LM,. Note relative proportions of radial enamel (r) and parazones (p) and diazones (d) of decussating enamel. Scale bar, 100 km. zone of nonprismatic enamel near the surface but no evidence of prism decussation (Fig. 2A). In contrast, the Lemur spp. and Varecia Schmelzmusters consist of three enamel types. From the surface to the EDJ, these are nonprismatic enamel (a very small component of the Schmelzmuster, most prominent towards the root but typically absent or negligible at more occlusal surfaces), 227 radial enamel, and decussating enamel (HSB) (Figs. 2B-D, 3A-C). This arrangement of enamel types is typical of both upper and lower molars, although the proportions of radial and decussating enamel vary among species (see Varecia, Lemur catta and L. macaco vs. L. fulvus subspp. below). In general, the distribution of enamel types is uniform throughout a single tooth (i.e., throughout mesial, distal, buccal, and lingual aspects, cusp tips, and basins), but in some cases a particular plane of section may obscure the Schmelzmuster. Therefore, care was taken to examine specimens in a variety of sectioning planes and using both light microscopy and SEM. In Lemur and Varecia, HSB typically have a slight cuspal inclination, forming angles of 70-80" with the EDJ. Towards the cervix, the enamel is thinner and HSB are more nearly perpendicular (90") to the EDJ. The arrangement of HSB in Lemur and Varecia is simple, with none showing bifurcation of zones. Numbers of prisms per zone range from 4-8, with variation within individual teeth. The change in prism orientation between parazones and diazones is gradual, occurring across a transitional band 2-3 prisms wide (Figs. 4, 5). The maximum difference in prism orientation between adjacent zones ranges from 30-50" but varies within single teeth. The transitions between zones are most distinct in Lemur catta and Varecia, which also have, on average, slightly narrower decussation zones (4-6 prisms wide) than Lemur macaco or Lemur fulvus. The Schmelzmuster of all subspecies of Lemur fulvus can be distinguished from those of Varecia, Lemur catta, and Lemur macaco by the proportions of radial enamel and decussating enamel. In Lemur fulvus, decussation zones are restricted to the inner 50-70% of the enamel (see Fig. 3) (except at cusp tips and wear facets, where the outer nonprismatic and radial enamel presumably are worn away). In Varecia, Lemur catta, and Lemur macaco, decussating enamel comprises a much larger proportion of the total enamel thickness (approximately 90%) (see Fig. 2B-D). Varecia has a molar Schmelzmuster most similar to that of Lemur catta. Varecia also has the largest Fig. 4. Micrographs of buccolingually sectioned molars, illustrating orientation of decussating prisms in parazones (p) and diazones (d). A Varecia variegata DUPC 1363, buccal slope hypoconid, RM,. B Lemur catta, DUPC 1183, lingual slope entoconid, LM,. C : Lemur Macaco, DUPC 1268, buccal slope hypoconid, RM,. Scale bar, 10 pm. Fig. 5. Micrographs of buccolingually sectioned molars, illustrating orientation of decussating prisms in parazones (p) and diazones (d). A Lemur fuluus rufus, DUPC 1192, lingual slope entoconid, RM,. B: Lemur fuluus fuluus, DUPC 1202, lingual slope entoconid, LM,. C: Lemur fuluus albifrons, DUPC 1265, lingual slope entoconid, LM,.Scale bar, 10 pm. LEMUR ENAMEL MICROSTRUCTURE 229 molars and thinnest enamel of all taxa included in this study; Varecia molar enamel thickness is only 50-75% that of Lemur or Hapalemur, measured perpendicular to the EDJ at homologous regions. Prisms and crystallites Figures 6-12 illustrate prism cross-sections for each species. Each species shows considerable variation in prism shape and packing arrangement, even within a single field of view, but the most distinct and consistent differences in prism pattern are between superficial prismatic enamel, where closed prisms (Pattern 1) predominate, and intermediate and deep enamel, where open prisms (Pattern 2 ' or Pattern 3) predominate. It should be emphasized, however, that all three prism patterns can occur in adjacent areas, at any enamel depth; thus, Pattern 1 prisms can be found in intermediate and deep enamel, though on average less frequently than open prisms. In some cases, the orientation of prisms relative to the surface-tangential plane appears to differ between depths. This can be seen most clearly in the different orientation of crystallite long axes relative to the surface. This may be an artifact of the preparation of surfacetangential sections or reflect intrinsic variation in prism orientation due to prism decussation. In most cases, open-boundary prisms, whether Pattern 2 or Pattern 3, are separated from one another by relatively broad regions of interprismatic enamel. Thus, the Pattern 3 prisms conform to Boyde's (1964) Pattern 3C, and the lemurid Pattern 2 prisms differ from Boyde's Pattern 2A, where interprismatic enamel is restricted to broad inter-row sheets. The angles between crystallites in interprismatic and prismatic enamel range from 30-60", regardless of prism pattern. In many instances, Pattern 2 and Pattern 3 arrangements were difficult to distin- Fig. 6 . Hapalemurgriseus, DUPC 1317, LM,.Surfacetangential sections illustrating prism packing arrangements for superficial(A),intermediate(B), and deep (C) enamel. Arrows in B indicate apico-cervically oriented regions of interprismatic enamel, typical of Pattern 2 enamel. Arrow in C indicates an enamel seam. Occlusal towards top. Scale bar, 10 pm. 230 M.C. MAAS Fig. 7. Varecia uariegata, DUPC 1127, LM,.Surfacetangential sections illustrating prism packing arrangements for superficial (A),intermediate (B), and deep (C) enamel. Occlusal towards top. Scale bar, 10 pm. Fig. 8. Lemur catta, DUPC 1183, LM,. Surface-tangential sections illustrating prism packing arrangements for superficial (A), intermediate (B), and deep (C) enamel. Occlusal towards top. Scale bar, 10 pm. LEMUR ENAMEL MICROSTRUCTURE Fig. 9. Lemur macaco, DUPC 1268, LM,. Surface-tangential sections illustrating prism packing arrangements for superficial (A),intermediate (B), and deep ( C ) enamel. Occlusal towards top. Scale bar, 10 pm. 23 1 Fig. 10. Lemur fulvus rufus, DUPC 1192, LM,. Surface-tangential sections illustrating prism packing arrangements for superficial (A), intermediate (B), and deep ( C )enamel. Scale bar, 10 pm. 232 M.C. MAAS Fig. 11. Lemur fuluus fuluus, DUPC 1202, RM,. Surface-tangential sections illustrating prism packing arrangements for superficial (A), intermediate (B), and deep (C) enamel. Arrow in C indicates cross-section of an enamel tubule. Note frequency of enamel tubules in deep enamel. Occlusal towards top. Scale bar, 10 pm. Fig. 12. Lemur fuluus albifrons, DUPC 1265, RM,. Surface-tangential sections illustrating prism packing arrangements for superficial (A), intermediate (B), and deep (C) enamel. Occlusal towards top. Scale bar, 10 pm. LEMUR ENAMEL MICROSTRUCTURE 233 guish. In part this is because of variation in and Kruskal-Wallis tests, P < 0.01). This prism orientation in regions of decussating was the case both for tests among depths for enamel. The Pattern 2 packing arrangement single individuals and for tests among was most clearly seen in the intermediate, depths for each species (with values for conradial enamel of Hapalemur (Fig. 6B) but specific individuals pooled). Similar patwas found in other taxa as well. terns of depth-related variation in prism Enamel seams were observed in deep size are also reported for other mammalian enamel of Hapalemur (see arrow, Fig. 6C) species (e.g., Fosse, 1968; Grine et al., 1986, but were not evident unequivocally in any 1987; Maas 1993b, in press), which points to other taxa. Enamel tubules, which in cross- the importance of comparing homologous section appear as small, dark holes within depths when making interspecific compariprism bodies, are common in the deep sons of prism size and shape. enamel ofLemur fulvus (see Figs. lOC, 11C, Results of comparisons among conspecific 12C) but were observed infrequently in sur- individuals at homologous depths have more face-tangential preparations of Lemur catta, serious implications for phylogenetic studVarecia, and Hapalemur. They were not ies. Significant differences between conspefound in Lemur macaco and were not seen in cific individuals were found in five of 54 poslongitudinal preparations. sible cases (comparisons for each variable at Lemurid prism cross-sections vary accord- each enamel depth for each of the six species ing to shape as well as packing pattern. In represented by more than a single individLemur spp. and Varecia molars, most ual): for Lemur catta deep enamel PA and prisms are round or ovoid in cross-section, ASA (ANOVA and Kruskal-Wallis test, although prisms with more angular apices P < 0.051, Lemur fulvus albifrons superfialso occur (see Figs. 7-12). There are differ- cial and intermediate enamel ASA (ANOVA ences in the completeness (degree of cervical and Kruskal-Wallis test, P < 0.05), and Leclosure) of prism boundaries, but there is no mur fulvus fulvus intermediate enamel ASA and Kruskal-Wallis test, regular pattern to this variation, which (ANOVA sometimes is seen between adjacent prisms P < 0.05). This amount of conspecific indiin a single field of view. In Hapalemur, some vidual variation, while not large, is greater superficial enamel shows a pattern of than what would, in most biological studies, “prisms within prisms” (Fig. 6A), but in gen- be expected from chance alone and suggests eral Hapalemur superficial and deep prism that, for lemurids, statistical tests of becross-sections conform to the predominately tween-species differences in metrical paovoid shape seen in other lemurids. In con- rameters may not be very informative. This trast, the intermediate enamel of Hapale- is, in fact, the case. While both ANOVA and mur is very distinctive (Fig. 6B). These Kruskal-Wallis tests demonstrate signifiprism cross-sections show greater apico-cer- cant among-species differences for all pavical distention than prisms of the other le- rameters a t all depths (individuals pooled murids. Most have a “waisted,” or medially for each species), multiple comparisons tests constricted, appearance; others have show that these significant differences perstraight sides and appear either rectangu- tain to different pairs of species, depending lar, with parallel sides and flat apical ends, on the enamel depth and the metrical paor triangular. rameter (Table 2). As might be expected from the within-speSome of the variation in prism paramecies variation in prism packing pattern and ters, both interspecific and intraspecific, unshape, there also is considerable within-spe- doubtedly reflects differences in the orientacies metrical variation in prism size. This tion of prism axes relative to the plane of the variation is expressed at two levels: among surface-tangential section. This is in part different enamel depths within individuals because of prism decussation and in part beand species and between conspecific individ- cause of sectioning artifact: a prism perpenuals a t the same enamel depth. For all spe- dicular to the EDJ, even if it pursues a cies but Vureciu, the three parameters (PA, straight course to the surface, will intercept ASA, and the PNASA index) differed signif- a plane tangential to the surface at slightly icantly among enamel depths (both ANOVA different angles at different enamel depths, 234 M.C. MAAS TABLE 2. Significant species pairs (P < 0.05) based on multiple comparisons (Tukey's studentized range test) for prism cross-sectional parameters PA (prism area), ASA (estimated ameloblast secretory area) and PAIASA' PA ASA Deep enamel LFA-HG; LFA-LC; LFA-LFR LFA-LM Intermediate enamel HG-LC; HG-LFA, HG-LFF HG-LFR; H G - W LC-LFA, LC-LM, LFA-LFR, LFA-LM, LFA-W, LFR-LM, LM-W Superficial enamel HG-LC; HG-LFA HG-LFF HG-LFR, HG-LM; W-LC; W-LFA, W-LFF; W-LFR; W-LM PAIASA HG-LFA HG-LFF; H G - W LFA-LC; LFA-LFR; LFA-LM HG-LC HG-LC; HG-LFA; HG-LFF; HG-LFR; H G - W , LM-LFA, LM-LFF; LM-LFR, LM-W HG-LC; HG-LFA; HG-LFF HG-LFR HG-W, LFA-LM HG-LC; HG-LFA, HG-LFF; HG-LFR HG-LM; LFA-LC; LFA-LM; L F A - W L F F - W HG-LC; HG-LFR; W-LC; W-LFA, W-LFR 'See text for discussionofprism parameters.Abbreviations: HG, Hapalemurgriseus;LC,Lemur catta; LFA,Lemurfulvus albifrons; LFF,Lemur fuluus fuluus; LFR, Lemur fuluus rufus; LM,Lemur macaco; W, Varecia uariegata. TABLE 3. Species values for PA (prism area), ASA (estimated ameloblast secretory area) and PAIASA x 100' Species Hapalemur grisew (n = 2) Depth S I D Varecia uariegata uariegata (n = 2) Lemur catta (n = 3) Lemur macaco macaw (n = 1) S I D S I D S I D Lemur fulvus rufus (n = 3) S Lemur fuluus fuluus (n = 3) I D S I D Lemur fuluus albifrons (n = 2) S I D PA (pm2) Mean 14.7 12.3 13.9 15.6 16.5 15.4 19.6 16.4 13.5 18.7 13.7 13.5 20.7 16.7 13.5 20.6 16.9 14.2 21.1 18.6 16.2 SD 4.0 2.4 2.0 2.9 2.8 2.8 2.8 2.6 2.7 2.2 2.2 2.2 3.1 2.1 2.0 2.9 2.5 1.9 3.2 3.1 3.0 ASA (prn') Mean SD 27.2 27.0 25.1 29.8 31.5 29.6 32.0 31.4 27.8 31.9 27.8 26.8 33.9 31.4 26.8 35.4 32.4 28.6 36.0 33.4 31.3 5.2 3.8 3.5 4.2 4.0 4.3 4.0 4.7 3.1 2.8 4.0 2.9 4.7 3.8 3.2 4.0 4.8 3.0 5.1 4.6 4.5 PNASA x 100 Mean SD 54 46 55 52 52 52 61 52 48 58 49 51 61 53 51 58 53 50 59 56 52 10 5 5 7 5 6 5 5 6 5 5 10 7 5 5 6 5 6 6 6 6 ~ See text for explanationof prism parameters.Means are based on measurement of ten prisms and surroundingestimated ameloblast secretory areas for each of n individuals.Abbreviations: S, Superficial;I, Intermediate;D, Deep. owing to the difference in radius of curvature of the EDJ and the tooth surface. While keeping in mind the intraspecific individual variation, the patterns of species averages of metrical parameters do shed some light on the nature of variation in lemurid microstructure. Species values for PA, ASA, and the PNASA index are reported in Table 3. As expected, given the intraspecific variation, standard deviations are large for all parameters, and there is considerable overlap in values, not only among depths but among species. For most taxa, PA and ASA show an average size in- crease from deep to superficial enamel (Fig. 13).Varecza,which has the thinnest enamel, is an exception. It shows little difference in average PA and ASA among enamel depths. Hapalemur, which has enamel comparable in thickness to Lemur spp., shows a deep to superficial size increase in average ASA but not PA. Among enamel depths, the average PAIASA index is greatest for superficial enamel in Lemur spp. but not Varecia and Hapalemur. Among the seven taxa, the range of prism areas is 8-29 pm2, corresponding to a range of prism diameters of 3.5-6 pm. Estimated ameloblast secretory LEMUR ENAMEL MICROSTRUCTURE 235 Fig. 13. Average species values for prism parameters PA (prism area), ASA (estimated ameloblast secretory area), and the index PNASA x 100. The relative amount of interprismatic enamel varies inversely with the index value. S = Superficial enamel; I = Intermediate enamel; D = Deep enamel. See text for further explanation. area ranges from 19-50 pm2, and the proportion of prismatic enamel (PNASA) ranges from 32-74%. mammals with prismatic enamel, absence of HSB due to evolutionary reversal has been documented in at least one group, Cetacea (Ishiyama, 1987; Maas and Thewissen, in DISCUSSION press). Alternatively, prism decussation The Schmelzmuster of lemurid molars is may be derived for lemurids. It has been more complex than previously reported. suggested that prism decussation, which With the exception of Hapalemur, decussat- serves the biomechanical function of a ing, radial, and nonprismatic enamel are all crack-stopping mechanism, evolved indepresent, contradicting previous reports of pendently in many lineages of mammals, in Lemur enamel that described absence of de- response to increased chewing stresses assocussation (Boyde and Martin, 1982, 1984b) ciated with increase in body size (Koenigor little or no decussation (Shellis, 1984). swald et al., 1987). The absence of decussaThe differences between Hapalemur and tion only in Hapalemur griseus, the smallest other lemurids also reflects greater be- of the lemurids, provides some support for tween-species variation in Schmelzmuster the idea of independent evolution of decusthan previously appreciated. sation within the larger-bodied lemurids. The phylogenetic implications of within- Decussation may have evolved indepenfamily variation in Schmelzmuster and dently in other strepsirhine lineages as well, enamel types are uncertain, since the char- as suggested by the presence of decussation acter polarities of enamel types currently in the relatively large-bodied Propithecus are unknown for Lemuridae, for Strep- (Shellis, 1984) and the absence of decussasirhini, and for Primates. Decussation may tion in the smaller Otolemur (Maas 1993b). be primitive for lemurids. If so, the absence If HSB were absent in the last common anof HSB in Hapalemur would represent an cestor of lemurids and other strepsirhines, evolutionary reversal. Although radial decussation could represent either a synapoenamel is almost certainly primitive for morphy of a Varecia-Lemur clade, according 236 M.C. MAAS to the phylogenetic hypothesis of Tattersall and Schwartz (1991), or an instance of parallel evolution if, as others have suggested (Groves and Eaglen, 1988; Simons and Rumpler, 1988), Hapalemur and Lemur cattu are sister taxa. It would not be surprising to find that some enamel characters, like many other characters of lemurid morphology and behavior, exhibit homoplasy, particularly since this is the case in other groups of mammals. The question then arises as to the extent that the distribution of enamel characters reflects functional constraints, and what the evolutionary history of this structure-function complex is within the Lemuridae, and within Strepsirhini, and Primates. As with descriptions of lemurid Schmelzmuster, previously published reports of lemur prism patterns offer conflicting interpretations. Boyde and Martin (1982, 1984a,b, 1987) reported predominance of Pattern 1 prisms, Shellis (1984; Shellis and Poole, 1977) described Lemur sp. enamel as a variant of Pattern 3, and Gantt (1980) described Lemur sp. enamel prisms as opensided and transitional between Pattern 1 and Pattern 3 prisms. In the specimens reported here, Pattern 1 prisms clearly predominate only in the superficial enamel of lemurid molars, and prisms with open boundaries are common in deeper enamel. In the case ofLemur and Varecia, these open prisms most commonly show a Pattern 3 arrangement, whereas Pattern 2 prisms are more common in Hapalemur. In all cases, prisms are widely separated by interprismatic enamel, as indicated by the relatively low PAIASA indices. This high proportion of interprismatic enamel may be primitive for mammals with prismatic enamel (Koenigswald and Clemens, 1992; Kozawa, 1984; Sahni, 1979, 1987). None of the lemurids showed any of what are probably the more derived prism pattern variants, where interprismatic regions are more restricted (i.e., the keyhole-shaped prisms of Pattern 3A, where interprismatic enamel is restricted to prism tail regions, or Pattern 2A, where interprismatic enamel is restricted to interrow sheets and the angle between prismatic and interprismatic crystallites is close to 900). The observation that all three prism patterns occur in all lemurids studied here, sometimes within a single field of view, is not surprising. This heterogeneity of prism packing arrangements has been reported for many other primates, including humans, macaques, and a variety of New World monkeys (Boyde, 1964; Boyde and Martin, 1987). It is probably typical of primates, as well as many other mammals. In light of the common occurrence of Pattern 2 and Pattern 3 prisms in all but the superficial level of lemurid molar enamel, the claim that strepsirhines are characterized by Pattern 1 enamel (Martin et al., 1988) should be abandoned. This, along with the fact that Pattern 2 and 3 prisms also are present in molar enamel of Otolemur (Maas, 1993b), Nycticebus, Perodicticis (Shellis, 1984), Propithecus (Shellis, 1984; Boyde and Martin, 1987), and Daubentonia (Shellis and Poole, 1979; Martin et al., 1988), indicates that there is no support for the suggestion that the distribution of prism patterns might provide evidence for either a or prosimian/ strepsirhinekaplorhine anthropoid dichotomy (Martin et al., 1988). As is the case with Schmelzmuster, use of prism patterns in phylogenetic studies requires determination of character polarities. Some have assumed that closed (Pattern 1) prisms are primitive for mammals, based on knowledge of the distribution of Pattern 1 prisms in some living groups (e.g., Boyde and Martin, 1984a). However, Pattern 1 prisms have been shown to be derived for Multituberculata (Krause and Carlson, 19871, and a report of open prisms in a dryolestid eupantothere, a primitive Cretaceous mammal (Lester and Koenigswald, 1989), suggests that open prisms may in fact be primitive for therian mammals as well. As pointed out by Koenigswald and Clemens (1992), determination of character polarity for prism patterns is complicated by the fact that a given pattern can include apparently primitive and very derived variants. Thus, Pattern 3 enamel includes the open prisms of the dryolestid eupantothere as well as the almost certainly derived Pattern 3A keyhole pattern of human enamel. Furthermore, the character transformation sequence among LEMUR ENAMEL MICROSTRUCTURE variants of a given prism pattern (e.g., Pattern 3C to 3B to 3A) or among prism patterns (e.g., Pattern 1 to Pattern 2 or Pattern 3) has yet to be demonstrated. The ubiquitous occurrence of open prisms in the lemurids studied here does provide some support for the notion that open prisms are primitive for Lemuridae, though this has yet to be tested by out-group comparisons. In any event, open prisms are clearly more widely distributed among strepsirhine taxa than previously thought and may well be primitive for strepsirhines, if not primates. In this context, the enamel microstructure of various early Tertiary primates is of considerable interest. Not only might it elucidate character states of primitive primate enamel, but such data also are relevant to the issue of the relationships of various Eocene primate groups to living haplorhines and strepsirhines. Gantt (1980) reported that a number of early Eocene primates (Necrolemur, Adapis, Smilodectes, and Notharctus) were characterized by Pattern 1 prisms. However, it is possible that Gantt’s study sampled only the superficial enamel, a region where, as has been shown here, Pattern 1 prisms are common, even for taxa where open prisms predominate at deeper levels. Gantt’s preliminary survey therefore cannot be considered conclusive. Prism shape also has received some attention in phylogenetic studies, although generally less than prism packing patterns. For example, prism cross-sections may appear apico-cervically distended or compressed, and their shapes may vary in angularity and symmetry. Qualitative differences in prism shape (rounded-ovoid,rectangular, triangular) were used to distinguish between prisms of colobines and cercopithecines and among various hominoids (Dostal, 1989; Dostal and Zapfe, 1986; Dostal et al., 1985). Martin et al. (1988:259) noted a “so far unique inverted V shape to the prism outline” of the callitrichid primate Leontopithecus. In the case of lemurids, the distinctive medially constricted prism cross-sections of Hapalemur intermediate enamel appear to be yet another instance of a unique prism shape. If this does represent an autapomor- 237 phy of Hapalemur, however, it is of little significance for lemurid or strepsirhine systematics. In terms of phylogenetic significance, quantitative assessments of differences in lemurid prism shape are even more problematic than qualitative assessments. Statistical analyses of prism shape variables have proved difficult for some taxa, largely because of the heterogeneity of variances of shape parameters (Grine et al., 1986,1987). Moreover, the utility of quantitative analyses of prism shape in systematics has been questioned, again largely because of the heterogeneity in prism shape in some species (Koenigswald and Clemens, 1992; Maas, in press). Quantitative prism shape analyses are particularly problematic for decussating enamel, where differences in prism orientation add an additional component of variation which can confound two-dimensional shape analysis (Maas, in press; Maas and Thewissen, in press). Likewise, differences in sectioning planes among specimens may influence prism shape comparisons (Maas, 1993b), and prism shape also can be affected by etching regime (e.g., Martin et al., 1988). For whatever reasons, the magnitude and nature of within-species variation in prism parameters for the lemurids studied here precluded meaningful statistical analyses of species differences. Nevertheless, some interesting trends were apparent when only species averages were considered. The increase in average area of estimated ameloblast secretion from the EDJ to the tooth surface, seen in all lemurids except Varecia, most likely reflects increase in tooth area from the EDJ to the surface (e.g., Fosse, 1964, 1968): estimated ameloblast secretion area increases from the EDJ to the outer surface of the tooth because these differentsized surfaces must be covered by enamel deposited by the same number of ameloblasts. Varecia, which has the thinnest enamel of any of the taxa studied and therefore presumably the least depth-related difference in enamel area, does not show this size-depth relationship. In sum, metrical variation in size or shape of enamel prisms and ameloblasts appears to have little systematic significance a t the 238 M.C. MAAS generic or specific level, at least for Lemuridae. Although the sample sizes for all taxa were small (1-3 individuals), it is unlikely that larger sample sizes would add any discriminating power to these variables, given the high individual variability in prism parameters and the extensive overlap in values among taxa. In some other mammalian groups, most notably Multituberculata, prism size variation has been shown to be useful systematically a t higher taxonomic levels (e.g., Krause and Carlson, 19871, but it has yet to be shown that this is the case for primates. The systematic significance of prism features such as enamel seams and enamel tubules in lemurids is unclear. Of the taxa studied here, seams were observed unequivocally only in the deep enamel of Hapalemur. Enamel seams have been observed in Otolemur crassicaudatus as well (Maas, 199313). Seams are common in chiropteran enamel (Lester and Hand, 1987) and may be a fairly common component of primitive prismatic and nonprismatic enamels (Lester and Koenigswald, 1989; Koenigswald and Clemens, 1992). As noted above, identification of tubules in cross-sectional preparations is sometimes problematic, and thus the apparent absence of tubules in some specimens examined here is not definitive. Shellis and Poole (1977), for example, found that in Lemur sp., tubules could be observed in thin section preparations but not in the SEM. In this study, tubules were fairly common in cross-sections of deep enamel of Lemur fulvus and Lemur catta and very close to the EDJ in Hapalemur. Shellis and Poole (1977) described Propithecus enamel and noted that tubules frequently appeared to extend to the tooth surface. This does not appear to be the case for lemurids. Although we cannot assess polarity of prism patterns or enamel types from the data reported here, microstructure is consistent with the interpretation, based on some other morphological criteria (Tattersall and Schwartz, 19911, that Lemur spp. and Varecia are distinct from Hapalemur. Moreover, there are no characters of enamel microstructure shared between Lemur catta and Hapalemur to the exclusion of the other lemurids. Koenigswald and colleagues (Koe- nigswald and Clemens, 1992; Koenigswald et al., 1993) have suggested that there is an inverse relationship between the hierarchical levels of structural complexity and taxonomic rank. They propose that the most complex structural levels (enamel types, Schmelzmuster) are useful in species and generic distinctions, whereas the least complex levels (crystallites, prisms) are most diagnostic at the taxonomic level of the family or order. The theoretical basis of this is unclear, but this analysis of lemurid enamel, where Hapalemur is most clearly distinguished from Lemur spp. and Varecia on the basis of Schmelzmuster, appears to be another case of this phenomenon. It is particularly important to understand that although enamel microstructural characters may in some cases be useful in testing phylogenetic hypotheses, this is not always the case. Care must be taken to define character states so as to facilitate distinction of homoplasies from shared derived characters. The traditional definitions of enamel characters are a first step but not, in themselves, adequate for this task. CONCLUSIONS The Hapalemur Schmelzmuster, which consists almost entirely of radial enamel, with a thin surface layer of nonprismatic enamel, is different from that of the other lemurids studied. The Schmelzmuster of Lemur spp. and Varecia contains an additional, prominent inner layer of decussating enamel (HSB). In Lemur fulvus subspp., HSB comprise a smaller proportion of the total enamel thickness than in Lemur catta, Lemur macaco, and Varecia, but this difference in Schmelzmuster is much less pronounced than that which distinguishes &mur spp. and Varecia from Hapalemur. All lemurids are characterized by an open (Pattern 2 or Pattern 3) prism pattern. Previous reports that Lemur enamel consists of Pattern 1 prisms are based on incomplete sampling of lemur enamel. Pattern 1prisms are predominant a t the surface of enamel of lemurids, but Pattern 2, or, most commonly, Pattern 3 prisms are found a t deeper levels. The Pattern 2 arrangement is most pronounced in Hapalemur molars, where, at in- LEMUR ENAMEL MICROSTRUCTURE termediate depths, prisms also have a unique, medially constricted cross-sectional shape. Although this study demonstrates similarity in enamel microstructure among Lemur spp. and Varecia and the distinctiveness of Hapalemur, the phylogenetic significance of the distribution of enamel microstructural features depends on interpretation of character polarity. Thus, whether radial enamel or HSB is primitive for Lemuridae or for Strepsirhini would affect phylogenetic interpretations based on Schmelzmuster. Likewise, open prisms may be an autapomorphy of Lemuridae or primitive for the group. To begin to resolve questions of character polarity and character transformation in lemurid enamel microstructure, one must first identify appropriate outgroups, which, for Lemuridae, is not an easy task (Eaglen, 1983). Previously suggested candidates include Indriidae (e.g., Eaglen, 1983; Groves, 1974), Cheirogaleidae (e.g., Charles-Dominique and Martin, 1970; Eaglen, 1983; Yoder, 1994), and the fossil Adapidae (e.g., Beard et al., 1988; Eaglen, 1983; Gregory, 1920; Szalay and Delson, 1979). Currently, little or nothing is known of enamel microstructure in any of these groups, but a survey of enamel microstructure of these and other strepsirhine taxa could resolve the question of polarity of enamel characters. Whether this information would in turn help resolve phylogenetic questions is less certain, however, given the likelihood of homoplasy in enamel characters. There is a growing body of evidence that the distribution of some enamel microstructure characters reflects functional, as well as phylogenetic constraints (e.g., Koenigswald et al., 1987; Maas, 1992, 1993a,b; Stern et al., 1989).Although enamel microstructure may not clarify phylogenetic questions for this group of mammals, an understanding of the distribution of enamel characters, within a phylogenetic context, could provide important insight into the evolution of this structure-function complex. ACKNOWLEDGMENTS I thank J. Muennig for photographic work, D. Overdorff and A. Yoder for discus- 239 sion of strepsirhine systematics, and R.F. Kay, P. Ungar, and two anonymous reviewers for comments on an earlier draft of this paper. This research was supported by BNS 9020788 from the National Science Foundation. LITERATURE CITED Beard KC, Dagosto M, Gebo DL, and Godinot M (1988) Interrelationships among primate higher taxa. Nature 331:712-714. Boyde A (1964) The Structure and Development of Mammalian Enamel. Ph.D. thesis, University College, London. Boyde A (1971) Comparative histology of mammalian teeth. In AA Dahlberg (ed.): Dental Morphology and Evolution. Chicago: University of Chicago Press, pp. 81-94. Boyde A, and Lester K (1984) Further SEM studies of marsupial enamel. In RW Fearnhead and S Suga (eds.): Tooth Enamel N.Amsterdam: Elsevier Science Publishers, pp. 442-446. Boyde A, and Martin L (1982) Enamel microstructure determination in hominid and cercopithecoid primates. Anat. Embryol. 165:193-212. Boyde A, and Martin L (1984a) A non-destructive survey of prism packing patterns in primate enamels. In RW Fearnhead and S Suga (eds.): Tooth Enamel. N.Amsterdam: Elsevier, pp. 417421. Boyde A, and Martin L (1984b) The microstructure of primate dental enamel. In DJ Chivers and A Bilsborough (eds.): Food Acquisition and Processing in Primates. New York: Plenum Press, pp. 341-367. Boyde A, and Martin L (1987) Tandem scanning reflected light microscopy of primate enamel. Scanning Microsc. 1:1935-1948. Carter JT (1922) On the structure of the enamel in the Primates and some other mammals. Proc. Zool. SOC. Land. 1922:599-608. Charles-Dominique P, and Martin RD (1970) Evolution of lorises and lemurs. Nature 227:257-260. Dostal A (1989) Dental enamel prism patterns of Old World monkeys and man. In H Splechtna and H Hilgers (eds.): Trends in Vertebrate Morphology. Fortschritte der Zoologies, Volume 35. Stuttgart: Gustav Fischer Verlag, pp. 251-255. Dostal A,, and Zapfe H (1986) Zahnschmelzprismen von Mesopithecus pentilicus Wagner, 1839, im Vergleich mit rezenten Cercopitheciden (Primates: Cercopithecidae). Folia Primatol. 46t235-251. Dostal A, Ruzicka F, and Zapfe H (1985) Comparative examinations of cercopithecid dental enamel prism pattern. Anzeiger der Osterreichischen Akademie 122t155-158. Dumont ER (1993) Functional and Phyletic Features of Mammalian Dental Enamel: A Key to Primate Higher-Level Relationships. Unpublished Ph.D. thesis, State University of New York at Stony Brook. Eaglen RH (1983) Parallelism, parsimony, and the phylogeny of the Lemuridae. Int. J. Primatol. 4:249-273. Fortelius M (1985) Ungulate cheek teeth: Developmen- 240 M.C. MAAS tal, functional, and evolutionary interrelations. Acta 2001.Fenn. 18O:l-76. Fosse G (1964) The number of prism bases on the inner and outer surface of the enamel mantle of human teeth. J. Dent. Res. 4357-63. Fosse G (1968)A quantitative analysis of the numerical density anbd distributional pattern of prisms and ameloblasts in dental enamel and tooth germs. V. Prism density and pattern on the outer and inner surface of the enamel mantle of canines. Acta Odontol. Scand. 26.501-544. Gantt DG (1979)A method of interpreting enamel prism patterns. Scanning Electr. Microsc. 19791975-981. Gantt DG (1980) Implications of enamel prism patterns for the origin of the New World monkeys. In RL Ciochon and AB Chiarelli (eds.): Evolutionary Biology of the New World Monkeys and Continental Drift. New York Plenum Press, pp. 201-217. Gantt DG (1983) The enamel of Neogene hominids: Structural and phyletic implications. In RL Ciochon and RS Corruccini (eds.): New Interpretations of Ape and Human Ancestry. New York: Plenum Press, pp. 249-298. Gantt DG (1986) Enamel thickness and ultrastructure in hominids: With reference to form, function, and phylogeny. In D Swindler and E Irwin (eds.): Comparative Primate Biology, Volume 1: Systematics, Evolution, and Anatomy. New York: Alan R. Liss, Inc., pp. 453475. Gantt DG, Pilbeam DR, and Steward GP (1977) Hominoid enamel prism patterns. Science 198:1155-1157. Gregory WK (1920) On the structure and relations of Notharctus, a n American Eocene primate. Mem. Am. Mus. Nat. Hist. 3:49-243. Grine FE, Fosse G, Krause DW, and Jungers WL (1986) Analysis of enamel ultrastructure in archaeology: the identification of Ouis aries and Capra hircus dental remains. J. Arch. Sci. 13:57!%595. Grine FE, Krause DW, Fosse G, and Jungers WL (1987) Analysis of individual, intraspecific and interspecific variability in quantitative parameters of caprine tooth enamel structure. Acta Odontol. Scand. 45:l23. Groves CP (1974) Taxonomy and phylogeny of prosimians. In RD Martin, GA Doyle, and AC Walker (eds.): Prosimian Biology. London: Duckworth, pp. 435448. Groves CP, and Eaglen RH (1988) Systematics of the Lemuridae (Primates, Strepsirhini). J . Hum. Evol. 17: 513-538. Ishiyama M (1987) Enamel structure in odontocete whales. Scanning Microsc. 1:1071-1079. Koenigswald Wv, and Clemens WA (1992) Levels of complexity in the microstructure of mammalian enamel and their application in studies of systematics. Scanning Microsc. 6:195-218. Koenigswald Wv, and Pfretzschner HU (1987) HunterSchreger bands in the enamel of mammalian teeth. Arrangement, orientation of the prisms. Zoomorphology 106t329-338. Koenigswald Wv, Rensberger JM, and Pfretzschner HU (1987) Changes in the tooth enamel of early Paleocene mammals allowing increased diet diversity. Nature 328:15&152. Koenigswald Wv, Martin T, and Pfretzschner HU (1993) Phylogenetic interpretation of enamel structures in mammalian teeth: Possibilities and problems. In FS Szalay, MJ Novaceck, and MC McKenna (eds.): Mammal Phylogeny: Placentals. New York: Springer-Verlag, pp. 303-314. Kozawa Y (1984) The development and evolution of mammalian enamel structure. In RW Fearnhead and S Suga (eds.): Tooth Enamel IV.Amsterdam: Elsevier Science Publishers, pp. 437441. Krause DW, and Carlson SJ (1987) Prismatic enamel in multituberculate mammals: Tests of homology and polarity. J . Mamm. 68:755-765. Lester KS, and Hand SJ (1987) Chiropteran enamel structure. Scanning Microsc. 1:421436. Lester KS, and Koenigswald Wv (1989) Crystallite orientation discontinuities and the evolution of mammalian enamel-or, when is a prism? Scanning Microsc. 3:645-663. Lester KS, Hand SJ, and Vincent F (1987) Adult phyllostomid (bat) enamel by scanning electron microscopy-with a note on dermopteran enamel. Scanning Microsc. 2:371-383. Maas MC (1992) Enamel structure and function in brachydont molars: The example of lemurid primates. J . Vert. Paleontol. 12 (Supplement to No. 3):41A(abstract). Maas MC (1993a) Constraints on enamel microstructure in lemurs. Am. J . Phys. Anthropol. 16(Suppl.): 136 (abstract). Maas MC (1993b) Enamel microstructure and molar wear in the greater galago, Otolemur crassicaudatus (Mammalia, Primates). Am. J . Phys. Anthropol. 92: 217-233. Maas MC (in press) Enamel microstructure in Notoungulata (Mammalia). In RF Kay, RH Madden, JJ Flynn, and R Cifelli (eds.): Paleobiology of an Extinct Neotropical Fauna: The Miocene Fauna from Colombia. Washington DC: Smithsonian Institution Press. Maas MC, and Thewissen JGM (in press) The enamel microstructure of Pakicetus (Mammalia: Archaeoceti). J. Paleontol. Martin LB, Boyde A, and Grine FE (1988) Enamel structure in primates: A review of scanning electron microscope studies. Scanning Microsc. 2:1503-1526. Petter J-J, Albignac R, and Rumpler Y (1977) Faune de Madagascar: Lemuriens. Paris: Centre Nationale de la Recherche Scientifique. Sahni A (1979) Enamel ultrastructure of certain North American Cretaceous mammals. Palaeontographica Abt. A 166:37-49. Sahni A (1987) Evolutionary aspects of reptilian and mammalian enamel structure. Scanning Microsc. 1 :1903-1912. SAS Institute Inc. (1987) SASISTAT Guide for Personal Computers, Version 6 Edition. Cary, North Carolina: SAS Institute Inc. Shellis RP (1984) Inter-relationships between growth and structure of enamel. In RW Fearnhead and S Suga (eds.): Tooth Enamel IV.Amsterdam: Elsevier Science Publishers, pp. 467471. Shellis RP, and Poole DFG (1977) The calcified dental tissues of primates. In CLB Lavelle, RP Shellis, and DFG Poole (eds.): Evolutionary Changes to the Pri- LEMUR ENAMEL MICROSTRUCTURE mate Skull and Dentition. Springfield, IL: Charles C Thomas Publisher, pp. 197-279. Shellis RP, and Poole DFG (1979) The arrangement of prisms in the enamel of the anterior teeth of the ayeaye. Scann. Electr. Microscop. II:497-506. Simons EL, and Rumpler Y (1988) Eulemur: New generic name for species of Lemur other than Lemur catta. C.R. Acad. Sci. I11 307r547-551. Sokal RR and Rohlf FJ (1981) Biometry: The Principles and Practice of Statistics in Biological Research, 2nd ed. New York: W.H. Freeman and Co. Stern D, Crompton AW, and Skobe Z (1989) Enamel ultrastructure and masticatory function in molars of the American opossum Dideelphis uirginiana. 2001. J. Linn. SOC.93:311-334. Szalay FS, and Delson E (1979) Evolutionary History of the Primates. New York: Academic Press. 241 Tattersall I (1982) The Primates of Madagascar. New York: Columbia University Press. Tattersall I, and Koopman K (1989) A further note on nomenclature in Lemuridae. J. Hum. Evol. 18:499500. .. . Tattersall I, and Schwartz J H (1991) Phylogeny and nomenclature in the “Lemur-group” of Malagasy strepsirhine primates. Anthropol. Pap. Am. Mus. Nat. Hist. 69:l-18. Vrba ES, and Grine FE (1978)Australopithecineenamel prism patterns. Science 202:890-892. Yoder AD (1994)Relative position of the Cheirogaleidae in strepsirhine phylogeny: A comparison of morphological and molecular methods and results. Am. J . Phys. Anthropol. 94:25-46.