I Reprinted from JOURNAL OF PALEONTOLOGY Vol. 36, No. 4, July, 1962 EVOLUTIONARY TRENDS AMONG MOLLUSK FECAL PELLETS LOUIS S. KORXICICER The A. & M. College of Texas, College Station In the study of invertebrate fossils such as animal. Muscle scars and the pallial sinus of gastropods and pclecypods direct knowledge the clam shell come in this category. Occa- of soft parts must come from inferences sionally the paleontologist's knowledge of drawn from features of the shell that reflect soft parts is aided by an unusual preserva- some quality or position of a soft part of the tion such as the fossilized gastropod intestine 830 7^jz,/tOArroz.O(77C/iz. woygj described by Casey (1960) from the Lower SAND Cretaceous, but such finds are indeed rare. 'GRAINS Another way to learn about soft parts of fossils is by studying fecal pellets. These have been described from many geologic forma- tions, but have not been viewed as objects useful in reconstructing the internal anat- (7 omy of fossil organisms. Many biologists and geologists have noted that invertebrates produce feces hav- ing a characteristic shape and consistency that, is quite uniform for members of a spe- 0 •vi-Uuiji.^v.. .%& cies (for review see .Moore, 1939). Manning & Kumpf (1959), after comparing feces TI'.XT-FIG. /—[Examples of fecal pellets. from many invertebrates, conclude that A. Oval pellet from Cerilluuiii muscariim. Pel- "Most of the animals in a given group, usu- let is light tan with brown spots. Length 1 mm. Pellet was obtained from a living ally at the generic level, shed similar pellets." specimen collected in the vicinity of Hi- Abbott (1954a) used the shape and position nt! in, Bahamas. of feces of the gastropod genus Echini)!us, B. Sculptured rod from Ciltariuni pica. Pellet along with other criteria, as a basis for re- is light brown. The pellet contains two long- itudinal surface grooves; one groove con- moving Echininiis from the family Modu- tains a spiral tube. Faint ridges run the lidae and placing it in the family Litto- length of the pellet. Coarser sediment riniclac. Abbott states that "The feces of grains are concentrated near center of pel- many, if not all, Littorinidae are relatively let beneath spiral tube. Pellet length is 4 mm. Pellet was obtained from a living short (2 or 3 times as long as wide) and arc specimen collected in the vicinity of Bi- lined up in the rectum, one directly behind nd ni, Bahamas. the other," but Abbott did not feel that this C. Unsculptured rod from Nerita fulgurans character is necessarily of phylogenetic im- (from Manning & Kumpf, 1959, fig. 3a). Pellet length varies from 1-3 mm., width portance. The present, report is the result of 0.71 mm. an investigation based upon the hypothesis I). Ribbonlike pellet from Brachidontes exus- that, shape of fecal pellets may indeed be of tus (from Manning & Kumpf, 1959, fig. phylogenetic importance and, therefore, be lc). Pellet length if 0.95 mm., width 0.25 mm. useful in deriving the relationship between invertebrate stocks. genera have not been treated fully in that GASTROPODA work, it has been more convenient to use The feces of Gastropoda can be divided here the classification in Abbott (1954). The according to shape into various types. gastropod subclass Prosobranchia is divided Manning & Kumpf (1959) classified gastro- into three orders: Archacogastropoda, Mcso- pod pellets as: ovoid or ellipsoid, ellipsoid to gastropodaand Xeogaslropoda (TheTreatise rodlike; rodlike, rod with external longi- on Invertebrate Paleontology includes the last tudinal sculpture, rod with spiral sculpture, two in the order Caenogastropoda). The rod with transverse and longitudinal sculp- order Archacogastropoda, which is con- ture, and pellets of undiscernable shape. sidered to be a primitive stock beginning in Pellet classification used in the present paper the Lower Cambrian, contains four super- is: oval pellets, unsculplurcd rods, sculp- families with representations in today's tured rods, and shapeless pellets for those seas, the Patellacea, the Neritacea, the pellets which have a loose consistency" and Pleurotomariacea and the Trochacea. The do not maintain a consistent form. Examples order Mesogastropoda (Caenogastropoda, ol fecal pellets are shown in Text-figure 1. part), which is considered to have sprung The classification of Gastropoda has re- from Archacogastropoda, ranges through cently been modified in The Treatise on In- Mesozoic rocks but increases in numbers in vertebrate Paleontology (1960), but as Recent the Cenozoic. The order Neogastropoda f.iAE0^V7"0Z,0G/C/lZ, #07^.5 S31 (Caenogastropoda, part) is mostly confined 8 to Cretaceous and post-Cretaceous de- posits. The fecal pellet forms produced by species of prosobranchs are summarized in a histogram in Text-figure 2. The data sug- /3,Yf/Vf#&46'7y%7/°(%?/4 gest that the rod pellet, both unsrulptured and sculptured, is the primitive type and 4 the oval pellet and shapeless pellet are more advanced types. PELECYPODA 2 The nature of the gills is used by the zool- ogist in subdividing the class Pelecypoda. The order Protobranchia, which contains 0 the suborder Nuculacea, is considered primi- tive; the order Filibranchia containing the suborders Taxodonla and Anisomyaria is 12 considered to be more advanced; the order ' //// Eulamellibranchia, which contains the sub- in _ P orders Schizodonta, Heterodonla, Anape- UJIO donta and Anomalodesmacea, is a further advance. Some pelecypods produce a ribbonlike _MES OGAS %W/06%%4 GO fecal pellet that has not been reported from gastropods; otherwise the pelecypod and gastropod pellets are quite similar. The SPEC fecal pellet forms produced by species of <X> Pelecypoda are summarized in a histogram i in Text-figure 3. The data suggest that the sculptured rod is the primitive type. °4 rr ANNELIDA UJ Yonge (1960, p. 4), after consideration of m 2 - the structure of Ncopilina, states "Mow, no 5 doubt remains that the Mollusca must have 3 separated from the Annelida after the '///. Wl appearance of segmentation; the affinities between the two phyla are much closer than was previously suspected." Therefore, of interest here are the feces of the annelid Eurythoe complanatus which is described by Manning & Kumpf (1959, p. 304) as a short NEOGASTROPODA TF.XT-FIG. 2—Histograms of types of fecal pel- lets produced by the gastropod orders Arch- aeogastropoda, Mesogaslropoda and Neo- gastropoda. Column A—shapeless feces. Col- umn B—oval pellets. Column C—tmsculp- tured pellets. Column D—sculptured pellets. Data mostly from literature. WA A B D 832 r/lZ,gOWTaLOG/C/IZ. AWEiS rod with a longitudinal groove on both the dorsal and ventral surfaces. Although all annelids do not produce sculptured rods, e.g., Arcnicola marina extrudes rods without sculpture (Moore, 1955, p. 516), the fact that an annelid is capable of excreting a sculptured rod, which is characteristic of many gastropods and pelecypods of prim- itive stock, supports the possibility of the sculptured rod being a primitive type. EVOLUTIONARY DEVELOPMENT OF FECAL PELLET'S Moore (1939, p. 517) considers the various shapes of fecal pellets to be modifications of the unseulptured rod. Because of the pre- dominance of sculptured rods among both primitive gastropod and pelecypod stocks, it is the present writer's opinion that the various types of focal pellets may be simpli- fications of the sculptured rod. The un- seulptured rod may be a type transitional between the sculptured rod and oval pellets and shapeless pellets, but the scarcity of tin- sculptured rods among pelecypods suggests that oval and shapeless pellets may have evolved directly from sculptured pellets. If, as suggested here, the primitive fecal pellet is the sculptured rod type, there has been considerable similarity in the development of local pellet types among diverse groups. Primitive feces seem to be extruded as long rods, which then tend to break up into shorter rods because of mechanical weakness _ EULAMELLI'BRANCH!A of the longer rod. A major change in organs producing fecal pellets must have taken place when oval pellets were evolved. In the Littorinidao the oval pellets are lined up in the gut (Abbott, 1954a), showing that in this group they are formed before reach- ing the anus. In general, oval pellets are not completely isolated from each other, but appear like strung beads, one connected to the other. After elimination, the slender connecting links are broken and the oval pellets become discrete bodies. The segregation of coarse sediment par- TEXT-Ftt". 3—Histograms of types of fecal pel- lets produced by pelecypod orders Proto- ticles In one part of a pellet and fine sedi- branchia, Filibranchia and Eulamellibranchia. ment in another has been noted by many Column A—shapeless pellets. Column 8— investigators {e.g., Manning & Kumpf, 1959; oval pellets. Column C—unscuIpLured pellets. Moore, 1932), and is illustrated in Text- Column D—ribbonlike pellets. Column K— sculptured pellets. Data mostly from litera- figure 115 of the present paper. In general, ture. the separation of pellet constituents seems to fVlZ./%W7#Z.0G7C/l,L 7V07\E.S 833 be a primitive trait with composition being stomach by means of the cilia with which more nearly uniform in advanced forms. it is lined," and "In some prosobranch and tectibranch genera anal glands open into EFFECT OF FOOD HABIT the rectum close to the anus, and are con- cerned with the preparation of the feces." Moore (1939) states "In general, carniv- Yonge (1960, p. 15) in discussing primitive orous animals tend to produce faeces of stages in the evolution of the Mollusca loose consistency, vegetable eaters firmer states "The stomach itself probably early ones, and deposit eaters the most resistant contained a ciliated sorting region, an area ol all." The Neogastropoda are primarily of cuticle forming a gastric shield, and a carnivores or omnivorous scavengers and 'style-sac' region initially concerned with their feces are generally of a loose consist- the consolidation of fecal material—a mat- ency and not likely to be preserved as fos- ter of prime importance when the gut opened sils.