An analysis of chewed food particle size and its relationship to molar structure in the primatesCheirogaleus medius andGalago senegalensis and the insectivoranTupaia glis.код для вставкиСкачать
An Analysis of Chewed Food Particle Size and Its Relationship to Molar Structure in the Primates Cheirogaleus medius and Galago senegalensis and the lnsectivoran Tupaia glis WENDY SUE SHEINE AND RICHARD F. KAY Department ofAnatomy, Duke University Medical Center, Durham, North Carolina 27710 KEY WORDS Lemuridae Dentitions - Lorisidae . Tupaiidae - Digestion - Molar tooth structure ABSTRACT The chewed food particle size and shearing capacity of the lower molars of two primate species, the fat-tailed dwarf lemur, Cheirogaleus medius and the bushbaby Galago senegalensis, and an insectivoran, the tree shrew, Tupaia glk, were compared. Differences in the shearing design of the lower molars correlate strongly with the chewed food particle size in these species: the greater the shearing capacity, the smaller the chewed food particles. These three species are of comparable size but differ greatly in diet in the wild. C. medius primarily eats fruit and nectar, while G. senegalensis and T.glis are largely insect-eaters. The lower molars of G. senegalensisand T! glisshow a much greater shearing capacity than do those of C. medius. The average length of chewed food particles of C. medius is significantly larger than that of G. senegalensis, while that of T.glis is intermediate between the two primates but is closer to that of G. senegalensis. Our findings that insect-eating species grind their food more finely than do fruit- and resin-eating species can be correlated with digestibility of foods: finely chewing foods such as fruits which are low in relatively undigestible cell wall components would not greatly improve their digestibility, so a highly efficient food processing apparatus would be less important to the animal's survival. Insect-eaters much more finely chew their foods, implying that there is some constituent of insect bodies difficult to digest, and that grinding increases its digestibility. We suggest that this constituent is chitin. Considerable attention has recently 'been given to differences in the functional organization of the primate dentition and its relevance to feeding behavior (e.g., Kay, '73, '75, in press; Kay and Hylander, in press; Hylander, '75; Zingeser, '73; Rosenberger and Kinzey, '76). However only Walker and Murray ('72, '75) actually attempted to establish some link between molar structure and size of chewed food particles in primate species with different diets. In this paper we attempt to refine and extend their analyses on the assumption that the most meaningful way to characterize the biological role of the dentiAM. J. PHYS. ANTHROP., 47: 15-20. tion, and particularly the triturition mechanism, is t o determine the particle size of chewed foods. Of particular concern here is identifying dental parameters which can be correlated with particle size or shape so that one can assign a "triturition efficiency" rating to a particular molar shape or construction. (Kay ('73, '75) and Kay and Hylander (in press) showed that there are major differences between the molar shapes of insectivorous and frugivorous mammals and inferred that these differences are related to the efficiency of digestion. We present evidence to support this notion. 15 16 WENDY SUE SHEINE AND RICHARD F. KAY TABLE 1 Size (in mm) of food particles in the feces of three species studied. N is the number of particles measured, X is the mean, M is the median, S.D. is the standard deuiutwn, SK is the skewness Length R N Species Cheirogaleus medius Galago senegalensis Tupaia glis 1,000 1,000 1,000 1.54 0.67 0.98 Breadth SD SK x M SD SK 1.32 0.51 0.72 2.73 3.60 1.79 0.81 0.30 0.50 0.63 0.25 0.39 0.63 0.22 0.37 1.90 3.66 1.82 M 1.15 0.54 0.78 TABLE 2 Mean second lower molar dimensions in the three species studied M2 Species (sample size) length Paracristid obliquecristid Cheirogaleus medius (2) Galago senegalensis (5) npaiaglis (4) 2.22 2.22 3.29 0.58 0.94 0.99 1.29 1.04 1.75 MATERIALS AND METHODS Three mature Cheirogaleus medius, the fattailed dwarf lemur, six Galago senegalensis, the bushbaby, and two Tupaia glis, the tree shrew were used in an experiment to determine the size of chewed food particles. The choice of animals was based on an approximate equivalence of body size (to minimize the effects of allometry) and diversity of diet in the wild. Galago senegalensis and Tupaia glis are primarily insectivorous while Cheirogaleus eats much more fruit and resin. A relatively rigid food with a high cellulose fraction, carrots, was chosen as the experimental food on the assumption that the chewed particles would be easily recognizable on the basis of color, and that the cellulose fraction, being virtually undigestible in small animals, would be minimally altered after swallowing. Most important, this food is eaten by virtually all captive primates facilitating future species comparisons. The animals were food deprived for about 12 hours prior to the experimental procedure. Then they were placed in wire bottomed cages and fed pieces of carrot cut into 1 mm X 10 mm x 25 mm slices. No other foods were presented for the duration of the experimental procedure. Ad lib feeding continued for 48 to 72 hours. Feces were collected at 12-hour intervals and preserved in 10% formalin. Feces analysis was undertaken rather than analysis of stomach contents due to the difficulty of stomach pumping operations with small animals (see note added in proof). A sample of feces was placed in a tray of water and shaken gently to separate the particles. The particles were spread on a gridded Hypocristid 0.64 0.70 1.39 Postmetacristid 0.54 1.10 1.12 Preentccristid 0.47 0.82 1.09 filter paper mounted in a Buchner funnel connected t o a filter flask and vacuum. The sides of the funnel were washed with water to insure that all particles rested on the paper. After suction, the paper was removed with forceps and placed under a microscope equipped with a calibrated reticle. Two linear dimensions were measured on each carrot particle-the maximum length and the maximum breadth orthogonal to the long axis of the particle. The total shearing capacity of the postcanine dentitions of each species was estimated by measuring the summed lengths of the M2 shearing blades, the cristid obliqua, post-hypocristid, protocristid, pre-entocristid, and postmetacristid. M, was chosen because its structure is usually representative of that of the postcanines as a whole (Kay, "73, '75). RESULTS A. Behavioral observations The feeding behaviors of the three species were very similar. Each animal picked up a carrot slice in its hand and conveyed it to its mouth. The slice was often held and visually inspected, especially by C.medius and G. senegalensis. A bite was taken by inserting the slice into the side of the mouth in the postcanine region. In most instances the chunk bitten off was masticated thoroughly and swallowed before another bite was taken. Often during chewing the animal held the remainder of the carrot slice in one or both hands. We have very little quantitative data on the number of masticatory cycles each species took for a comparable sized carrot slice. However, a small number of chewing counts CHEWED FOOD PARTICLE SIZE 17 A B C Fig. 1 Lateral views of the lower P,-MBof A, Galago senegalensis (left P,-MJ anterior to the left, M, length = 2.22 mm; B, Cheirogaleus medias, right P,-M, anterior to the right, M2length = 2.22 mm; C, f i p a t a g h (left P,-MJ anterior to the right, Mzlength = 3.29 mm. By comparison with Galago and Tupaia, the molars of Cheirogaleus have low cusps and short shearing blades. did not reveal apparent contrasts among the species. B. Chewed food particle size and shape The mean of the maximum dimension of the chewed carrot particles recovered in feces (table 1) differ significantly. (Probabilities in students t test are p < <0.001.) Although skewed distributions are not legitimately testable with students t, the deviation of the probabilities is probably negligible with such large sample sizes. Cheirogaleus medius had the longest food particles, Galago senegalensis the smallest, with Tupaia glis food particles intermediate in size but closer to G.senegalensis. The mean breadth of food particles of Cheirogaleus medius is significantly larger than those of other species (p<<O.OOl). Tupaia glis food particle breadths are significantly larger than those of Galago senegalensis. All food particle lengths and breadths are right-skewed (table 1).In recognition of this, the shape of each particle was separately calculated as a ratio of length to breadth. Then the ratios were averaged. Thus the ratios presented cannot be calculated from the mean particle length and breadth in table 1. The shapes of the chewed food particles of the species are also significantly different (p< 0.01). Galago senegalensis particles are 2.66 times as long as they are wide, Cheirogaleus rnedius particles are 2.09 times as long as they are broad. Tupaia glis particles are intermediate in shape (2.19). The functional design of the second molars of the species studied (representative of the triturition mechanism as a whole) are quite different (table 2, figs. 1 A-C). The second molars of Cheirogaleus medius have low cusp relief compared with those of Galago senegalensis and Tupaia glis. The crests (which function as shearing blades) are in general shorter and less trenchant: the sum of the five measured crests divided by tooth length is 1.43 for Cheirogaleus, compared with 2.10 for Galago senegalensis and 2.02 for Tupaia glis. Irrespective of diet, the total M2 shearing blade length of primates (estimated by summing the length of the same 5 crests used in table 2) is negatively allometric with respect to M2 length (log, total shearing = 0.34 + 0.91 log, M 2length for 41 species of non-cecopithecid primates) (Kay, unpublished data). 18 WENDY SUE SHEINE AND RICHARD F. K A Y In other words, smaller primates have proportionally longer shearing blades than large primates. This allometric effect must be taken into account in comparisons of the molar shearing capacity of species with different sized molars (Kay, '75). From the equation, an "average" primate with a tooth lengthened the same a s G. senegalensis would have a summed M, shearing blade length of 2.91 mm. Galago senegalensis has 160% of the expected shearingcrest length for a primate with its M, length, Cheirogaleus medius has 109% and the insectivoran, Tupaia glk, is intermediate with 145% compared with the primate model. The differences in the shearing design of the second lower molars correlate strongly with the size of chewed food particles in these comparably sized mammals: the greater the shearing capacity, the smaller are the food particles. DISCUSSION We have shown that the three species studied differ significantly in the size and shape of the food particles in their feces. This observation can be accounted for in several ways. In terms of oral food preparation, either the teeth of some of the species are more efficient in reducing food particle size, or some species simply chew the foods more finely than others. The data in table 2 illustrates that there are major differences in the shapes of the molars and t h a t those species with bett e r developed shearing blades also have smaller food particles in their feces. We have not investigated whether some species are chewing comparably sized items more times than other species. This may be a very important consideration, particularly when the species differ in body size. The possibility must also be considered that the size of food particles in the feces may not be solely a reflection of differences in oral food preparation. One way to approach this problem would be to extract and analyze the food particles immediately after swallowing. However, this imposes some amount of risk to the survival of the subjects, some of which are endangered in the wild. The other way to minimize post-oral food particle size change is to select foods unlikely to be significantly digested. We chose carrots, which have a high cellulose fraction, because there is little likelihood t h a t the species could efficiently process it. For precisely the opposite reason, chitin was unacceptable, because, as we will mention below, we strongly suspect t h a t chitin is relatively highly digestible in some primate species. Carrots have the additional advantage as a n experimental food because their cellulose constituent may closely approximate the physical constancy of chitin, t h e s t r u c t u r a l component of insect exoskeleton. The frugivorous species used in this study chews its foods more coarsely than do the more insectivorous ones. The diet of Cheirogaleus medius in the wild is principally fruit and nectar although they will accept and eat insects and mammals, a t least in captivity (Petter, '62). Judging from stomach contents, Zhpaia glis eats a variety of insects as well a s some fruit (Medway, '66). Galago senegalensis appears to be much more insectivorousinsect fragments were found in abundance in all stomach contents of wild collected specimens examined by Haddow and Ellice ('64), although vegetable matter was also present. Booth ('60) confirms that they eat mainly insects. Why should insect-eaters chew their food more finely than comparably sized fruiteaters? Our proposed explanation requires a review of some factors of carbohydrate digestion. Vegetable substances may be divided into two fractions, t h a t contained in cell walls and that in the metabolically active part of the cell (Van Soest, '66). Cell contents (mainly lipids, sugars, organic acids, starch, nonprotein nitrogen, soluble protein, and pectin) are virtually completely and rapidly digestible by mammals as are some cell wall constituents such as attached protein. Some additional cell wall constituents are undigestible in mammals: lignin, lignified nitrogen compounds, keratin and silica (Van Soest, '66). Finally the digestibility of the hemicellulose and cellulose component of the cell wall depends on the presence of intestinal microorganisms capable of producing the appropriate enzymes (Van Soest, '661, and varies according to the physical and chemical environment for digestion, the length of the period of time allowed for digestion, and the relative surface area to volume ratio of the ingested food particles (McLeod and Minson, '69). The importance of food particle size for digestibility is illustrated by data abstracted from the work of McLeod and Minson (fig. 2). They show that: As food particle size is decreased, digestion is speeded and the total digestibility of the food (related to the percent of structural carbohydrate) increases slightly. 19 CHEWED FOOD PARTICLE SIZE ultimate digestibility: 2 particle i size(crn1 0 Fig. 2 The effects of food particle size on the rate of digestion. The food particle size is plotted against standardized in vitro digestibility at 48 hours for four plant species. For example, if 0%of food a was digested in 48 hours with a particle size of 2 cm, then 11%was digested in 48 hours with a particle size of 1 cm and 18%with a particle size of 0.4 cm. Data from McLeod and Minson (‘691. Plant species: a, Chloris gayana;b, Cenchrus ciliaris; c, Digitaria spp.; d, Sataria spp. End point or “ultimate” digestibility is taken for 0.4 mm particl size for 72 hours. The higher the ultimate digestibility of the food, the less is the effect of reducing particle size. Foods a and b in figure 2 are 56 and 55% digestible and the effect of reducing particle size is quite large. For foods c and d with much higher digestibility the effect of reducing particle size is much less. Since the ultimate digestibility of the foods is related to the percent of cell-wall constituents they contain (Van Soest, ’66) it is clear that reducing the particle size of foods is a more important factor in digestibility among animals which eat high cell-wall constituent foods like leaves, bark, buds and grasses than among those which eat foods low in that fraction. This suggestion is supported by Walker and Murray’s (’75) demonstration that leaf-monkeys grind their foods like leaves, bark, buds and grasses, more finely than do fruit-eating ones. Our findings show that insect-eating species more finely grind their foods than fruit and resin-eating species. Finely grinding fruits and gums generally low in cell wall components would not greatly improve a species’ digestive efficiency. Soft material such as fruits could be crushed to extract the nutritive elements; finely subdividing fruit would not be necessary. The low-cusped dentition of a fruit-eater such as Cheirogaleus medius would be wellsuited t o rupturing plant cell walls without necessarily reducing particle size. Conversely we conclude that there must be some constituent of insect bodies, analogous to the plant cell-wall constituent, which becomes more digestible when it is more finely ground. We suggest that that constituent is chitin. Chitin is a polysaccharide with physical and chemical properties similar to those of cellulose (Wainright e t al., ’76). Although we know of no data on the in vivo digestibility of chitin in mammals, the digestive enzyme chitinase has been isolated from the digestive tracts of several mammals including the lorisoid primate Perodicticus potto which eats insects and is not found in species which rarely feed on insects (Goffart, ’76; Jeuniaux, ’63). (Jeuniaux considered and eliminated the possibility that the production of chitinase is stimulated by the presence of chitin and retarded by its absence in the gastrointestinal tract.) We hypothesize, on the basis of our findings 20 WENDY SUE SHEINE AND RICHARD F. K A Y and those of Jeuniaux, first, that chitin digestion must occur in many small primates and tree shrews, and second, that chewing insect exoskeleton more finely increases its digestibility for small insectovirous mammals. The subject of dental evolution has always been a major focus of primate paleontology. Recently, attention has centered on the interpretation of dental trends in biomechanical terms and, from this, making inferences about feeding changes in the past. These interpretations are increasingly being based on studies of how living animals move their teeth, how forces are transmitted to food and the implications of this for the organization of jaw musculature and tooth shape. Implicit in all this is that the function of the postcanine dentition is to break down food, increasing its surface area and facilitating enzymatic processing. This study, to our knowledge, is the first of its kind to approach the question of how the dentition actually alters food particle size in a controlled experimental situation. When more of this work is done, i t should be possible to interpret the selective factors which have influenced dental evolution. ACKNOWLEDGMENTS This work was supported by NSF Grant GS43262 to Richard F. Kay. LITERATURE CITED Booth, A. H. 1960 Small Mammals of West Africa. Longmans, London. Haddow, A. J., and J. M. Ellice 1964 Studies on bushbabies (Galago spp.) with special reference to the epidemiology of yellow fever. Trans. Roy. Soc. Trop. Med. and Hyg., 58: 521-538. Hylander, W. L. 1975 Incisor size and diet in anthropoids with special reference to Ceropithecidae. Science, 189: 1095-1098. Jeuniaux, C. 1961 Chitinase: an addition to the list of hydrolases in the digestive tract of vertebrates. Nature, 192: 135-136. 1963 Chitin et Chitinolyse. Masson, Paris, 180 pp. Kay, R. F. 1973 Mastication, Molar Tooth Structure, and Diet in Primates. Ph.D. Thesis, Yale University. 375 pp. 1975 The functional adaptations of primate molar teeth. Am. J. Phys. Anthrop., 43: 195-216. - Molar structure and diet in extant Cercopithecidae. In: Development, Function and Evolution of Teeth. P. Butler and K. Joysey, eds. Academic Press, London, in press. Kay, R. F., and W. Hylander The dental structure of Arboreal folivores with special reference to Primates and Phalangeroidea (Marsupialis). In: Arboreal Folivory. G. G. Mongomery, ed. Smithsonian Inst. Publ., in press. McLeod, M. N., and D. J. Minson 1969 Sources of variation in the in uitro digestibility of tropical grasses. J. Brit. Grassland Sw., 24: 244-253. Medway, Lord 1966 Observations on the faune of Pulau Tioman and Pulau Tulai, Part 2, the mammals. Bull. National Mus. Repub. Singapore, 34: 9-32. Petter, J. J. 1962 Recherche3 sur I’kologie e t l’ethologie des lemuriens Malagaches. Mem. Mus. Hist. Nat. (Paris), Ser. A: 1-146. Rosenberger, A. L., and W. G. Kinzey 1976 Functional patterns of molar occlusion in platyrrhine primates. Am. J. Phys. Anthrop., 45: 281-297. Van Soest, Peter 1966 Nonnutritive residues: a system of analysis for the replacement of crude fiber. Jour. A.O.A.C., 49: 546-551. Walker, P., and P. Murray 1972 Particle size of stomach contents a s an indication of energy expenditure proportioning in feeding-foraging activities of selected Anthropoidea. (Abstract). Am. J. Phys. Anthrop., 37: 454. 1975 An assessment of masticatory efficiency in a series of anthropoid primates with special reference to the Colobinal and Cercopithecinal. In: Primate Functional Morphology and Evolution. R. Tuttle, ed. Mouton, Hague, pp. 135-150. Wainwright, S. A,, W. D. Biggs, J. D. Currey and J. M. Gosline 1976 Mechanical Design in Organisms. Wiley, New York, 423 pp. Zingeser, M. R. 1973 Dentition of Brachyteles arachnoides with reference to Alouattine and Atelinine affinities. Folia Primat., 20: 351-390. Note added in proof: Recently we developed a technique which allows us t o sample stomach contents directly, without harm to the animals. Subjects were subdued with a small dosage of Ketamine. A flexible tube was passed orally into the stomach. A small volume of water was introduced into the stomach via the tube, and the stomach contents were withdrawn by exerting gentle pressure on a syringe attached to the free end of the tube. Our results indicate that the average length of carrot particles is not significantly altered by the gastrointestinal tracts of the species studied here. Carrots recovered from the stomachs of G. senegalensis have a mean particle length of 0.44mm, compared with a mean of 0.67mm in fecal particles. The two lengths are not significantly different.