University of Nebraska - Lincoln DigitalCommons@University of Nebraska - Lincoln
Papers in Natural Resources Natural Resources, School of
September 1992
Morphometrics of the Family Emballonuridae
Patricia W. Freeman University of Nebraska-Lincoln, [email protected]
Cliff A. Lemen University of Nebraska-Lincoln, [email protected]
Follow this and additional works at: https://digitalcommons.unl.edu/natrespapers
Part of the Natural Resources and Conservation Commons
Freeman, Patricia W. and Lemen, Cliff A., "Morphometrics of the Family Emballonuridae" (1992). Papers in Natural Resources. 19. https://digitalcommons.unl.edu/natrespapers/19
This Article is brought to you for free and open access by the Natural Resources, School of at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in Papers in Natural Resources by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln.
Published in Contributions to Mammalogy in honor of Karl F. Koopman (T.A. Griffiths and D. Klingener, Eds.), Bulletin of the American Museum of Natural History, No. 206, pp. 54–60. Copyright © 1992 American Museum of Natural History.
Used by permission.
American Museum of Natural History http://www.amnh.org/
Bulletin of the American Museum of Natural History http://library.amnh.org/pubs/bulletin.html
R N\~)H, ISo. 206 : 5'4-*b lqq'L
Morphometrics of the Family Emballonuridae
PATRICIA W. FREEMAN1 AND CLIFF A. LEMEN2
ABSTRACT
Morphometric analysis revealed three distinc- all other emballonurids. Phylogenetic studies also tive groups among the genera of emballonurids. separated Taphozous-Saccolaimus as distinctive Taphozous-Saccolaimus is a group distinctive in but included diclidurids among other New World size and shape, particularly cranially. Diclidurids species. Compared with molossids, emballonurids are distinctive in appendicular characters only, es- are morphometrically quite homogeneous. pecially those in the wing. The third group includes
INTRODUCTION Miller (1 907) thoroughly described most recently Barghoorn ( 19 77) examined cranial of the extant families of bats and illustrated morphology of a fossil genus, Vespertiliavus, several of the genera with exquisite line draw- and all Recent genera for their possible phy- ings. His work remains valuable for its clarity logenetic relationships, and Robbins and Sar- and for establishing a description of and qual- ich (19 88) produced a phylogenetic study of itative differences among the morphologies the family using protein electrophoresis and of families and genera of Chiroptera. immunology. We have been interested for some years in the quantitative morphological differences among genera within families and among families (Freeman, 19 8 1; Lemen and Free- Thirty-eight meristic characters (27 cranial man, 1981, 1984). In this morphometric and 1 1 appendicular) were studied on 37 spe- treatment of the family Emballonuridae, the cies of emballonurid bats. These are standard sheath-tailed bats, we examine the quanti- measurements and, for the most part, a sub- tative differences among the species and gen- set of those in Freeman (1981). There are era within the family, compare the morpho- some changes, however, because of structural metric groupings with groupings from two differences between molossids and emballo- phylogenetic studies, and describe the mor- nurids. Because there is no comparable third phometric variation between the families phalanx and cartilaginous tip of digit I11 in Ernballonuridae and Molossidae, the free- emballonurids, the quantity measured for tailed bats. both families is the length from the second Early generic treatments of emballonurids phalanx to the tip of that digit. Postorbital include Troughton's (1 925) revision of the and interorbital breadths in some emballonu- Australasian genera Taphozous and Sacco- rids had to be measured inferior to any su- laimus, Sanborn's (1 937) study of American praorbital bone overhanging those breadths species, and Tate and Archbold's (1939) ex- in order to measure the least constriction. amination of the genus Emballonura. More This was particularly true with the diclidu-
' Curator of Zoology and Associate Professor, Division of Zoology, University of Nebraska State Museum, Lincoln, Nebraska 68588-05 14. * Research Associate, Division of Zoology, University of Nebraska State Museum, Lincoln, Nebraska 68588-05 14. 1991 FREEMAN, LEMEN: EMBALLONURID MORPHOMETRICS 55 rids. The quantity SIZE equals the sum of the opogon, G. Taphozous nudiventris, E. natural logs of greatest skull length (which Taphozous perforatus, I. Taphozous theobal- differs slightly from condylocanine length; di, J. Saccolaimus flaviventris, K. Saccolai- used in Freeman, 1984, 1988), zygomatic mus peli, and L. Saccolaimus saccolaimus. breadth, and height of braincase. SIZE cor- The following measurements were taken relates well with weight of the animal (Free- (descriptions and illustrations are in Free- man, 1988). man, 198 1): Cranial: greatest skull length, Both bivariate and multivariate analyses palatal length, maxillary toothrow, upper were used to assess the data, including prin- molariform row, lacrimal width, interorbital cipal components analysis with and without width, postorbital width (POSTORB), zy- a "shearing" function. The shear method de- gomatic breadth, breadth at mastoids, breadth scribed by Bookstein et al. (1985) is partic- of braincase, height of braincase, height of ularly useful in our analysis because it gen- upper canine, length M3 (M3LENGTH), erates a size factor based on within-genus width M3, width at upper canines, width at comparisons and not the first principal com- upper molars, dentary length, dentary-con- ponent of the entire data set, as used in "size- dylocanine length, condyle to M3 length, out." The distinction between the sizeout ap- lower toothrow, moment arm of temporal, proach and shearing is particularly important moment arm of masseter, height of coronoid, in this data set because the bats in the Ta- dentary thickness, height of condyle above phozous-Saccolaimus group are much larger toothrow, height of lower canine, and length than the majority of Emballonuridae. The of condyle; Appendicular: tibia, forearm, third sizeout approach will tend to define shape metacarpal, third metacarpal first phalanx differences between Taphozous-Saccolaimus (PHAL 1M3), third metacarpal second pha- and other bats as size-related differences. This lanx to tip, fourth metacarpal, fourth meta- would indicate that bats of the Taphozous- carpal first phalanx, fourth metacarpal sec- Saccolaimus group are not different in shape. ond phalanx, fifth metacarpal, fifth Shearing does not use the size differences metacarpal first phalanx (PHAL 1MS), and among groups in its definition of size, and in fifth metacarpal second phalanx. this case Taphozous-Saccolaimus is found to Abbreviations used in the text include PC 1, have considerable shape differences com- principal component one; PC2, principal pared to other Emballonuridae (fig. 1). Fi- component two; PC3, principal component nally, simple regression analyses were per- three; H2, sheared component 2; and H3, formed on each character versus the SIZE sheared component 3. quantity. We examined the following species for this study: a. Balantiopterx io, b. Balantiopterx plicata, c. Centronycteris maximiliani, d. Co- We thank curators of the American Mu- leura afra, e. Cormura brevirostris, f. Dicli- seum of Natural History, the Field Museum durus albus, g. Diclidurus ingens, h. Dicli- of Natural History, and the National Muse- durus isabella, i. Diclidurus scutatus, j. um of Natural History for use of their spec- Emballonura alecto, k. Emballonura atrata, imens. Early data-gathering trips by Freeman 1. Emballonura beccarii, m. Emballonura were supported by the Field Museum. Com- monticola, n. Emballonura nigrescens, o. puter analyses were performed and graphics Emballonura raflrayana, p. Emballonura produced by equipment in the newly estab- semicaudata, q. Emballonura sulcata, r. Per- lished Mary B. Totten Center for Biosyste- opteryx leucoptera, s. Peropteryx kappleri, t. matics Technology located in the University Peropteryx macrotis, u. Peropteryx trinitatis, of Nebraska State Museum. Drs. Thomas A. v. Rhynchonycteris naso, w. Saccopteryx bil- Griffiths and Don E. Wilson kindly reviewed ineata, x. Saccopteryx canescens, y. Saccop- the manuscript and Mark Marcuson, staff art- teryx leptura, A. Taphozous australis, B. ist, assisted with graphics. Finally and most Taphozous hamiltoni, C. Taphozous hilde- importantly the senior author thanks Karl gardeae, D. Taphozous longimanus, E. Ta- Friedrich Koopman (pronounced "Cope- phozous mauritianus, F. Taphozous melan- mun" as would KFK) for serving as one of BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 206
A CRANIAL AND APPENDICULAR B CRANIAL AND APPENDICULAR
CRANIAL D APPENDICULAR
Fig. 1. The first three principal components of an analysis of the emballonurid data set with cranial and appendicular measurements (A and B). Separate analyses using sheared components were run on cranial measurements alone and on appendicular measurements alone (H2 and H3 in C and D). Species are indicated by letters, which are listed in Methods.
her mentors at the American Museum; for who unselfishly shares that knowledge; and patiently wading through new methodology for being a continual source of stimulation and giving excellent help as a member of her and friendship. dissertation committee; for leaving charac- teristic, cryptic, and dependable notes on RESULTS specimen tags of bats and other mammals in most of the major collections around the Morphological trends in the data, revealed country, particularly at Field Museum; for by principal components analysis, revolve being a colleague of encyclopedic knowledge around general size, several wing measure- 1991 FREEMAN, LEMEN: EMBALLONURID MORPHOMETRICS 57 ments, width-of-face measurements, and a ponents two and three can be seen as ex- tooth measurement. Eighty-seven percent of tremes from the regression line to a greater the variation in the family can be explained or lesser degree in the bivariate plots (fig. 2). by the first component, which is related to a The simple plots clarifl and verify the mul- change in size. Size typically explains most tivariate picture so that it is easy to see what of the variation in morphological studies on characters influence the principal compo- quantitative characters and is typically the nents. first principal component. Component two explains 3.5 percent of the variation and is influenced primarily by the length of the first DISCUSSION phalanx of digit I11 (shortest at positive end), Morphological relationships within the postorbital width (widest at positive end), emballonurids parallel the phylogenetic hy- length of first phalanx of digit V (longest on potheses of Robbins and Sarich (1 988) in positive side), and interorbital width (widest some cases, and run contrary to them in oth- at positive end). Component three explains ers. The most basic split discovered by Rob- 1.9 percent of the total variation and is in- bins and Sarich was Taphozous and Sacco- fluenced by length of second phalanx of digits laimus versus the rest of the emballonurids. IV and V (longest at positive end) and length Our data show that Taphozous and Sacco- of M3 (longest at positive end). laimus are a distinctive group in size and The placement of species on the first two cranial shape. This is a case where time has components, shows species of Taphozous- increased morphological distinction between Saccolaimus well separated from all other groups. The next most distinctive group is genera except Diclidurus because of their larg- the genus Diclidurus. Its skull morphology is er size (fig. 1A). Diclidurus is distinct from similar to that of Balantiopterx, and both oc- all other genera because of its wing configu- cupy an extreme of H2 (fig. 1C). However, ration (unusually short first phalanx of digit Diclidurus is distinct in wing morphology. I11 and long first phalanx of digit V), wide The recognition of this genus as a separate postorbital and interorbital breadths, and subfamily is based largely on postcranial somewhat longer M3s. On PC3 Diclidurus is morphology, particularly the shape of the less cohesive because Diclidurus isabella has clavicle and the construction of the tibia, but a short second phalanx of digit IV, and of also, because the cranium has a wide supra- digit V to a lesser degree, while the other three orbital bone that overhangs the interorbital/ species in the genus have long ones (fig. 1B). postorbital region (Miller, 1907; Koopman, All other emballonurid genera- Emballonu- 1984b). Electrophoreticdata indicate that Di- ra, Coleura, Rhynchonycteris, Saccopteryx, clidurus belongs within the large group of New Cent ronycteris, Peropteryx, Cormura, and World genera. If Robbins and Sarich (1 988) Balantiopteryx-are in one large indistin- are correct, this is a case where morphological guishable group. distinctiveness does not indicate phylogenet- Using the shearing method to remove the ic distance. effect of size gives somewhat different results. Another finding of Robbins and Sarich The main difference is that the Taphozous- (1 988) was recognition of the Emballonura- Saccolaimus group forms a distinctive mor- Coleura group of Old World bats versus the phological entity based on cranial characters New World genera. This differs from Barg- but not appendicular characters (fig. 1C). Di- hoorn's (1 9 77) placement of Coleura with the clidurus is highly distinctive based on fea- New World forms. However, his placement tures of the wing but much less so for cranial of Coleura was based on the loss of an incisor. features (fig. 1D). Tooth reduction may occur in unrelated taxa, In examining the makeup of the multivar- reducing the reliability of this character. iate analyses, we regressed each of the 38 Overall, we prefer the grouping of Robbins characters against a composite quantity to and Sarich (1988). Actually, the electropho- represent size (see Methods). A sample of the retic data indicate that Emballonura is para- characters that are heavily loaded on com- phyletic, with Coleura included within. Our BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 206
SlZE SlZE
7:0 8.0 9.0 SlZE Fig. 2. Bivariate plots for variables (natural logs) first phalanx of digit I11 (A), postorbital breadth (B), first phalanx of digit V (C), and length of M3 (D) against SIZE (see Methods for explanation of SIZE). Lines plotted in each scattergram are linear regression lines; the relevant statistics for the lines are (A) a = 0.43, b = 0.29, P < 0.0001; (B) a = - 1.52, b = 0.40, P < 0.0001; (C) a = -0.36, b = 0.44, P < 0.0001; and (D) a = -0.56, b = 0.07, P < 0.03. Species are indicated by letters, which are listed in Methods. morphological data indicate a close similarity teryx, Peronymus, Centronycteris, and Bal- between Emballonura and Coleura. There is, antiopteryx) under the name Saccopteryx and however, no great distinction between these stated in a footnote that "As in many other two genera and the New World forms. cases, but to an exaggerated degree, I here Miller (1907: 85) stated under principal unite a number of units almost universally subdivisions in the family that "the genera called genera by modern mammalogists. They of Emballonuridae as a whole form a very are however, manifestly and closely allied, homogeneous group, but the South American cover less morphologic range than do many Diclidurus is so different from the others that genera, and include so few species that ge- it must be regarded as forming a distinct sub- neric separation has no practical value. This family." Simpson (1945: 55) lumped many seems an obvious case, one of many, in which of the New World genera (Cormura, Perop- subgeneric rank has everything to be said for 1991 FREEMAN, LEMEN: EMBALLONURID MORPHOMETRICS
A CRANIAL
+Or
-1.54) -!2 -1 0 12 3 C CRANIAL AND APPENDICULAR
Fig. 3. Principal components analyses of the two-family data set, run on cranial measurements alone (A), appendicular measurements alone (B), and cranial and appendicular measurements together (C). Emballonurids are solid squares and molossids are open circles. Emballonurids are as variable as mo- lossids in size (PC1, first column), but are much less variable in shape (PC2 and PC3). 60 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 206 it, both as better representing the real situa- data, we conclude that emballonurids when tion and as practically more convenient to compared to molossids morphometrically are everyone but the Saccopteryx specialist." a homogeneous group. This homogeneity may We investigated these qualitative claims of help explain why the emballonurids have been homogeneity by using the same 38 characters difficult to classify above the species level. measured previously by Freeman (1 98 1) for the family Molossidae and comparing entire families with one another. Both families are REFERENCES insectivorous, both occur worldwide, and for Barghoorn, S. F. both we had over 75 percent of the total spe- 1977. New material of Vespertiliavus Schlos- cies in the family represented in our analysis. ser (Mammalia, Chiroptera) and sug- However, emballonurids are thought to be gested relationships of emballonurid bats primitive and are placed in Koopman's in- based on cranial morphology. Am. Mus. fraorder Yinochiroptera, whereas the molos- Novitates 26 18: 29 pp. sids are derived and are in the Yangochirop- Bookstein, F. L., B. C. Chernoff, R. L. Elder, J. tera (Koopman, 1984a). M. Humphries, G. R. Smith, and R. E. Strauss 1985. Morphometrics in evolutionary biolo- One way to compare size and shape di- gy. Acad. Nat. Sci. Philadelphia Spec. versity in two families is to compare the Publ. 15: 277 pp. amount of variation that is and is not ex- Freeman, P. W. plained by the first principal component. In 198 1. A multivariate study of the family Mo- emballonurids, the total variation of char- lossidae (Mammalia, Chiroptera): mor- acters is 3.5 1, of which 3.09 (88%) is ex- phology, ecology, evolution. Fieldiana, plained by the first principal component. This Zool., n.s., 7: 173 pp. leaves only 0.4 18 (1 2%) for the "shape" com- 1984. Functional cranial analysis of large an- ponents. In molossids, the total variation is imalivorous bats (Microchiroptera). 1.95, with the first component explaining Biol. J. Linn. Soc. 2 1: 387-408. 1988. Frugivorous and animalivorous bats 1.182 (6 1°/o), and the remaining 0.769 (39%) (Microchiroptera): dental and cranial on the "shape" axes. The conclusion that can adaptation. Ibid., 33: 249-272. be drawn is that the emballonurids are not Koopman, K. F. as variable in shape as are the molossids, and 1984a. A synopsis of the families of bats -part this lack of variation can be seen in a variety VII. Bat Res. News 25: 25-27. of graphical representations. 1984b. Bats. In S. Anderson and J. K. Jones, We have run the two-family data set for Jr. (eds.), Orders and families 3f Recent variation in cranial measurements alone, ap- mammals of the world, pp. 145-1 86. pendicular measurements alone, and cranial New York: Wiley. and appendicular measurements together (fig. Lemen, C. A., and P. W. Freeman 1981. A test of macroevolutionary problems 3). The size component (PC1) in the cranial with neontological data. Paleobiology 7: run shows that although there are more 311-315. smaller-sized species of emballonurids than 1984. The genus: a macroevolutionary prob- molossids, variation in size across the fam- lem. Evolution 38: 1219-1237. ilies is similar. Sizes among the two families Miller, G. S. from the appendicular measurement run show 1907. The families and genera of bats. Bull. a similar degree of variation. U.S. Natl. Mus. 57: 282 pp. However, it is in the shape components, Robbins, L. W., and V. M. Sarich here represented by PC2 and PC3, that em- 1988. Evolutionary relationships in the family ballonurids show much less variation (fig. 3). Emballonuridae (Chiroptera). J. Mam- Although molossids are more variable than mal. 69: 1-13. Sanborn, C. C. emballonurids in shape in each of the three 1937. American bats of the subfamily Em- runs, the most dramatic difference in varia- ballonurinae. Zool. Ser. Field Mus. Nat. tion can be seen in the graph of the two shape Hist. (Chicago) 20: 321-354. components in the run with all 38 characters Simpson, G. G. (fig. 3C; PC2 versus PC3). Based on these 1945. The principles of classification and a FREEMAN, LEMEN: EMBALLONURID MORPHOMETRICS
classification of mammals. Bull. Am. Troughton, E. L. G. Mus. Nat. Hist. 85: 350 pp. 1925. A revision of the genera Taphozous and Tate, G. H. H., and R. Archbold Saccolaimus (Chiroptera) in Australia 1939. A revision of the genus Ernballonura and New Guinea, including a new spe- (Chiroptera). Am. Mus. Novitates 1035: cies and a note on two Malayan forms. 14 PP- Rec. Australian Mus. 14: 3 13-339.