The Committee on Collections of the American Ornithologists' Union is preparing for publication three composite inventories of anatomical collections of the world. One will cover the 45 largest skeleton collections; another, the collections of spirit specimens; and the third, smaller skeleton collections. The inventories will be useful for managing collections and for planning research. Analysis of these inventories shows that some families and orders are well represented in collections but have not been extensively studied; by contrast, many are poorly represented at the generic and specific levels. About a third of the species of birds of the world are not available as skeletons, and the same is true for spirit specimens. Continued collecting is needed to replace specimens that are dissected, to enlarge the series available for most species, to fill gaps that now prevent comparative studies in some families, to build useful research collections in various parts of the world, and to salvage important species in advance of habitat destruction. Received 12 July 1982, accepted 2 August 1982.

National Museum of Natural History, Washington, D.C. 20560 USA; 2The University of Oklahoma, Norman, Oklahoma 73019 USA; and 4Museum of Natural History, University of Kansas, Lawrence, Kansas 66045 USA THE Committee on Collections of the Amer- ican Ornithologists' Union has been active since its inception in publishing information about avian museum collections. Two such reports provided general narrative descriptions of 283 North American collections (Banks et al. 1973, Clench et al. 1976), and a similar project, ex- panded to world-wide coverage, is now under way. In addition, the committee is now com- piling a composite inventory of avian anatom- ical specimens in collections throughout the world. The purposes of this article are to de- scribe the anatomical inventory, provide a brief history of its development and projected plans for the future, summarize some of its data, and discuss the inventory in relation to collection management and research. HISTORY The idea of a composite inventory of avian anatomical specimens in North American col- lections began in 1974 with Mary H. Clench, who then chaired the AOU Committee on Col- lections. Before the project could be initiated, her energies were diverted to a study of major museum collections and their users for the Panel 3 Present address: Carnegie Museum of Natural History, Pittsburgh, Pennsylvania 15213 USA. 740 on Systematics Collections of an NSF-AOU Workshop on a National Plan for Ornithology (King and Bock 1978). Separately, but as part of that panel's report, Zusi distributed a ques- tionnaire to 66 North American museums con- cerning the nature, uses, and curatorial prob- lems of skeleton and spirit collections. From this questionnaire we learned that virtually all of the 44 museum curators who responded were willing to provide lists of their anatomical col- lections. One of the recommendations of the report was that the AOU should publish a composite inventory of anatomical collections (Zusi 1978), and this project then became Zu- si's main responsibility as a member of the AOU Committee on Collections. Beginning in 1979, Zusi requested invento- ries from 66 curators of North American col- lections. Scott Wood became a member of the committee in 1980 and joined him in the in- ventory project. As the new head of the com- mittee, Marion Jenkinson continued support for the project and played an increasingly ac- tive role in its consummation. Wood urged that the inventories be computerized, and he de- signed the necessary algorithms. Gary D. Schnell at the University of Oklahoma gener- ously offered access to computer facilities and contributed personnel for both programming and data entry. He will continue to maintain The Auk 99: 740-757. October 1982 Fig. 1. Reduced photocopy of a sample page from the inventory of avian skeletal specimens. Columns are museum inventories. Letters represent holdings of different size classes (see text), and a blank space indicates the lack of a specimen. and up-date the computer files. Although some inventories were compiled on coding forms by the curators of their respective collections, most were transcribed from the submitted invento- ries by Wood, Mary McKitrick, and Max C. Thompson. Zusi determined the list of species of birds of the world to be used in the inventory, and Jenkinson entered the list onto tape. At her urging we also decided to include major mu- seums from all parts of the world in the survey. Zusi then wrote to 45 curators in 24 countries, and, later, Wood, Jenkinson, and others wrote to additional museums. Curators were re- quested to submit separate inventories of skel- etal and spirit specimens for each species in their collections. We have been much encour- aged by the positive response from most cu- rators. The first published inventories will in- clude a minimum of 40 North American collections and 25 collections from other parts of the world. Before the inventory project was initiated, Zusi (1978) recommended to the AOU that taxonomic and geographic guides to the ana- tomical specimens most critically needed should be prepared. The inventories, because they are in taxonomic order with all species of birds of the world listed, will provide an idea of the need for specimens at various taxonomic levels. In addition, the AOU has received a grant (DEB- 8205935) from the National Science Foundation that will permit us to analyze the inventories from a geographic standpoint. That work is now under way, and the results should be pub- lished in about two years. THE COMPOSITE INVENTORIES Three kinds of composite inventories will be published separately as bound computer print- outs (with citable titles and publication dates). Their titles will be: Inventory of avian skeletal specimens, 1982; Inventory of avian spirit specimens, 1982; and Supplementary inven- tory of avian skeletal specimens, 1982. The for- mat for the first two is described below; that for the third will differ and be as condensed as possible. Figure 1 shows a sample page from the in- yentory of larger skeleton collections. Each of the 46 numbered columns contains a museum inventory (in decreasing order of size from left to right). In the published version, column 46 will hold the totals from all museums in the inventory of smaller skeleton collections. Be- cause each column except the first is only one character wide, all entries greater than 9 are represented by the following codes: A = 10- 15; B = 16-25; C = 26-50; D = 51-100; E = 101-500; F = (>500). The first column, how- ever, will give the actual total number of spec- imens for each species. The systematic species list follows the no- menclature and arrangement of Morony et al. (1975), with a few minor changes and the fol- lowing major exceptions: volume 1 (Mayr and Cottrell 1979) and volume 8 (Traylor 1979) of Peters' "Check-list of birds of the world" are followed, and new names have been added from Bock and Farrand (1980) and from card files in the National Museum of Natural His- tory. The numbers presented in the inventories represent, to our knowledge, complete speci- mens. Because of lack of time or assistance, however, some curators did not separate par- tial from complete skeletons, carcasses from complete spirit specimens, or chicks from adults in their specimen counts. Persons using these inventories to locate specimens for study should verify with the curator the precise nature of the specimens in question. We shall add new data and new inventories to the computer files as they are received. When the changes in a composite inventory become significant, a new edition will be published. Instructions for ordering the composite inven- tories will appear in The Auk, The Condor, The Wilson Bulletin, The Ornithological News Letter, and in major ornithological journals through- out the world. SUMMARY OF DATA In comparison with skin collections, ana- tomical holdings are very small. The largest skin collections are approximately 35 times the size of the largest skeleton collections, and in many museums the number is much higher. Table 1 summarizes the specimen holdings of the 36 largest skeleton collections for which we re- ceived inventories as of 1 June 1982, and Table 2 provides specimen totals for all of the spirit collections inventorled as of that date. We have also received 28 skeleton inventories from col- lections that either contain fewer than 1,000 skeletons each or, if larger, were received too late to be included in Table 1. These museums are (in alphabetical order): Acad. Nat. Sci. Philadelphia, Arch.-zool. Arbeitsgruppe (Schleswig), Australian Mus., Barbados Mus., Bolton Mus. (England), British Columbia Pro- vincial Mus., Bulgarian Acad. Sci., Charleston Mus., Czechoslovakian Acad. Sci., Cleveland Mus. Nat. Hist., Forschungsinstitut Sencken- berg (Frankfurt), Glasgow Mus., Koninklijk Mus. Midden-Afrika (Tervuren), Merseyside County Mus. (Liverpool), Midwestern State University (El Paso), Mus. Dierkunde (Gent), Mus. Nac. Hist. Nat. (Santo Domingo), Nat. Mus. Bulawayo, Polish Acad. Sci., St. Bona- venture Univ., Univ. Connecticut, Univ. Hel- sinki Zool. Mus., Univ. Montana, Univ. New Mexico, Univ. Texas E1 Paso, Uppsala Univ. Zool. Mus., Virginia Commonwealth Univ., Washington State Univ., Wayne State Univ. (Detroit). In Table 3 the anatomical data are summa- rized by family and subfamily. The number of these families or subfamilies in which all gen- era or species are represented by skeletons or spirit specimens is high (71% for both skele- tons and spirit specimens), but most of these are very small families or subfamilies contain- ing fewer than 4 genera or 5 species. Families with more than 10 genera, all of which are rep- resented by skeletons, are the Procellariidae, Threskiornithidae, Anatidae, Scolopacidae, Alcidae, Hirundinidae, and Mimidae; for spir- it specimens they are the Procellariidae, Scol- opacidae, Hirundinidae, Laniidae, Mimidae, Estrildidae, Ploceidae, and Corvidae. Based on the inventories we have received, skeletons are apparently lacking only for the Oxyruncidae, Neodrepanidinae, and Hypoco- liinae, and spirit specimens for Atrichornithi- dae. Only 1-5 skeletons have been reported for each of the following families or subfamilies: Pedionomidae, Dromadidae, Micropsittinae, Phodilinae, Brachypteraciidae, Philepittinae, Atrichornithidae, Pseudochelidoninae, Salpor- TABLE 1. Total numbers of specimens and species in the 36 largest skeleton collections for which inventories were received by 1 June 1982. Museum Specimens Species Royal Ontario Mus. Univ. Kansas Natl. Mus. Nat. Hist., Washington, D.C. Univ. Michigan Mus. Vert. Zool., Berkeley Louisiana State Univ. American Mus. Nat. Hist. Peabody Mus., Yale Florida State Mus. Carnegie Mus. Nat. Hist. British Mus. (Nat. Hist.), Tring Washington State Mus. Field Mus. Nat. Hist. Univ. Oklahoma Univ. South Florida Los Angeles County Mus. Delaware Mus. Nat. Hist. Mus. Comp. Zool., Harvard Rijksmus. Nat. Hist., Leiden a Inst. Haustierkunde, Kiel Univ. Miami, Florida Cornell Univ. Natl. Mus. Canada Mus. Para. Emilio Goeldi, Belem San Diego Mus. Nat. Hist. Paleo. Inst., USSR Acad. Sci., Moscow Mus. Nat. Hist., G6teborg C.S.I.R.O., Canberra Univ. Arizona California Acad. Sci. Natl. Mus. New Zealand, Wellington Univ. California, Long Beach Mus. Alexander Koenig, Bonn Univ. California, Los Angeles Moore Lab. Zool., Occidental College Alberta Prov. Mus. 25,234 1,816 22,533 1,633 20,459 3,409 16,646 3,145 10,704 1,752 8,194 2,182 7,893 2,261 7,124 1,587 7,077 598 6,133 864 5,580 2,000 4,489 535 4,332 1,412 3,987 502 3,922 1,169 3,706 892 3,423 717 3,284 1,221 3,228 1,510 3,186 230 2,687 689 2,663 459 2,619 283 1,798 285 1,796 638 1,762 802 1,634 365 1,607 520 1,558 460 1,466 500 1,382 231 1,336 401 1,255 467 1,134 394 1,069 456 1,029 173 From van Oort (1907). nithinae, Rhabdornithidae, Catamblyrhyn- chinae, and Cnemophilinae. Apparently only 1-5 spirit specimens exist for the Casuariidae, Anseranatinae, Pedionomidae, Aramidae, Ibid- orhynchidae, Strigopinae, Phodilinae, Lepto- somatidae, Oxyruncidae, Neodrepanidinae, Hypocoliinae, Tichodromadinae, Salporni- thinae, and Cnemophilinae. Table 3 reveals that some families and subfamilies for which ana- tomical specimens are available in large num- bers are, nonetheless, poorly represented at the generic and specific levels. Table 4 lists the number of species repre- sented by a given number of specimens. Roughly a third of the species of birds of the world are not available as skeletons nor a sim- ilar fraction as spirit specimens. Many more species that are available in collections are rep- resented by 10 or fewer specimens than by more than 10 for both skeletons and spirit speci- mens. The species best represented by skeletal specimens is the House Sparrow (Passer do- mesticus), with approximately 9,000. A few oth- er species number between 1,000 and 3,000: Savannah Sparrow (Passerculus sandwichensis), Horned Lark (Eremophila alpestris), European Starling (Sturnus vulgaris), Dark-eyed Junco (Junco hyemalis), Ovenbird (Seiurus aurocapillus), Red-eyed Vireo (Vireo olivaceus), Yellow- rumped Warbler (Dendroica coronata), Brown- headed Cowbird (Molothrus ater), and Northern Fulmar (Fulmarus glacialis). No species is represented by as many as 1,000 spirit speci- TABLE 2. Total numbers of specimens and species in collections of spirit specimens for which inventories were received by 1 June 1982. Museum Specimens Species Natl. Mus. Nat. Hist., Washington, D.C. Koninklijk Mus. Midden-Afrika, Tervuren British Mus. (Nat. Hist.), Tring Peabody Mus., Yale Univ. American Mus. Nat. Hist. Royal Ontario Mus. Delaware Mus. Nat. Hist. Univ. Kansas Field Mus. Nat. Hist. Louisiana State Univ. Carnegie Mus. Nat. Hist. Mus. Vert. Zool., Berkeley Los Angeles County Mus. Mus. Alexander Koenig, Bonn Univ. Michigan Washington State Univ. Natl. Mus. Canada C.S.I.R.O., Canberra Australian Mus., Sydney Univ. Connecticut South Australian Mus., Adelaide Natl. Mus. New Zealand, Wellington Texas A&M Univ. Moore Lab. ZooL, Occidental College California Acad. Sci. Univ. Uppsala Mus. Nacional, Santo Domingo 14,362 2,438 14,352 735 12,395 2,909 9,192 1,997 8,949 2,372 3,672 957 3,602 665 3,453 698 3,121 990 2,945 1,114 2,666 840 2,497 658 2,130 480 1,294 474 1,266 503 993 504 955 161 717 253 634 289 549 150 353 128 288 85 229 113 206 143 189 83 92 76 82 48 mens. The five species most numerous in col- lections are: Red-headed Quelea (Quelea ery- throps), Village Weaver (Ploceus cucullatus), Common Bulbul (Pycnonotus barbatus), Vireo olivaceus, and Passer domesticus. Some species have been collected in large numbers for studies of skeletal variation (e.g. Passer domes- ticus, Johnston 1969, Johnston and Selander 1971; Passerculus sandwichensis, Rising, in pro- gress; Eremophila alpestris, Niles 1973). Others are received in large numbers by museums be- cause of TV-tower kills (e.g. Seiurus aurocapil- lus, Vireo olivaceus, Dendroica coronata ). MANAGEMENT OF ANATOMICAL COLLECTIONS The inventories prepared by curators for this project will undoubtedly prove useful to them in reviewing the adequacy of their collections with respect to their particular goals and in judging the strengths and weaknesses of their collections in relation to regional and world- wide resources. Focusing attention on anatom- ical collections will, we hope, lead to improve- ments in their care and management. We also hope that the deficiencies exposed by the com- posite inventory will provide a stimulus for the collecting of new material. The publication of synoptic inventories will certainly stimulate exchanges among museums and provide more even coverage of material in various parts of the world. Arranging ex- changes can be very time-consuming, but the inventories can be used to maximize the ben- efits of this work. It is far more useful for a curator of a North American museum, for ex- ample, to add a representative of a genus not found in any other North American collection than simply to add one new to his or her mu- seum. The inventories will allow curators of large collections that raditionally receive the major burden of loan requests to suggest additional sources to borrowers and allow a more equi- table distribution of loans. This will clearly benefit the borrower and the large museum. Also, the small collection will thereby become more useful to the research communiby, which may be an important criterion for its continu- ing support. October 1982] Inventory of Avian Anatomical Specimens 745 746 Zus, WOOD, AIID JE:so [Auk, Vol. 99 October 1982] Inventory of Avian Anatomical Specimens 747 uu * ,-t-.c* * * * c* * * uu* * * * . 748 ZUSI, WOOD, AND JENKINSON [Auk, Vol. 99 October 1982] Inventory of Avian Anatomical Specimens 749 750 Zusi, WOOD, Arab JmaIttasota [Auk, Vol. 99 October 1982] Inventory of Avian Anatomical Specimens 751 752 Zust, WOOD, AND JENKINSON [Auk, VoL 99 October 1982] Inventory of Avian Anatomical Specimens 753 754 Zvs, WooD, ^tad JENKNSON [Auk, Vol. 99 TAnLE 4. Numbers of species represented by var- ious amounts of anatomical material. Preparation Numbers of Spirit specimens Skeletons specimens 0 2,706 2,957 1 919 1,027 2-5 1,757 1,949 6-10 1,026 1,086 11-25 1,216 1,128 26-50 614 501 51-100 364 '237 101-200 214 89 >200 189 31 Finally, we hope that presentation of these composite inventories, incomplete as they are, will stimulate those in charge of unreported collections to submit their inventories to one of us for inclusion in a future printing. RESEARCH Undoubtedly, the inventories will foster an- atomical research by facilitating the finding of specimens and the design of feasible research projects and by encouraging the collecting of new material. They dearly demonstrate that, while a few species are represented by large numbers, most are represented by few speci- mens or are lacking. This prompts us to ask whether or not we need large series of anatom- ical specimens (>100 per species), and whether or not existing collections meet the research needs of the present and immediate future. The answer to the first question is "yes" for the following reasons. 1. Analyses of infraspecific variation in the avian skeleton are rare for lack of large samples of most species. Series of the same age, sex, season, physiological state, or geographical or- igin may be required. Such series may be im- possible to assemble, even from large numbers of skeletons or spirit specimens. It is not rea- sonable to expect museums always to have the necessary specimens on hand, but they should have enough to indicate the potential of a pro- posed project. 2. Skeletal specimens are often poorly cleaned, incomplete, or damaged; simple totals do not reflect the number of useful specimens. Furthermore, skeletons are either articulated or disarticulated. Paleontologists, archaeologists, and those analyzing hawk and owl pellets need disarticulated, well-cleaned specimens; sys- tematists and functional anatomists often need articulated skeletons as well. Spirit specimens are sometimes poorly fixed or stored, and many were in substandard condition before they were fixed and are thus too hard, soft, or deterio- rated for proper dissection. Difficult dissec- tions may have to be done repeatedly to estab- lish an anatomical relationship and to reveal a range of variation. 3. Skeletal specimens are relatively durable, but delicate ones may be broken with use, and small bones can be lost in packing and un- packing. Some spirit specimens are kept more or less intact as distributional records or doc- umentation for published statements, but in general they achieve usefulness only in pro- portion to the degree to which they are dis- sected. Study of one anatomical system may ruin that specimen for other studies. Apart from such legitimate destruction, Burton (1980) has emphasized the hazard of loss of spirit speci- mens from drying. Thus, we see that both skel- etons and spirit specimens should continue to be collected as replacements, even if augmen- tation of the collection is not an objective. 4. Several circumstances suggest a need to develop strong collections throughout the world. The collections useful for broad system- atic research are mostly concentrated in North America and Europe. Laws, designed to pro- tect populations of birds from over-exploita- tion by commercial trade, have inadvertently imposed burdens on museums by requiring permits or excessive documentation for routine international loans. Specimens are lost or dam- aged by postal and other systems with increas- ing frequency. Some museums prohibit loan of specimens represented by only one or a few individuals in their collections, and their cu- rators have a responsibility to maintain collec- tions for resident scientists and visitors. All of these facts point to the need for a higher degree of regional independence in satisfying the need for specimens. Do existing collections meet the needs for present and future research? The answer is "only in part" for the following reasons. 1. The emphasis of systematic studies based on comparative anatomy in recent decades has changed from broad surveys of a few species or genera per family to detailed comparisons of many species and genera within families or orders. Many of the "characters" such as pal- atal types, leg muscle formulae, and toe ar- rangements used by earlier systematists are not necessarily homologous throughout birds. Furthermore, the states of such characters are less diagnostic of orders and families than was previously thought. It is important to survey as many species and genera as possible in any higher taxon to determine the variation and evolutionary trends in such features. Some or- ders and families that need further study are well represented in anatomical collections (e. g. Procellariiformes, Podicipediformes, Pelecani- formes, Anseriformes, Gruiformes, Charadri- iformes, Tyrannidae, and many oscine fami- lies). Reference to Table 3, however, will show that other groups are not well enough repre- sented in spirit or skeletal collections to permit extensively comparative studies in systemat- ics, functional morphology and adaptation, ecomorphology, and other fields. Stresemann (1959) thought that studies based on anatomical evidence have solved relatively few phylogenetic problems and that they are usually incapable of doing so. We think, how- ever, that anatomical studies sufficiently com- parative to support phylogenetic arguments are only beginning to be made and that compar- ative anatomy will become increasingly im- portant in the elucidation of avian phylogeny. 2. Limitations of museum collections for studies on infraspecific variation, or even for pilot projects on such studies, have already been mentioned. CONCLUSIONS For broadly comparative studies or analysis of variation within species, the world's anatom- ical collections are often inadequate. Further selective collecting is needed on a broad scale. The forthcoming composite inventories are in- tended to maximize the efficiency of this en- deavor. In the course of collecting new speci- mens it is essential to keep in mind that full data with each skeleton and spirit specimen are as important as they are with a skin. Field col- lectors, if not thoroughly familiar with the sci- entific uses of specimens, should become ac- quainted with the principles and practices presented in such papers as Miller (1940), Van Tyne (1952), Parkes (1963), Zusi (1969), and Burton (1980). We also remind collectors and curators that, when they obtain a rare or little known species as a fresh specimen, it may make a more significant contribution to knowledge as a spirit specimen or skeleton (and partial skin) than as a traditional skin. We must emphasize that loss of habitat, de- clining avian populations, restrictions on col- lecting, and political hostilities are all obstacles to the development of collections. It is vitally important that ornithologists in all parts of the world build strong anatomical collections of their local avifaunas and of other avifaunas readily accessible to them, before the habitats and birds are gone. Otherwise, many ornitho- logical questions of great significance will re- main unanswerable for lack of material. ACKNOWLEDGMENTS We extend special thanks to all of the individuals who sent us inventories of the collections in their care. Others who helped us obtain inventories through correspondence or personal contacts were Walter Bock, Joe Marshall, Peter Morgan, Kenneth Parkes, and Lester Short. We are indebted to Chris- topher Wood for inventorying a major skeleton col- lection that would otherwise not be included. The University of Oklahoma has generously provided the computer services that make this project feasible. Mary Clench read the manuscript and made some valuable suggestions. LITERATURE CITED BANItS, R. C., M. H. CLENCH, & J. C. BARLOW. 1973. Bird collections in the United States and Canada. Auk 90: 136-170. Boci(, W. J., & J. FARND, JR. 1980. The number of species and genera of Recent birds: a contri- bution to comparative systematics. Amer. Mus. Novit. 2703: 1-29. BURTON, P. J. K. 1980. Studies on functional anat- omy in birds utilizing museum specimens. Pp. 190-194 in Actis 17th Congr. Intem. Ornithol. (R. Nohring, Ed.). Berlin, Deutsch Ornitholo- gen-Gesellschaft. CLENCH, M. H., R. C. BANItS, &J. C. BARLOW. 1976. Bird collections in the United States and Canada: addenda and corrigenda. Auk 93: 126-129. JOaNSTON, R.F., 1969. Character variation and ad- aptation in European sparrows. Syst. Zool. 18: 206-231. , & R. K. SELANDER. 1971. Evolution in the House Sparrow. II. Adaptive differentiation in North American populations. Evolution 25: 1- 28. KING, J. R., & W. J. Boche. 1978. Workshop on a national plan for ornithology. Final report to Na- tional Science Foundation and American Orni- thologists' Union. M,3/4R, E., & G. W. COTrRELL, Eds. 1979. Check- list of birds of the wodd, vol. 1, 2nd edition. Cambridge, Massachusetts, Mus. Comp. Zool. MLrER, A. H. 1940. Field techniques in collecting for a research museum. Mus. News 17(17): 643. MORON3/4, J. J., JR., w. J. BocI<, & J. FARRAND, 1975. Reference list of the birds of the world. New York, Amer. Mus. Nat. Hist. NiLs, D.M. 1973. Adaptive variation in body size and skeletal proportions of Horned Larks of the southwestern United States. Evolution 27: 405- 426. OORT, E. D. 1907. Catalogue ostologique des oiseaux, Mus. Hist. Nat. des Pays-Bas, Cat. des Collections, vol. 10, pt. 1. Leiden, E. J. Brill. P,mcs, K. C. 1963. The contribution of museum collections to knowledge of the living bird. Liv- ing Bird 2: 121-130. STRSM,NN, E. 1959. The status of avian system- atics and its unsolved problems. Auk 76: 269- 280. TRAYfoR, M. A., JR., Er). 1979. Check-list of birds of the world, vol. 8. Cambridge, Massachusetts, Mus. Comp. Zool. V,N TYNE, J. 1952. Principles and practices in col- lecting and taxonomic work. Auk 69: 27-33. Zusi, R. L. 1969. The role of museum collections in ornithological research. Proc. Biol. Soc. Washington 82: 651-661. 1978. Anatomical collections. In Workshop on a national plan for ornithology. Final report to National Science Foundation and American Ornithologists' Union (J. R. King and W. J. Bock, Co-Directors).