New Directions for Bioacoustics Collections Author(s): Sandra L. L. Gaunt, Douglas A. Nelson, Marc S. Dantzker, Gregory F. Budney, and Jack W. Bradbury Source: The Auk, 122(3):984-987. 2005. Published By: The American Ornithologists' Union DOI: http://dx.doi.org/10.1642/0004-8038(2005)122[0984:NDFBC]2.0.CO;2 URL: http://www.bioone.org/doi/full/10.1642/0004-8038%282005%29122%5B0984%3ANDFBC %5D2.0.CO%3B2

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. Commentary

The Auk 122(3):966–971, 2005 © The American Ornithologists’ Union, 2005. Printed in USA.

Bird Collections: Development and Use of represent the original “big science” expendi- a Scientifi c Resource.—Bird collections were tures in the life sciences, antedating by cen- founded and built during the heyday of global turies large contemporary endeavors such as exploration. The mission of these collections genome projects. Long-term investments in the through most of their history has been to docu- development and maintenance of collections ment avian diversity and its distribution and to have produced a resource equivalent in many serve as a resource for research and education. respects to the mega-science facilities found in As bird collections became established and grew, other disciplines (e.g. large telescopes or super- ornithology itself became a scientifi c discipline colliders). A key diff erence is that this resource and broadly expanded its purview. Today, there has a useful lifespan that greatly exceeds that of are more professional ornithologists than at any mechanical facilities. Bird collections need to be time in history, and collections-related research viewed as a highly versatile and indispensable represents only a small portion of the discipline. resource integral to the continued successful This is healthy. Collections are but one means (and economical) pursuit of a wide range of through which we study birds. But we cannot be subjects. Importantly, these subjects are no lon- lulled into a view that the day of the collection is ger restricted to ornithology. past—a decided risk when fewer ornithologists Unlike other, contemporary “big science” have direct experience either with collections projects, biological collections establish an or with the multidimensional strengths that a object legacy—continuing sources of data that collections-based approach brings to science. are repeatedly tapped to provide answers to Too li le a ention has been paid to the role of questions about birds and environmental con- bird collections in science. This role is changing. ditions. Many of these questions were not even Because we understand avian diversity be er imagined by those who have built these collec- than that of most other classes of organisms, tions. Indeed, as the ways in which museum the central goal for establishing bird collections specimens are used multiplies with conceptual, would seem to be largely accomplished. As any- technological, and environmental changes, it one studying avian diversity knows, however, is clear that we need to refocus on how best to much remains to be done: systematics, pa erns continue developing this resource. and processes of diff erentiation, and geographic variation in birds remain vigorous areas of new Sujhnrjsx fsi Shnjshj ymwtzlm Tnrj learning. But while those of us close to collections remain connected to these important questions, Mismatches in temporal relevance.—The classic to others what we do is increasingly arcane. analogy that natural history collections are like AV er all (tongue in cheek), isn’t there already libraries invites direct comparisons between a fi eld guide? To an undiscriminating public, specimens and books. There is some mate- much of this work would seem to be done once rial similarity to these objects; old books and a fi eld guide appears. And, indeed, explorations bird skins are products of animal skin and at that level in most regions are largely complete. plant materials. With care, books can easily But in other respects the scientifi c strengths of last for many centuries, and a useful lifespan collections-based research are blossoming—with of more than a millennium is likely. Eff ective no end in sight to the fruits that can be borne. preservation of bird skins has been practiced Bird collections are probably the strongest for only about two centuries, but the skins will and most dependable shared resource in clearly last far longer. Beyond simple object ornithology. Biological collections in general comparisons, consider use: it is routine in

966 July 2005] Commentary 967 collections-based work to use specimens 100 “Arrogance of the present.”—The scientifi c capa- years old or older, and historical specimens are bilities and accomplishments of today are truly increasingly used to conduct retrospective stud- impressive when compared with those of yester- ies—research asking questions about changes year. But this will be equally true in the future, in birds and the environments they live(d) in. when the accomplishments of today will be over- As time passes, preserved specimens increase shadowed by those of our successors. Thus, to in scientifi c value. By contrast, the value of be er serve our science, we should consider how journal articles and books we produce today we might contribute to a future 50 or 100 years will be relatively short-lived (see ISI Journal from now. What do we have that they will not? Citation Reports; www.isinet.com); aV er only a A key asset of the present is access to a biota that few decades most of our papers will no longer is still probably half intact. There is strong evi- directly contribute to science. Thus, in terms dence that passing samples of this biota forward of utility and relevance, these two products a to future researchers is one of the most eff ective researcher can leave behind—specimens and ways to contribute to the accomplishments of publications—have strikingly diff erent trajec- future science, and that archiving specimens tories. Publications are important. But the long- will enhance the eff ectiveness of future wildlife term scientifi c value of specimens is widely management and conservation (e.g. through the underappreciated, and we probably place too growing importance of retrospective studies). much emphasis on producing publications, the more ephemeral of the two products that we Gtnsl Ftw|fwi might leave as our scientifi c legacy. “Biological fi lter paper.”—Just as historical Collections are helping to answer widely anthropogenic objects like books refl ect changes important questions about birds and our shared in human history, so too can preserved specimens environments, and meeting the needs of users enable us to appreciate and measure changes. is a central purpose of a collection. As the user One of the most important developments in col- community grows, so too should support and lections-based science is the retrospective study, participation in continued collections develop- a very powerful approach for assessing changes ment. And the strengths of collections must be in populations and environments. Specimens considered broadly, separately from individual document life in three dimensions: geographic research programs. space (locality), biodiversity space (taxonomy), With increasing frequency, specimen loan and time (date). The last dimension is becom- requests ask diff erent questions of preserved ing increasingly important, because historical material than the questions for which that mate- samples enable us to enlist this strong analytic rial was originally preserved (e.g. feather pluck- approach to measure and understand change. ing of skins for genetic, isotopic, or contaminant Probably the broadest reward this science brings analyses, and disease screening of genetic sam- to society as a whole is through the increasing ples). In fact, one cannot predict what question, use of specimens as “biological fi lter paper,” specimen type, or taxon the next loan request documenting “experiments” in the environments will hit upon. This is a double-edged sword: it in which these animals lived. In many studies, a reassures us that collections are broadly useful, species or genus is chosen because it represents an but it suggests that collections growth is becom- important trophic level at which to measure bio- ing increasingly out of touch with collections- accumulation or magnifi cation of contaminants. based science. As the array of possible uses has These measurements have profound implications increased, our ability to foresee what the speci- for humans and the natural resources we man- men needs of tomorrow will be has declined. age, and these studies repeatedly demonstrate There is li le doubt that there will be need, that historical samples are crucial. Birds are excel- however. Important questions about changes in lent bioindicators of environmental conditions, populations and environments will expand the and bird specimens should continue to be at the need for specimens, and adequate sample sizes forefront of this fi eld. Continued acquisition of from today will be required. new specimens should be seen as a priority. The So it is clear that the resource itself must time dimension can be studied only with contin- continue to be developed. Collections were ued sampling of the avifauna. built on general acquisition policies, and their 968 Commentary [Auk, Vol. 122 broad usefulness today refl ects this. Continued is science that tests hypotheses has in some growth should follow this course (in a guided respects been detrimental to collections. Simple manner), and care must be taken to foster this exploration, description, and comparison—the growth without exclusive reliance on focused types of science upon which collections were research programs. Individual research pro- established—are today “poor cousins” in terms grams are important components of collections of recognition, funding, and publication. This growth, but healthy growth requires a broader has occurred despite the obvious importance of diet. Wider user needs are not likely to be met collections in establishing the baselines of our by a focus on the scientifi c questions of today, understanding of biodiversity and its distribution which, although important at present, tend to in space and time. Specimens and collections can, be either taxonomically narrow or broad but with planning, continue to establish and provide shallow in sample size. Even the best-focused the baselines from which hypotheses are devel- research of today will not meet tomorrow’s oped and tested, in dimensions that are both tra- specimen needs—unless, in aggregate, we work ditional (e.g. biodiversity) and nontraditional (e.g. to increase that likelihood. Thus, most collec- environments, contaminants, and diseases). Our tions-based biologists and curators advocate science will be strongest when we recognize and a general acquisitions policy for our shared support these multiple legitimate pathways and collection resources. But it is important to note contributions to how knowledge is developed. that we recognize costs and commitments and Hypothesis testing is an important component that we work to maximize the gains that each of individual research programs. But hypothesis archived specimen represents. Most eff ort goes testing alone, except in the most general manner to fi ll gaps (taxonomic, geographic, and tem- (e.g. “things will change”), is an inadequate basis poral) and to increase sample sizes to enhance for continued collections growth. statistical power. The challenge is that there An altogether diff erent phenomenon has are a lot of gaps, especially in temporal and emerged around humans killing birds for numerical dimensions. science. Fastidiousness in this regard wildly The preserved objects themselves, as continu- outstrips our responses to other, much greater ing sources of new information, have primacy sources of avian mortality, and collecting for over associated data. Recognizing that diverse, science is singled out for astonishing levels of long-term scientifi c gains are achievable, more restriction and scrutiny. The contortions through components of individual birds are being pre- which many permi ing agencies are willing to served, such as skin, partial skeleton, tissues, go to squeeze scientifi c sampling (nearly out of and stomach samples. Given the surprises existence in many cases) is truly amazing. Most that technology and science have derived from of this regulation is not biologically defensible, bird specimens thus far, it is not too outlandish and it is not conservation. Permi ing is useful to suppose that interdisciplinary teams (e.g. and necessary, but it oV en seems to be a politi- ornithologists, entomologists, parasitologists, cal means of imposing belief systems on others, virologists, isotope ecologists, computational aff ecting government and wildlife management and systems biologists, and community geneti- to the detriment of science and society. It is cists) will one day delve into a treasure trove of ironic that the very agencies that would benefi t preserved avian stomach and tissue samples to most from the knowledge that specimens would extract complex network analyses of environ- deliver to their management programs are so ments, communities, and biospheres. The fact oV en hostile to this knowledge development. that modern bird specimens are preserving so This is self-handicapping behavior. many a ributes of today’s biota speaks to the Individual researchers also need to examine strengths of whole-organism sampling and how their beliefs aff ect their work and their sci- preservation. Ornithology has yet to develop entifi c commitment. The act of killing a bird for a more economical or effi cient method than science is diffi cult for those who do it and seem- whole-organism sampling to accomplish such ingly a personal barrier for those who choose important, multidimensional documentation. not to. Those not directly involved oV en have This must be more widely understood. strong beliefs surrounding this act. Many have Fashion trends are trumping strong science.—The not sorted out their feelings and beliefs about fashionable concept that the only good science scientifi c collecting, nor have they thought July 2005] Commentary 969 through the long-term consequences of their will maximize scientifi c gain for the resources choices to either support or not support this expended in avian research. activity. Two beliefs are oV en involved: belief in Specimen salvage.—Many birds are killed the sanctity of the life of an individual bird, and inadvertently by humans (e.g. in collisions with belief in an erroneous worldview of conserva- towers, windows, or vehicles, or by pet cats) or tion in which every individual ma ers. I do not die naturally. These are potential specimen “sal- wish to demonize these beliefs, but it is wrong vages,” and this source is greatly underutilized. to impose them on others. Too oV en, permit- On average, however, salvaged specimens are of ting systems are being used to do so, which less value than specimens actively collected for is detrimental to research and management. science. Concentrations in geographic and taxo- Equally problematic are choices by researchers nomic space limit the eff ectiveness of salvage, as that cause these belief systems to diminish the does the oV en mutilated or ro ing condition of scientifi c eff ect of their own work. Too many salvaged birds. An active salvage program can workers go out of their way to avoid collecting rapidly fi ll local gaps and become saturated in birds, even if their studies would be be er for the areas (geographic and taxonomic) repre- it. Usually, science in general would be be er sented in the salvage stream. Thus, museum served if they collected as part of their eff orts. interest in salvage usually comes below the inter- I long ago integrated collecting with banding, est in actively collected specimens, because the because I realized that I was releasing most of la er have been taken with science specifi cally the data I was working so hard to obtain. in mind and are thus more useful and in be er Except when noncollecting is necessary, condition. But salvage has value, and in one owing to factors such as small, fragile popula- respect this value is underexploited. The relative tions or the requirements of a study to follow concentrations of salvaged birds in geographic living birds, specimens should be an expected and taxonomic space represent an opportunity product of most fi eld ornithology. By not deliv- to obtain birds from areas not usually collected ering fully on the scientifi c promise of an eff ort, (e.g. cities and parks), to archive rare species time and money are oV en ineffi ciently used and from captivity, and to develop large sample sizes thus—in part—wasted. This can be viewed as of some species. These opportunities seem to be misallocation of scarce resources. Some will rarely exploited, perhaps because the questions argue that time does not allow them to both these samples can be used to address tend not collect and accomplish their goals. If those goals to be in the realm of traditional museum stud- are strong science, some change is warranted. ies. However, their value for biomonitoring, for Even banders should be preserving all acciden- example, is high, and this is an area where agen- tal casualties. Researchers should understand cies, governments, and individuals unwilling to that by choosing not to collect (or, if killing kill birds should be actively developing partner- birds violates a personal belief, choosing not to ships with museums. preserve salvaged specimens), they are dimin- Monitoring and surveillance.—Birds play a ishing their own scientifi c legacy. One can retain prominent role in environmental monitoring, a strong respect for life and conduct good sci- yet we oV en lack good baseline data against ence that includes preservation of specimens. which to measure change. Too much avian Everyone working with birds should consider monitoring and surveillance involves only the details of these issues. Conservation must counting animals. This is like taking health focus on populations; every individual dies. and disease statistics without addressing cau- In most populations, every individual does not sation. Collecting, preparing, and archiving ma er. Study population biology. Do the math. bird specimens through time is an economical Recognize that bird populations are a renewable way to enable implementation of retrospective resource, and that scientifi c collecting represents studies when perceived changes occur, allowing a practically insignifi cant (and non-additive) detection of correlative changes that may have proportion of annual avian mortality. In turn, happened in such things as contaminants, food, specimens provide multiple benefi ts to science, and habitat use (e.g. through stable-isotope wildlife management, conservation, and society. analyses), diseases, parasites, genetic diversity, The vast majority of bird populations can easily sex and age structure, and traditional pheno- support the small amounts of collecting that typic parameters. This approach is eff ective for 970 Commentary [Auk, Vol. 122 the monitoring and surveillance of populations, of them unpredictable), and adequate numbers species, and environmental change. And the util- of specimens can be highly eff ective in docu- ity of these materials is not directly correlated menting these changes, for both contemporary with time. The other two dimensions that col- and retrospective studies. lections document (positions in biodiversity and Individual researchers also have much to geographic spaces) can produce important base- gain by partnering with a repository. Properly line information almost immediately. Specimens archiving your specimens and samples for do not need to “ripen” for years to be useful. As perpetuity has widely recognized scientifi c ben- a discipline, we have not eff ectively planned efi ts. Such partnerships are best arranged before how best to do this job of using science for eff ec- writing proposals for funding and permits (and tive environmental stewardship. Nonlethal fi eld certainly before initiating fi eldwork). Proposals work, counting, monitoring, and surveillance all are strengthened by an archival component, and remain important, but it is imperative to couple most funding agencies support the added costs these approaches with a sample-based compo- that ensure proper preparation and preserva- nent if we are to maximize our success in bird tion. Those costs are generally a small portion conservation and management. of the total, and writing them into proposals, Leveraged research and partnerships.—Collecting with input from the repository, is now routine. and sending specimens or specimen loans, espe- Reviewers are increasingly (and properly) cially to people at other institutions and in other expecting to see this. In their turn, reposito- countries, indirectly leverages research support ries need to grant researchers specifi c rights to (i.e. time and money) for the species being col- research priority, with a sunset clause (e.g. rights lected. Thus, through specimens, a management of refusal on potentially competing user requests agency, an institution, a state, or a country can being guaranteed for fi ve years or as long as the oV en get work done that they themselves could researcher is conducting active research on the not do or could not aff ord to do, providing material). These partnerships are among the easi- increased knowledge about that resource. (And I est to generate, because the immediate gains are emphasize the importance of basing such eff orts apparent to all. Moreover, this approach ensures on vouchered specimens.) Indeed, it is surpris- that important research material and associated ing that more resource management agencies data are not lost to science because of a local and their permi ing personnel do not recognize freezer failure or a lab or offi ce cleaning. this important and eff ective means of inexpen- Supporting the resource.—Archiving speci- sively learning about the resources they manage. mens costs money and time, and museums can- The most eff ective managers will be those with not carry the burden alone. For example, just the best science to apply to their management sending dead birds to a museum does not get plans; courting the appropriate researchers with the job done. Preparation capacities at museums specimens and samples is highly eff ective in are always saturated, and this is a bo leneck. developing such partnerships. Enhancing preparation capacity would have Two other important partnerships are those immediate benefi ts, and a broader distribution between institutions (e.g. museum and state or of preparation activities would work to the federal agencies) and those between research- strengths of museums as repositories. This is a ers and a repository. Agencies responsible for community resource, and it is time to develop resource management are increasingly dis- community-wide solutions to collections devel- sociated from the sample-based perspective opment and maintenance. These solutions of museums, a situation that is harming us should be local, regional, and national, but a all. With our shared goals of understanding key basis is that those who are users or ben- and successfully managing and conserving efi ciaries of specimens and related data need to wildlife, we are natural partners, and we need become supporters and contributors. to bridge this divide and begin working more eff ectively together to obtain and archive the Ctshqzxntsx specimen resources that will enable the very best science and scientifi c management. The Bird collections are a community research best “bridge” is general and specimen-based: resource and provide broad benefi ts to our changes will occur in many dimensions (many science, to the management and conservation July 2005] Commentary 971 of birds, and to society. In addition to strong maintenance-free in comparison with skins, contributions in traditional research, collections spirit specimens, or tissues. They are data-rich are making important, long-term contributions and, to some, aesthetically striking—yet avian to issues that have li le to do with the reasons osteological specimens have always been a for their establishment. These contributions are minor constituent of museum bird collec- oV en more important to society than the original tions. Systematic ornithology of the previous reason(s) for making the collections, and this centuries focused on plumage and external needs to become part of the planning and reward morphology, and birds collected in the fi eld processes for continued collections growth. were perforce transformed into the compact, Presently, most growth is focused on short-term round skin specimen that forms the bulk of the gains; yet, as products of science, the specimens world’s museum collections. These traditional themselves have a much longer useful life than skin specimens were considered of paramount the publications generated from them. We recog- importance and, to the collectors and curators, nize this, in part, simply by maintaining existing represented “value—money value and scientifi c collections. The next step is to make new invest- value” (Coues 1874). ments to enhance future gains. There is clear The traditional preparation of skin specimens indication today of a need to collect, prepare, leaves only the cranium and the distal elements and archive specimens, and to do so in a way that in the wings and feet, and early collectors usu- increases the array of preserved components (i.e. ally discarded the remaining bones and the animal parts), sample sizes, and dimensions (in torso. Sometimes, however, partial osteological biodiversity, geographic, and temporal spaces) specimens comprising the axial skeleton, femur, available to present and future researchers. A lot and humerus or other combinations were pre- of the world’s biodiversity will ride the conserva- pared from the torso as an ancillary step in the tion coa ails of successful avian management scientifi c collection process, usually only if there and conservation. As ornithologists, we have the were time available aV er the higher-priority opportunity and obligation to lead in this area. skins were prepared. OV en, single elements Specimens have been and should continue to be were preferred, and some collections special- an integral part of the science behind monitor- ized in synoptic collections of crania or sterna. ing, managing, and conserving our biological Olson (2003) points out that because sterna resources. Together we can direct collections were easy to obtain from skinned torsos, they growth to establish the baselines that we know were oV en the element of choice in anatomical will enhance our eff ectiveness in sound environ- collections, even in cabinets of curiosity (Fig. mental stewardship.—Kj{ns Wnspjw, University 1). The great French encyclopedist l’Herminier of Alaska Museum, 907 Yukon Drive, Fairbanks, constructed a classifi cation of birds based solely Alaska 99775, USA. E-mail: [email protected] on sterna (l’Herminier 1827), and Coues and other 19th-century ornithologists distinguished Ahpst|qjilrjsyx several taxa (e.g. Pelecaniformes, Alcidae) on the presence or absence of a perforate nasal I thank D. Causey, J. A. Cook, G. R. Graves, (Coues 1872). and J. V. Remsen, Jr., for discussions and T. Although exceptions exist, avian skeletons of Braile, D. D. Gibson, R. A. Z. Meier, R. M. Zink, the past were most oV en prepared as mounted and an anonymous reviewer for comments. displays and thus, as is oV en the case even today, data were secondary to the aesthetics of presentation. Consequently, osteological speci- mens collected before the mid-20th century The Auk 122(3):971–979, 2005 are oV en incomplete or data-poor, or comprise © The American Ornithologists’ Union, 2005. mixed proveniences—particularly those used Printed in USA. as reference collections for bone identifi ca- tion. Olson (2003) provides a succinct history Old Bones in New Boxes: Osteology of avian osteological collections and should Collections in the New Millennium.— be consulted for a more complete background Skeletons and bones are the most durable on their development and the nature of early specimens in avian collections. They are nearly specimens. Here, we explore some possible 972 Commentary [Auk, Vol. 122

The disadvantage of these skeletal specimens is that their preparation precludes nearly any other type of specimen, with the exception of tissue specimens, because the skin and inter- nal anatomy of the collected bird are usually destroyed in the process. Partial skeletons have elements missing; the most common partial skeleton prepared in modern treatments leaves only a leV or right set of distal elements (e.g. ulna, radius, carpometacarpus, hallux) in the skin, the distal end of the maxilla, and other bones necessary for a durable skin specimen (Winker 2000). Partial skeletons encompass a Fnl. 1. Peafowl (Pavo cristatus) MCZ 342364. broad range of included elements (sometimes Sternum with collection information noted on only the torso elements, sometimes articulated ♀ specimen: “Pavo indicus Linn./16 years old./ sterna and clavicles, etc.). Mounted skeletons ♂ which assumed the plumage of a /Specimen are nearly always complete (but not entire; see mounted./F. Peabody. Jan, 1864”. below) and usually were intended for display, but some of the earliest research specimens reasons for the relative unpopularity of avian were mounted with the elements articulated for osteological specimens, examine the value and movement. Mounted specimens are diffi cult to use of bony specimens in modern ornithologi- use for research, because many of the character- cal research, and suggest possible directions rich locations on bones near their articulating and solutions for the future. surfaces are oV en obscured by wires or holes, and early preparators would oV en use bones Czwwjsy Syfyzx tk A{nfs Oxyjtqtlnhfq from several specimens to “part out” missing or Ctqqjhyntsx pathological elements. Lots, or bulk collections of bones, are most Avian osteological specimens come in sev- commonly associated with subfossil or archeo- eral diff erent forms, from complete skeletons logical se ings. The Museum of Comparative with all major bones to single isolated elements Zoology (Harvard University) and the U.S. (Table 1). Complete skeletons are the founda- National Museum (Smithsonian Institution), for tion of most modern osteological research col- example, have numerous large boxes of many lections: in these, every major bone of the bird thousands of Great Auk (Alca impennis) bones is preserved, including leV and right elements. gathered from slaughter sites on Grand Funk

Tfgqj 1. Classes and features of avian osteological material.

Class a Entire b Associated c Articulated d Vouchered e Complete Usually Yes ? Usually Partial No Yes ? Usually Mount Usually ? Yes ? Lots No No No ? Elements (Recent) No No No OV en Elements (Fossil) No No No No a Complete specimens have all major bones (paired elements plus axial elements) usually kept for scientifi c study. Partial specimens lack some major elements, purposefully or through a rition. Mounts are assembled skeletons intended for display. Lots are bulk collections of bones, sometimes sorted by element type (e.g. humerus), some not. Elements are single bones of recent or fossil origin. b Entire specimens have all bones, including minor elements such as phalanges and hyoids; thus, most complete specimens are entire, whereas not all mounts are entire. c Associated specimens comprise bones from only one individual. d Articulated specimens have adjacent elements joined, oV en by dried ligaments or skin and rarely by wire, string, etc. e Vouchered osteological specimens were originally identifi ed by reference to the intact bird when collected; thus, by defi nition, no fossil is vouchered. July 2005] Commentary 973

Island, Canada. While nearly every major bone also be articulated by dried ligaments and skins, is represented, very few of these specimens are as in some early complete and partial specimens associated; that is, there is no way to determine or in semi-prepared skeletons. which bones came from the same bird. Single ele- By far the most problematic issue with osteo- ments are most commonly associated with fossil logical specimens is vouchering. A vouchered specimens or specialized reference collections for specimen was identifi ed by reference to the comparative studies. As mentioned above, early actual bird when it was collected or prepared. avian anatomists focused on sterna or crania, for Partial skeletons oV en lack easily diagnosable example, to the exclusion of other bones. elements, and mistakes in identifi cation or other Each of these forms of osteological specimens lapses during preparation can lead to errors can be characterized by four primary features: very diffi cult to detect later (see below). whether they are entire, associated, articulated, or vouchered. Entire specimens have every bone Szw{j~ tk Mfotw Ctqqjhyntsx preserved from the collected bird. Although most complete specimens are entire, there are Overall, skeletal specimens constitute ~7% of many which, as a consequence of collection or the total specimens held in 10 of the largest orni- preparation mishaps, lack a few very small or thological collections (Table 2). There are some delicate bones, such as the hyoid, the cranial notable variations from this general pa ern, xiphoid of pelecaniform birds, the alula, and so a ributable to the particular history of a museum on. Entire specimens off er the greatest research collection. For example, the number of osteologi- value, but are rarest in the early specimens. An cal specimens at Florida State Museum exceeds associated skeleton originates from a single bird, the number of skin specimens by ~30%, whereas and nearly every complete or partial specimen those at the British Museum represent only ~1% with data is associated; mounts may or may not of the total collections. The Florida collections be; and lots and single elements, by defi nition, were, in large part, formed by Pierce Brodkorb, a are never associated. Articulated elements are leading avian paleontologist of the 20th century. what make a prepared mount, but elements can The British Museum’s collection started out as,

Tfgqj 2. Relative proportion of skeletal specimens to traditional skin specimens at selected top avian collections, ranked by number of skeletal specimens. Data from various sources, including Wood et al. (1982), Mearns and Mearns (1998), and J. Hinshaw (pers. comm.).

Skeletons/ Skeletons/ Museum a Skeletons Skins skins (%) b total (%) c USNM 51,931 513,000 10 9 FMNH 49,294 360,199 14 12 ROM 44,268 135,972 33 25 AMNH 25,000 850,000 3 3 FL 23,238 17,794 131 57 UMMZ 23,086 171,225 13 12 KU 21,463 53,401 40 29 LSU 21,000 142,000 15 13 MVZ 19,537 159,283 12 11 BM ~15,000 ~1,000,000 ~2 ~1 MCZ ~6,000 ~340,000 ~2 ~1 Total 300,000 3,782,874 8 7 a AMNH: American Museum of Natural History; BM: British Museum of Natural History; FL: Florida State Museum; FMNH: Field Museum of Natural History; LSU: Louisiana State University; KU: University of Kansas Museum of Natural History; MCZ: Harvard University Museum of Comparative Zoology; MVZ: University of California, Berkeley, Museum of Vertebrate Zoology; ROM: Royal Ontario Museum; UMMZ: University of Michigan Museum of Zoology; USNM: United States National Museum. b Proportion of skeletal specimens to skin specimens. c Proportion of skeletal specimens to total number of specimens. 974 Commentary [Auk, Vol. 122 and remains, the premier collection of bird skins; than the comparative approach. Second, space skeletons never had a chance there. Similarly, is always a factor, and some collections are osteological collections represent a large part of unable to expand beyond their current size. the University of Kansas collections, because of But expansion is needed—more specimens are the long history of skeletal preparations begun needed in avian museum collections. by Charles Bunker in the fi rst decade of the 20th More specimens are needed because many century (Hall 1951, Johnston 1995). species of birds are still unrepresented by The number of osteological specimens in even a single skeletal specimen. For example, a collection does not necessarily correlate 30% of tinamou species have no osteologi- with the depth of taxonomic coverage. The cal specimens, and 67% of the genera have U.S. National Museum has by far the greatest unrepresented species (Table 4). In more spe- diversity of specimens, with 5,109 species rep- ciose orders, the pa ern is similarly bad. In resented (Table 3), more than half the known Apodiformes, for example, 33% of species have species of birds. The Field Museum of Natural no specimens, 46% of genera have species with History and the Royal Ontario Museum rank at no skeletal specimens, and 74% of species have the top of the list for mean number of specimens ≤10 specimens. For Caprimulgiformes, 40% of per species, an index that relates to the collect- species have no specimens, 50% of genera have ing eff ort for series of specimens rather than species without any specimens, and overall, single examples. The British Museum collec- 84% of species have ≤10 specimens. tion, while ranking about seventh in number of Clearly, osteological specimens are under- species, has only ~5 specimens per series. These represented in the top museum collections, but data refl ect several interacting factors. First, the the global situation is likely much worse than top-ranked institutions have dynamic collecting this. Osteological specimens have migrated from programs that continue to preserve osteological smaller museums and natural history cabinets specimens. By contrast, osteological collections into the collections of major museums, thus begun in the late 19th and early 20th centuries skewing the skeleton:skin ratio higher in those were usually intended for the study of com- few institutions. The frequency of skeletal col- parative function and morphology. The mod- lections among all scientifi c collections may be ern emphasis on the study of geographic and much lower than 8%. For example, several recent population variation demands more specimens estimates of the total number of bird specimens existing in world collections range from 8 to 10 million (Banks et al. 1973, Goodman and Lanyon Tfgqj 3. Depth of skeletal specimen holdings in 1994, Mearns and Mearns 1998), whereas the ten top avian collections. Data from Wood et total number of skeletons and other osteologi- al. (1982). cal specimens probably does not exceed 500,000 Mean (Wood et al. 1982, Wood and Schnell 1986, specimens/ Mearns and Mearns 1998). In other words, <5% Museum a Species b species Rank c of the world’s ornithological collections are rep- resented by osteological material of any kind— USNM 5,109 10.16 3 single elements to entire complete skeletons. FMNH 3,151 15.65 4 ROM 3,020 14.66 6 Smtwyhtrnslx fsi Bjsjknyx tk Oxyjtqtlnhfq AMNH 4,000 6.25 2 Mfyjwnfq FL 2,889 8.07 16 UMMZ 3,524 6.55 7 There may be many reasons for the discrep- KU 2,769 7.75 14 ancy in specimen preferences, but one aspect LSU 3,175 6.64 11 unrelated to scientifi c use of osteological mate- MVZ 2,148 9.09 10 rial is the relatively high labor cost and delay BM 3,000 5.00 1 associated with preparation. Skilled preparators a See abbreviations in Table 2. Collections are ranked by the can make skin specimens very quickly in the number of species represented by osteological material. b Number of species represented by skeletal specimens. fi eld; thus, for example, Ellio Coues and c Ranking in world based on total number of specimens Henry Henshaw, in the 1880s, competed for the (including skins, skeletons, spirit anatomicals, etc.). quickest preparation on a friendly wager. The July 2005] Commentary 975

Tfgqj 4. Taxonomic coverage by osteological specimens of selected avian orders. See text for explanation. Data from Wood et al. (1982).

Species frequency Species frequency Genus frequency Order (no specimens) a (1–10 specimens) b (no specimens) c Tinamiformes 0.30 0.24 0.67 Procellariformes 0.04 0.23 0.08 Sphenisciformes 0.00 0.06 0.00 Gaviiformes 0.00 0.00 0.00 Podicipidiformes 0.00 0.30 0.00 Pelecaniformes 0.02 0.15 0.17 Ciconiiformes 0.07 0.39 0.16 Falconiformes 0.19 0.45 0.19 Anseriformes 0.01 0.16 0.05 Galliformes 0.19 0.40 0.27 Gruiformes 0.22 0.50 0.30 Charadriiformes 0.05 0.28 0.12 Columbiformes 0.26 0.47 0.48 Psi aformes 0.15 0.42 0.30 Cuculiformes 0.29 0.39 0.35 Strigiformes 0.27 0.46 0.53 Caprimulgiformes 0.40 0.44 0.50 Apodiformes 0.33 0.42 0.46 Coraciiformes 0.20 0.44 0.32 Piciformes 0.18 0.53 0.48 Total 0.19 0.40 0.31 a Frequency of species with no osteological specimens. b Frequency of species with 1–10 osteological specimens, including partial skeletons. c Frequency of genera having species without osteological specimens. winner (Henshaw) completed a study skin of a Bunker at the turn of the last century (Ma hiesen recently collected House Sparrow in one minute 1989, Johnston 1995). A few specialized applica- and thirty-fi ve seconds; Coues took fi ve seconds tions may require bacterial maceration, chemical longer (Cutright and Brodhead 1981). Under treatment, or boiling for cleaning bones; but these normal conditions, Coues felt that four speci- techniques are rarely employed, because of unde- mens an hour was an acceptable rate (Coues sirable eff ects on the bones. The consensus is that 1874); with today’s more rigorous requirements, bone-cleaning with dermestid beetles (or other one specimen an hour is fairly typical (Winker carnivorous invertebrates, like marine crusta- 2000). By contrast, the fastest completion of an ceans) is more effi cient (Ma hiesen 1989, Winker entire skeleton by D.C. took three-and-a-half 2000), but the process is generally held to be noi- days, including about three hours of dedicated some and undesirable (Weed 2003). technician time in preparation; the rest of the AV er the bones are cleaned, the skeleton is time was taken by beetles cleaning the bones usually disarticulated and then soaked in various of extraneous tissue. This additional burden of solutions, depending on its condition—a weak time and personnel costs dissuades most collec- ammonia solution to reduce odor, for example. tors and museums from casually adding osteo- Most importantly, each element is annotated logical material to the collection mix. with the specimen acquisition or register num- The bo lenecks in osteological preparation ber. An experienced preparator can number are bone-cleaning and element-numbering. Most (nearly) every element of a robin-sized bird in skeletal preparators now use dermestid beetles about an hour; smaller birds and larger birds can (Dermestes maculatus) to remove fl esh and connec- take longer, because of small bone size or addi- tive material from the skeleton, a technique fi rst tional preparation time associated with greasier developed at the University of Kansas by Charles bones. Numbering of elements is a critical step 976 Commentary [Auk, Vol. 122 in the process, because if it is not done there is barcodes (Fig. 2). One great limitation is that a great danger of mixing or losing elements in the engraving physically alters the surface of use. Given the more numerous steps in skeletal the bone by removing material through carbon- preparation as compared with skin preparation, ization, which may be objectionable for many there are many opportunities to lose elements, types of research application. to exchange bones with other specimens of the Precision microfi bers (or microtaggants) as same species, or to intermix diff erent species small as 5 µm in diameter carrying up to 107 dif- under preparation at the same time. Despite this, ferent codes can be applied to the external sur- and because of the high demands on personnel face of bony elements through a spray adhesive. in numbering, many osteological specimens in The advantage is that an entire skeleton can be the world’s museums are unnumbered or only marked in a single spray; the grave disadvan- partially numbered (perhaps as high as 25% tage is that the microfi bers must be read using a overall; D. Causey pers. obs.). microscope and decoded. Loss of the codebook Nonetheless, osteological material has many would make this type of system unintelligible— positive aspects. In contrast to study skins, which a shortcoming shared with barcoding and other off er few standard morphological measurements symbolic marking. and are subject to wear, avian skeletons make Microprecision inkjet printers off er a close possible many more quantitative measurements replicate of manual numbering, and as the with a high degree of replication (see Olson 2003 resolution increases (in 2005, 1,200 dpi [dots per for more details). Osteological specimens are inch]), and with computerized control for print- low-maintenance, have high durability, and are ing on curved surfaces like bone shaV s, become much less susceptible to variations in storage increasingly more useful for rapid numbering regime, insect damage, or post-preparation deg- of elements. radation than other specimen types (Ma hiesen All these technological alternatives to manual 1989, Winker 2000, Olson 2003). numbering, and others, have the potential to speed the preparation process of osteological Nj| Dnwjhyntsx ns Czwfynts material, but all suff er the same problem of high cost. Even a relatively cheap precision inkjet Of the two main impediments in preparation printer ($40,000 in 2005) is likely beyond the bud- of osteological specimens, element-numbering gets of most museums, so technological solutions seems the most tractable for improvement. may have to await the creation of a centralized, Several new technological developments off er entrepreneurial specimen-processing facility, promising alternatives to numbering each similar to what has evolved in molecular biology element by hand with pen and ink. Precision for oligonucleotide synthesis and DNA sequenc- laser engraving can be quite quick, about 1 s per ing. Many institutions now outsource that work, element, using numbers, le ers, symbols, even which used to be done in individual laboratories, to central facilities or for-profi t enterprises.

Fzyzwj Rjxjfwhm ns Oxyjtqtlnhfq Ctqqjhyntsx

Osteological collections continue as a resource for current avian research (Fig. 3). Traditional uses are focused on the bony morphology, and examples of recent published research include comparative anatomy and morphology (Ponton et al. 2004, de Margerie et al. 2005), paleobiology (Holdaway et al. 2003, Causey et al. 2005), pale- ontology (Bourdon et al. 2005, Clarke et al. Fnl. 2. Great Currasow (Crax nigra) MCZ 2005), avian systematics (Mayr 2003, Zhou and 340401. Coracoid shaft marked by laser engrav- Zhang 2003), and zooarcheology (Plug et al. ing. Note the traditional pen and ink numbering 2003, Fiori et al. 2004). “401” that was applied over 100 years ago. The Recently, avian osteological material has engraving “MCZ340401” is 8.5 mm long. served as a resource for research far removed July 2005] Commentary 977

Fnl. 3. Published research based primarily on osteological material since 1990. Solid bars: com- parative ecological and evolutionary studies; shaded bars: paleontology; open bars: systematic studies. Data abstracted from citations listed in The Zoological Record. The histogram marked with an asterisk represents citations for only the first quarter of 2005. from the purposes originally assigned to bony Ahpst|qjilrjsyx material. For example, modern material- analysis of avian bones, tendons, and other con- We thank A. Pirie and R. Stymeist for assis- nective tissues has greatly facilitated medical tance in gathering data on the MCZ osteo- and veterinary treatments, as well as provided logical collections. B. C. Livezey and R. L. Zusi new insights into vertebrate evolution (Naldo et amiably off ered comments on the text, as well al. 2000, Summers and Koob 2002, Tully 2002). as skeptical observations on laser engraving Bone has proved to be an excellent source of of bones. J. Hinshaw (University of Michigan DNA, and subfossil bone has been used to Museum of Zoology) kindly provided data on enable molecular study of extinct populations the size and scope of the world’s avian collec- and species of birds (Terbu and Simons 2002). tions. We are grateful to the following cura- It should be pointed out that studies focused on tors and staff for information relating to the the ancient DNA contained in bone oV en use osteological collections of their institutions: J. specimens collected before DNA was known CracraV , G. Graves, J. Hinshaw, B. C. Livezey, to science (i.e. 1860). Osteological specimens S. L. Olson, R. P. Prys-Jones, J. V. Remsen, Jr., D. collected today are just as likely to serve as Willard, K. Winker, and R. L. Zusi. a resource for presently unknown scientifi c technologies 150 years in the future.—Dtzlqfx Lnyjwfyzwj Cnyji Cfzxj~, Department of Biological Sciences, University of Alaska, Anchorage, Alaska 99508, Bfspx, R. C., M. H. Cqjshm, fsi J. C. Bfwqt|. USA (e-mail: [email protected]) and 1973. Bird collections in the United States Jjwjrnfm Twnrgqj, Museum of Comparative and Canada. Auk 90:136–170. Zoology, Harvard University, Cambridge, Btzwits, E., B. Btz~f, fsi M. Ifwthmjsj. 2005. Massachuse s 02138, USA. Earliest African neornithine bird: A new 978 Commentary [Auk, Vol. 122

species of Prophaethontidae (Aves) from the q’Hjwrnsnjw, F. J. 1827. Recherches sur l’appareil Paleocene of Morocco. Journal of Vertebrate sternal des oiseaux. Actes Societe Linneenne Paleontology 25:157–170. Paris 6:3–93. Cfzxj~, D., D. G. Ctwgjyy, C. Ljk•{wj, D. Wjxy, Mfyymnjxjs, D. G. 1989. The curation of avian A. B. Sf{nsjyxp~, N. K. Knxjqj{f, fsi B. H. osteological collections. Pages 71–110 in Kmfxxfst{. 2005. The paleoenvironment of Notes from a Workshop on Bird Specimen humans and marine birds of the Aleutian Preparation (S. L. Rogers and D. S. Wood, Islands: Three millennia of change. Fisheries Eds.). Carnegie Museum of Natural History, Oceanography, no. 14. Pi sburgh, Pennsylvania. Cqfwpj, J. A., C. P. Tfrgzxxn, J. I. Ntwnjlf, G. M. Mf~w, G. 2003. Phylogeny of early Tertiary swiV s Ewnhpxts, fsi R. A. Kjyhmfr. 2005. Defi nitive and hummingbirds (Aves: Apodiformes). fossil evidence for the extant avian radiation Auk 120:145–151. in the Cretaceous. Nature 433:305–308. Mjfwsx, B., fsi R. Mjfwsx. 1998. The Bird Ctzjx, E. 1872. Key to North American Birds. D. Collectors. Academic Press, San Diego, Estes, Boston. California. Ctzjx, E. 1874. Field Ornithology. Compiling Nfqit, J. L., T. A. Bfnqj~, fsi J. H. Sfrtzw. a Manual of Instruction for Procuring, 2000. Radiographic analysis of the growth Preparing, and Preserving Birds and rate of long bones in bustards. Research in a Checklist of North American Birds. Veterinary Science 69:233–240. Naturalists’ Agency, Salem, Massachuse s. Oqxts, S. L. 2003. Development and uses of Czywnlmy, P. R., fsi M. J. Bwtimjfi. 1981. Ellio avian skeleton collections. Bulletin of the Coues: Naturalist and Fronteir Historian. British Ornithological Club 123A:26–34. University of Illinois Press, Urbana. Pqzl, I., P. Mnyhmjqq, fsi G. Bfnqj~. 2003. ij Mfwljwnj, E., S. Sfshmj, J. Czgt, fsi J. Animal remains from Likoaeng, an open-air Cfxyfsjy. 2005. Torsional resistance as river site, and its place in the post-classic a principal component of the structural Wilton of Lesotho and eastern Free State, design of long bones: Comparative mul- South Africa. South African Journal of tivariate evidence in birds. Anatomical Science 99:143–152. Record 282A:49–66. Ptsyts, F., A. Eqfst|xpn, J. Cfxyfsjy, A. Fntwn, I., M. Gfqf, fsi A. Tflqnfhtt. 2004. Cmnsxfr~, E. ij Mfwljwnj, A. J. ij Rnhvqjx, Ecology and subsistence strategies in the fsi J. Czgt. 2004. Variation of the outer eastern Italian Alps during the Middle circumferential layer in the limb bones of Paleolithic. International Journal of birds. Acta Ornithologica 39:137–140. Osteoarchaeology 14:273–286. Szrrjwx, A. P., fsi T. J. Kttg. 2002. The evo- Gttirfs, S. M., fsi S. M. Lfs~ts. 1994. lution of tendon: Morphology and mate- Scientifi c collecting. Conservation Biology rial properties. Comparative Biochemistry 8:314–315. and Physiology. Part A. Molecular and Hfqq, E. R. 1951. Charles Dean Bunker. Integrative Physiology 133A:1159–1170. University of Kansas Museum of Natural Tjwgzyy, S. J., fsi C. Snrtsx. 2002. Gene History, Miscellaneous Publication, no. 3. sequences from New Zealand’s extinct Htqif|f~, R. N., M. D. Jtsjx, fsi N. R. B. Huia. Journal of the Royal Society of New Aymknjqi. 2003. Establishment and extinc- Zealand 32:327–335. tion of a population of South Georgian Tzqq~, T. N., Jw. 2002. Basic avian bone growth Diving Petrels (Pelecanoides georgicus) at and healing. Exotic Animal Practice 5: Mason Bay, Stewart Islands, New Zealand, 22–30. during the late Holocene. Journal of the Wjji, W. S. 2003. The worst jobs in science. Royal Society of New Zealand 33:601–622. Popular Science 2003 (10):44–50. Jtmsxyts, R. F. 1995. Ornithology at the Wnspjw, K. 2000. Obtaining, preserving, and University of Kansas. Pages 95–112 in preparing bird specimens. Journal of Field Contributions to the History of North Ornithology 71:250–297. American Ornithology (W. E. Davis, Jr., Wtti, D. S., fsi G. D. Shmsjqq. 1986. and J. E. Jackson, Eds.). Publications of the Revised World Inventory of Avian Skeletal Nu all Ornithological Club, no. 12. Specimens, 1986. American Ornithologists’ July 2005] Commentary 979

Union and Oklahoma Biological Survey, of birds. In many cases, fi eld and storage prac- Norman, Oklahoma. tices have changed li le since their origin in Wtti, D. S., R. L. Zzxn, fsi M. A. Jjspnsxts. 1982. ornithology in the 1970s and 1980s (Johnson World Inventory of Avian Skeletal Specimens, et al. 1984). Because sampling from genetic 1982. American Ornithologists’ Union and resources is destructive and nonrenewable Oklahoma Biological Survey, Norman, without further collecting, there are a number of Oklahoma. issues regarding loan policies and reciprocation Zmtz, Z., fsi F. Zmfsl. 2003. Anatomy of the that are specifi c to these collections. The fate of primitive bird Sapeornis chaoyangensis from the GRCs is tied, even more intimately than the fate Early Cretaceous of Liaoning, China. Canadian of voucher collections, to the future of fi eld col- Journal of Earth Sciences 40:731–747. lecting; whereas traditional GRCs consisting of frozen tissues must eventually be renewed by continued fi eldwork, current voucher collec- tions will, in principle, remain intact and valu- The Auk 122(3):979–984, 2005 able without any further fi eldwork. Particularly © The American Ornithologists’ Union, 2005. for small to midsize museums with li le inter- Printed in USA. nal funding for the upkeep of GRCs (such as the Burke Museum), it remains a challenge to pro- Future of Avian Genetic Resources vide for the increasing demand on GRCs while Collections: Archives of Evolutionary and at the same time recouping costs for fi eld col- Environmental History.—In the past 30 years, lecting, curation, and storage of tissues. These genetic resources collections (GRCs) have collections and others like them face a unique shiV ed position within ornithology, from a set of challenges: how to balance the activities novel supplement to traditional voucher col- that build, preserve, and promote use of their lections to a major core source of raw material collections with an eye toward maintaining fueling multiple subdisciplines. The demand optimal use for future researchers. for specimens from GRCs now greatly exceeds Genetic resources collections demand li le both the demand for traditional voucher space, but take substantial staff time to organize specimens and, in many cases, the resources and are expensive to maintain. Frozen collec- available to museums to maintain GRCs. The tions need almost constant vigilance even with projection for the next decade is ever-increasing an alarm system installed (Dessauer et al. 1996). use. Here, we present a brief update on modern Because they are newer than traditional col- principles and challenges of collection, storage, lections, they usually represent a small (≤35%) organization, use, and dissemination of genetic overall proportion of specimens, but are none- resources and electronic information associated theless heavily used. Loan activity can become with such collections, drawing heavily on the a large investment for the host institution: for experience of building, loaning, and curating example, in 2003 the Burke Museum loaned the GRC of the Burke Museum at the University subsamples of 5% (1,500 tissues) of its collec- of Washington. The Burke Museum was estab- tion to researchers at other institutions, with a lished under the curatorship of Sievert Rohwer substantial outlay in both staff time and sup- in 1986 and is now the second-largest such col- plies. At the Burke, the upward trend in activity lection for birds in the United States, aV er that has been consistent over the past 10 years and of Louisiana State University. In addition, we shows no sign of diminishing. Because these make a number of recommendations for ensur- loans are to individuals at institutions all over ing the long-term sustainability and value of the world, they indicate a general increase in avian GRCs. demand on tissue collections. Unique challenges for avian genetic resources Field collecting and molecular protocols.—Since collections.—There are now several large their inception, avian GRCs have been used (5,000–60,0000 individual specimens) avian primarily in the arena of systematics, including GRCs in North America, Europe, and Australia, molecular phylogenetics and phylogeography. and many other museums and individuals More recently, common uses have come to have smaller GRCs. These collections typically include conservation genetics and stable-isotope consist of frozen tissues (heart, liver, muscle) analysis, in which chemical signatures derived 980 Commentary [Auk, Vol. 122 from tissues can help determine recent diet or complicated, still represents the gold standard habitat from which the tissue was collected for preservation of avian tissues in the fi eld (see Rocque and Winker 2005). In the past 25 (Engstrom et al. 1999). Storage of tissues in lysis years, the uses of avian GRCs have changed buff er (Seutin et al. 1991) has the advantage of dramatically, from protein, DNA hybridization, not requiring deep freezing and is very eff ective and RFLP (restriction fragment length poly- for isolating high-molecular-weight DNA, but morphism) studies requiring relatively large lysing cells makes isolation of RNA or even of amounts of blood or other tissues to polymerase purifi ed mitochondrial DNA a problem. Some chain reaction (PCR) based DNA sequence and protocols and storage buff ers off er the ability fragment analyses requiring only picogram to preserve RNA for PCR assays (Miller and quantities of DNA (e.g. amplifi ed fragment Lambert 2003). However, even nitrogen storage length polymorphism (AFLP) analysis; Wang will be inadequate for many molecular proto- et al. 2003). Ironically, because of their exquisite cols if the tissues are leV at ambient temperature sensitivity even with degraded DNA templates, for hours aV er the blood sample is obtained or PCR methods have, in our view, contributed to the individual sacrifi ced. Thus, an appropriate the decline of meticulous fi eld collection and goal for GRCs would be to gather a synoptic col- archiving practices, because the threshold of lection of one or several RNA-quality samples quality for PCR methods is oV en lower than for per species. other molecular biological approaches. Tissue Genetic resources collections will undoubt- culture methods have the advantage of provid- edly play a large role in “DNA barcoding,” an ing an unlimited supply of genomic material initiative whose goal is to genetically character- but are labor-intensive to set up and, to our ize many existing museum voucher specimens knowledge, have not been adopted by ornithol- with a short DNA sequence(s) to facilitate future ogists as they have been by mammalogists (e.g. fi eld identifi cation and species discovery. DNA the Zoological Society of San Diego’s Center for barcoding is controversial, not only because it Reproduction of Endangered Species [CRES]). is closely linked with the controversial idea We conducted an informal survey of fi ve of that DNA sequences can form the sole basis for the major avian GRCs in the United States to taxonomy (DNA taxonomy), but also because of determine trends in loan activity and research the many well-known theoretical shortcomings use. Our fi ndings suggest that 60–70% of current of short, single-locus molecular characteriza- loans are for phylogenetic studies (i.e. involving tions of biodiversity for purposes of species one or a few exemplars of diff erent species) and assignment (Moritz and Cicero 2004). We sug- that the vast majority of remaining loans are to gest that curators and users of GRCs scrutinize researchers studying population genetics (i.e. carefully the claims of DNA barcoding and draw many individuals of a single species). Loans a distinction between the theoretical issues sur- for other types of projects (e.g. stable-isotope rounding species designation by DNA and the analysis, studies in basic molecular evolution) potential practical benefi ts to the additional are currently uncommon. Sadly, researchers information provided by DNA sequences. By using techniques such as BAC (bacteria artifi - maintaining a utilitarian view of this contro- cial chromosome) library construction (which versy, genetic resources curators and collections requires very high molecular weight DNA) or stand to leverage substantial resources if DNA microarrays and expressed-sequence-tag (EST) barcoding is conducted on the large scale out- surveys of gene expression (which require intact lined in some schemes (Stoeckle 2003), and few RNA transcripts) cannot make use of most avian would deny that even a single DNA sequence GRCs because the DNA and RNA have not been a ached to a voucher can only increase the stored appropriately. With this in mind, it is information content of that voucher. imperative that the method of preservation, Organization and archiving of genetic resources both in the fi eld and in the GRC itself, maxi- collections.—Most avian GRCs store tissues mize the potential uses of the tissue, especially in cryogenic conditions—either the vapor as specialized techniques in genomics become phase of liquid nitrogen or in electric freez- more taxonomically widespread (Couzin 2002, ers set at around –80°C (Prindini et al. 2002). Edwards et al. 2005). Flash-freezing fresh tis- The major advantage of liquid-nitrogen sue in liquid nitrogen, though logistically systems is that they increase the long-term July 2005] Commentary 981 stability of macromolecules and the breadth especially student helpers who may be unfamil- of uses to which the tissues can eventually be iar with specifi c taxonomies. Organization sys- put. However, they oV en take up more fl oor tems become crucially important as collections space—an important consideration for col- grow in size, complexity, and loan activity, and lections with space limitations. Also, samples even managers may fi nd themselves caring for are sometimes more diffi cult to see and access tissues from organisms outside their area of in liquid-nitrogen freezers, and it is more dif- taxonomic expertise. Because Burke research- fi cult to accommodate samples in nonstandard ers frequently collect generally rather than for containers, which may be a problem for col- a specifi c research project, we found that add- lections with very active loan and acquisi- ing new tissues at the end of a number series tion programs. It is known that archiving in is substantially easier than threading these tis- mechanical freezers maintains materials above sues one by one among those already installed the critical preservation temperature for many in numerous boxes. Numerical organization biomolecules (Franks 1985); this, in conjunction also minimizes the diffi culty of incorporating with a frequent lack of backup freezer space, future taxonomic revisions and, because precise puts many GRCs in jeopardy. Indeed, the past location of a given tube is always known, loan decade has seen the thawing and eventual processing remains rapid. On the other hand, loss of several large and vital avian GRCs. numerical organization can be a hindrance The storage system chosen for GRCs will vary when sampling multiple samples from a single depending on the use and resources available taxon, which may be distributed over several to the collections. For example, freezers are collectors and accessions throughout one or generally less expensive to operate; when the multiple freezers. Burke Museum decided to increase its stor- Traditional and digital vouchers for genetic age capacity for tissues in the late 1990s, we resources collections.—Because of a growing chose increased freezer space over nitrogen, acknowledgment of the importance of voucher primarily because it was cheaper to set up and specimens for molecular research (Winker et maintain. By contrast, the American Museum of al. 1996, Ruedas et al. 2000), a primary goal for Natural History’s Ambrose Monell Collection many collections is to have all or most of their for Molecular and Microbial Research is housed samples vouchered with traditional specimens in an endowed, state-of-the-art storage system (Thomas 1994). However, for frozen-tissue based entirely on nitrogen—maintenance costs repositories, this traditional defi nition of a typically run ~$40,000 per year (R. Desalle pers. voucher can become impractical and—for many comm.). Hopefully, institutions wishing to collection endeavors involving endangered spe- switch to nitrogen storage can convince those cies or in countries where permits to conduct who pay the utility bills for freezers that they destructive sampling are diffi cult to obtain— can at least partly recoup electrical costs by hard to implement. Such nonvouchered samples investing in nitrogen. In either case, tissues are are undeniably valuable, oV en have substantial typically kept in uniform-sized (2 mL) cryovi- associated data, and in most cases are identifi ed als and organized in boxes and racks for easy correctly to species, yet museums are naturally retrieval. We expect that, for tracking and map- reluctant to absorb large numbers because of ping purposes, most large GRCs will comple- space constraints and lack of vouchers. ment traditional hand-wri en vial labels with In some cases, such samples are associated computer-generated labels or bar codes, which with fi eld voice recordings or photographs, or are permanent, easier to standardize, and less both, to increase their reliability. The term “e- susceptible to degradation. voucher,” coined by Monk and Baker (2001), Collections are usually organized taxonomi- applies to such documentation: “An e-voucher cally or numerically (by museum or collector is a digital representation of a specimen…[it] number), and taxonomic organization has been may be ancillary to a classical voucher speci- recommended elsewhere (Dessauer et al. 1996). men or it may be the only representative of The Burke Museum GRC has adopted a numeri- the specimen in the collection.” The goal of cal organization scheme because we felt it per- the collector should be to document the collec- mi ed rapid retrieval of tissues and valued that tion event with all means available. Collection the scheme can be used effi ciently by anyone, events involving multiple levels of vouchering 982 Commentary [Auk, Vol. 122

(e.g. morphological, molecular, digital) will collaboration and communication on advances inspire greater confi dence and permit a broader in tissue collection and preservation protocols, array of scientifi c inquiry by enhancing their along with current best practices associated evidentiary value. with repository management. The Organization Digital access and a global genetic resources for Economic Cooperation and Development‘s network.—Maximal use of biomaterials in con- (OECD) Working Party on Biotechnology is call- temporary research demands sophisticated ing for a global network of biological resource coordination of collection records married to collections to be established (Organization for primary data (molecular biology-based data, Economic Cooperation and Development 2001). digital images, etc.) via electronic and computer The Global Biodiversity Information Facility technology. Future methods in taxonomy need (GBIF) is similarly calling for the establishment to be integrated by a transparent, “virtual” orga- of an international network of biodiversity col- nizational schema that provides unity to taxon- lections with online databases to provide coor- omy and molecular systematics (Godfray 2002). dinated electronic access to their catalogues. Currently, avian GRC databases are heteroge- Conclusions.—Given the diffi culty of pro- neous in structure and organization. However, curing funds for collections-based research, many more museum collections will be coming the oV en greater diffi culty of obtaining the online in the future, and networking them could necessary collecting permits, and, fi nally, the be facilitated by harmonizing vocabularies and concomitant destruction of habitats for birds developing standards early on. Coordination globally, it is not diffi cult to imagine that col- of existing collections and information will lections of organisms made today may well enhance the value and accessibility of collec- be the last opportunity the scientifi c commu- tions (Hoagland 1997, Cambon-Thomsen 2003, nity has to obtain archival material for many Peterson 2005), and awareness of the inventory of the world’s species. Continued eff orts to of tissues available, or lack thereof, may help secure GR samples from all species, both stimulate needed fi eld collecting. Several pre- threatened and common, are justifi ed insofar liminary eff orts for a common digital framework as each specimen represents a unique record for GRCs are in the works, such as an initiative of environmental and evolutionary history from the AOU Commi ee on Bird Collections (Sheldon and Di mann 1997, Sheldon 2001). currently being organized by Carla Cicero (C. Thankfully, a modern paradigm of preserva- Cicero pers. comm.). Modern bioinformatics tion that maintains not only the collecting initiatives will ultimately link tissue-specimen locality and morphological identity of speci- collection records with bibliographic citations, mens, but also the integrity of the biomolecules competing taxonomic determinations, and geo- within them, is generally accepted. Hopefully, spatial referencing information; indeed, some societal acknowledgment of the value of these GRCs, such as those at the Museum of Vertebrate biomolecules will translate into increased sup- Zoology, Berkeley, already have such capabili- port for GRCs and the museums and other ties in place. The ultimate goal is to develop a institutions that maintain them.—Shtyy V. national infrastructure capable of supporting Ei|fwix, Burke Museum of Natural History and research involving genetic resources by promot- Culture, Box 353010, University of Washington, ing the linkage of biological resource collections’ Sea le, Washington 98195, USA (Present online specimen records with the publications address: Department of Ornithology, Museum of and data derived from those specimens. Comparative Zoology, Harvard University, 26 To achieve maximum value, tissue reposi- Oxford Street, Cambridge, Massachuse s 02138, tories need to be networked with one another USA. E-mail: [email protected]); Smfwts and with collections containing voucher speci- Bnwpx, Burke Museum of Natural History and mens (Dessauer et al. 1988). Such digital net- Culture, Box 353010, University of Washington, works for voucher collections, such as ORNIS Sea le, Washington 98195, USA; Rtgg T. (Ornithological Research Network Information Bwzrknjqi, Museum of Natural Science, Louisiana System), promise an exciting future for those col- State University, Baton Rouge, Louisiana 70803, lections. The International Society for Biological USA; and Rtgjwy Hfssjw, Corriel Institute for and Environmental Repositories (ISBER; see Medical Research, 403 Haddon Avenue, Camden, Acknowledgments) provides a forum for such New Jersey 08103, USA. July 2005] Commentary 983

Ahpst|qjilrjsyx Rossman, Eds.). Association of Systematics Collections, Washington, D.C. We thank the following individuals who pro- Jtmsxts, N. K., R. M. Znsp, G. F. Bfwwt|hqtzlm, vided helpful comments, answered questions fsi J. A. Mfwyjs. 1984. Suggested tech- about their collections, or provided comments niques for modern avian systematics. on the manuscript: F. Sheldon, D. Di mann, Wilson Bulletin 96:543–560. C. Cicero, M. Robbins, S. Hacke , N. Rice, M. Mnqqjw, H. C., fsi D. M. Lfrgjwy. 2003. An Peck, M. Braun, P. Sweet, S. Rohwer, R. DeSalle, evaluation of methods of blood preserva- and K. Winker. The International Society for tion for RT-PCR from endangered species. Biological and Environmental Repositories Conservation Genetics 4:651–654. website is at www.isber.org. Mtsp, R. R., fsi R. J. Bfpjw. 2001. e-Vouchers and the use of digital imagery in natural his- Lnyjwfyzwj Cnyji tory collections. Museology 10:1–8. Mtwny, C., fsi C. Cnhjwt. 2004. DNA barcod- Cfrgts-Tmtrxjs, A. 2003. Assessing the impact ing: Promises and pitfalls. PLoS Biology 2: of biobanks. Nature Genetics 34:25–26. 1529–1531. Ctzns, J. 2002. NSF’s ark draws alligators, Owlfsnfynts ktw Ehtstrnh Cttujwfynts fsi algae, and wasps. Science 297:1638–1639. Dj{jqturjsy. 2001. Biological Resource Djxxfzjw H. C., C. J. Ctqj, fsi M. S. Hfksjw. Centers: Underpinning the Future of Life 1996. Collection and storage of frozen tis- Sciences and Biotechnology. [Online.] sues Pages 29–47 in Molecular Systematics Available at www.sourceoecd.org. (D. Hillis, C. Moritz, and B. Mable, Pjyjwxts, A. T., C. Cnhjwt, fsi J. Wnjhtwjp. Eds.). Sinauer Associates, Sunderland, 2005. Free and open access to bird specimen Massachuse s. data: Why? Auk 122:987–990. Djxxfzjw, H. C., M. S. Hfksjw, R. M. Znsp, Pwnsinsn, L., R. Hfssjw, fsi R. DjSfqqj. 2002. fsi C. J. Ctqj. 1988. A national program Obtaining, storing, and archiving specimens to develop, maintain and utilize frozen for molecular genetic research. Pages 176– tissue collections for scientifi c research. 248 in Methods and Tools in Biosciences and Association of Systematics Collections Medicine (MTBM) Techniques in Molecular Newsle er 16:9–10. Systematics and Evolution (R. DeSalle, G. Ei|fwix, S. V., W. B. Jjssnslx, fsi A. M. Giribet and W. Wheeler, Eds.). Birkhauser, Smjiqthp. 2005. Phylogenetics of modern Basel. birds in the era of genomics. Proceedings Rthvzj, D. A., fsi K. Wnspjw. 2005. Use of of the Royal Society of London, Series B: in bird collections in contaminant and stable- press. isotope studies. Auk 122:990–994. Eslxywtr, M. D., R. W. Mzwum~, fsi O. Rzjifx, L. A., J. Sfqffw-Bwf{t, J. W. Dwfltt, Hfiiwfym. 1999. Sampling vertebrate col- fsi T. L. Yfyjx. 2000. The importance of lections for molecular research: Practice being earnest: What, if anything, consti- and policies. Pages 315–330 in Managing the tutes a “specimen examined”? Molecular Modern Herbarium: An Interdisciplinary Phylogenetics and Evolution 17:129–132. Approach (D. Metsger and S. Byers, Eds.). Sjzyns, G., B. N. Wmnyj, fsi P. T. Btfl. 1991. Elton-Wolf, Vancouver, British Columbia. Preservation of avian blood and tissue sam- Fwfspx, F. 1985. Biophysics and Biochemistry at ples for DNA analysis. Canadian Journal of Low Temperatures. Cambridge University Zoology 69:82–90. Press, Cambridge, United Kingdom. Smjqits, F. H. 2001. Molecular collections for Gtikwf~, H. C. J. 2002. Challenges for taxonomy. basic research: Museums, methods, and Nature 417:17–19. morality. Pages 331–346 in Proceedings of Htflqfsi, K. E. 1997. Access to specimens the International Conference on In-situ and and genetic resources: An Association for Ex-situ Biodiversity Conservation in the Systematics Collections position paper. New Millenium (Z. Yaacob, S. Moo-Tan, Pages 317–330 in Global Genetic Resources: and S. Yorath, Eds.). Yayasan Sabah, Kota Access, Ownership, and Intellectual Kinabalu, Sabah. Property Rights (K. E. Hoagland and A. R. Smjqits, F. H., fsi D. L. Dnyyrfss. 1997. The 984 Commentary [Auk, Vol. 122

value of vertebrate tissue collections in applied A specimen in a bioacoustics collection is a and basic science. Pages 151–162 in Global recording of one target animal or group of ani- Genetic Resources: Access, Ownership, and mals and the associated metadata. The sounds Intellectual Property Rights (K. E. Hoagland produced by the animal(s) are usually recorded in and A. R. Rossman, Eds.). Association of one session for a variable length of time (seconds, Systematics Collections, Washington, D.C. more oV en minutes, or even days, as technological Sytjhpqj, M. 2003. Taxonomy, DNA, and the advances improved storage capacity). Specimens bar code of life. BioScience 53:796–797. are obtained on master fi eld recordings that may Tmtrfx, R. H. 1994. Molecules, museums and contain multiple specimens and multiple species vouchers. Trends in Ecology and Evolution from multiple locations. A “label” for an acoustic 9:413–414. specimen, separating it from other specimens on Wfsl, Z., G. E. Hnqq, A. J. Bfpjw, fsi S. V. a master tape (or other media), is the narration Ei|fwix. 2003. Reconciling actual and by the recordist (Kroodsma et al. 1996). In the era inferred population histories in the House of reel-to-reel tape, specimens were cut out of the Finch (Carpodacus mexicanus) by AFLP master tape. Thus, specimens in bioacoustics col- analysis. Evolution 57:2852–2864. lections are termed “recordings” or “cuts.” More Wnspjw, K., M. J. Bwfzs, fsi G. R. Gwf{jx. 1996. recently, especially with the advent of analog cas- Voucher specimens and quality control in se es, cuts were duplicated from the master fi eld avian molecular studies. Ibis 138:345–346. recordings, preserving the integrity of the master fi eld tape. A white leader tape was added to each speci- men obtained from the master fi eld tape. This The Auk 122(3):984–987, 2005 © The American Ornithologists’ Union, 2005. leader served as a visible label onto which was Printed in USA. wri en information about species, location, and date. The specimen was then spliced onto a tape New Directions for Bioacoustics reel containing cuts from the same species. This Collections.—Bioacoustics collections contain species reel organization simplifi ed retrieval of recordings of sounds produced by animals. The specimens and until very recently was the way all technology that made possible the capture of major sound collections maintained their sound ephemeral sound events appeared more than 100 specimens. The three major collections, listed in years ago (Koch 1955). However, for biologists alphabetical order, are (1) Borror Laboratory of who sought to record animal sounds in the fi eld, Bioacoustics (BLB), The Ohio State University technological innovations in truly portable sound (blb.biosci.ohio-state.edu); (2) Macaulay Library equipment and reliable media emerged only aV er (ML), Laboratory of Ornithology, Cornell World War II. Nevertheless, before the introduc- University (birds.cornell.edu/lns); and (3) tion of the portable magnetic tape recorder, pio- National Sound Archive (NSA), Wildlife neers at Cornell University experimented with Division, The British Library (www.bl.uk/nsa). recording sound on motion picture fi lm (Brand Other important collections include (and see 1935). A recording fi eld-trip required a truck- Ke le 1989): Bioacoustics Laboratory and load of equipment, and it took weeks to get the Archive (BLA), Florida State Museum; Center fi lm developed. But there were successes with for Sound Communication, Odense University, this cumbersome technology, including the only Denmark; Sound Library, The Australian known recording of the Ivory-billed Woodpecker National Wildlife Collection; and Library (Campephilus principalis), made in 1935 by the of Wildlife Sounds, Museum of Vertebrate Cornell expedition to Louisiana (Kellogg 1962). Zoology, University of California. Biologists who rediscovered the Ivory-billed Analog magnetic tape, depending on the for- Woodpecker in Arkansas in 2004 were trained mulation, has a life expectancy of 10–40 years to listen for the bird with this recording, and and degrades with each use through magnetic it is crucial to researchers in the Bioacoustics particle loss. Thus, analog tape collections started Research Program at Cornell in evaluating in the late 1940s were recently faced with loss if more than 17,000 hours of automated record- not duplicated. Duplication to new analog tape ings made to detect calling individuals since stock has the same limitations, is labor-intensive, December 2004. and is becoming costly as digital media erodes July 2005] Commentary 985 the market for analog media. Digital duplication digital fi les are then wri en to optical media to fi les on a computer hard drive or transfer to (CD-R at the BLB and NSL, DVD-R at the ML) other digital storage media such as optical disk in the order processed rather than parsing them is also labor-intensive. However, digital storage to species reels, thus reducing labor. possesses many advantages—including lack of A protocol of error-testing each optical disk degradation, rapid streaming to new media, and when it is wri en and sampling the collection random access—that make it a superior archival across its life must be established to measure solution. media integrity. Because access on optical disk Digital technology has initiated a new era in is random rather than sequential, specimen animal sound recording, rivaling in importance retrieval, once the optical disk is located, is far the introduction of magnetic tape recording more rapid than with tape media. Processing of and aff ecting everything from how sound is loan requests is thus simplifi ed, but necessitates obtained to how it is stored, documented, and a system to track a specimen’s location. Thus, examined. Though digital sound recording has a database—to which an optical disk number been available for more than two decades, the and other data are automatically added to each archive community moved cautiously until specimen’s record and from which data retrieval standards for digitizing were established. is simple—is critical to the functioning of a digi- Critical to the accurate digitizing of ani- tal archive (Nelson and Gaunt 1997, Nelson et mal sounds was the availability of computer al. 2001). A byproduct of this systematic transfer hardware capable of sample rate and amplitude of all specimens from analog to digital format resolution (bits) suffi cient to accommodate the has been verifi cation that all specimens exist, full spectral and dynamic ranges of most animal that they are playable, and that the metadata sounds. Archives also require reliable, aff ordable are correct and complete in the database. media that maintain integrity over time. Optical The ability to make recordings accessible over disk media that tested to archival requirements the internet is a major benefi t to storing sound did not become available until the mid-1990s recordings digitally. The fi rst step—giving users (technical reports addressing these and other access to the full catalogue data via searchable issues can be found in Grotke [2004] and on the databases—has been or soon will be accom- BLB website [see Acknowledgments]). Storage plished for the major collections. on optical disk reduces housing space and stor- The second phase of internet accessibil- age requirements, and life expectancy is on the ity is online access to the sound specimens order of 100+ years (judging from accelerated themselves. The BLB site currently has sample aging tests). Redundant backups are made (a sounds wherever a sound spectrogram is working and an archival disk, at minimum), displayed (click on the spectrogram), and all and the original analog tapes are compactly recordings of sparrows, tanagers, and New stored under optimal conditions off site. The World warblers are available for audition (the current digital revolution in sound archives full collection will follow shortly). Streaming should bring about (1) creation of a digital sound capability allows users to place orders archive, (2) development of internet accessibil- for auditioned recordings that meet their needs. ity, and (3) improvement in acoustic analysis Distribution of specimens through our websites tools for research and conservation eff orts. for offl ine use is also available. In creating a digital archive, we must pre- In the past, users had to depend on our staff serve the historical analog collection, strive to make these decisions. With auditioning and to streamline the addition of new specimens, downloading of fi les available, whole new and improve access to specimens. In a digital projects and user bases can be accommodated. archive, specimens are digitized from the ana- Users come with diverse goals, from research- log species reels and new material from master ers wanting sounds for descriptive, mechanistic fi eld recordings of various analog or digital (learning, sound production and transmission), formats on various media. The time-consuming comparative, and evolutionary approaches, to editing process cannot be avoided, but digital educators looking for samples for lectures, dem- data stored to computer fi les are easily accessed onstrations, and student projects. Increasingly, for editing; unwanted segments are deleted and we supply material for exhibits by museums and specimens are copied into their own fi les. The government agencies. We continue to service the 986 Commentary [Auk, Vol. 122 general public, as well as commercial producers Ahpst|qjilrjsyx of fi lms and documentaries, CDs of animal sounds, devices to a ract or repel animals, and The BLB digital project is funded by the so on. Easy access to the sounds will facilitate National Science Foundation (DEB 9613674); research, conservation, and educational eff orts. the Macaulay Library digital eff ort by the This era of major changes to our archives, National Science Foundation (DBI 9977149, supported by funding from the National Science DBI 008441, IBN-03347507, and DUE-0332872), Foundation and the Offi ce of Naval Research the Offi ce of Naval Research (N00014-02-1- (to ML), is clearly enhancing our ability to 0467, N00014-02-1-0620, and N00014-04-01- process accessions and access existing speci- 0663), and the Mellon Foundation. K. Beeman, mens. Sounds are but one component of animal Engineering Design, supported the BLB eff ort, behavior. The ML has developed a sister collec- and the following graduate students pro- tion to the sound collection that includes video vided invaluable assistance; C. L. Bronson, S. images of animal behavior (Dantzker 2004). Burne , C. Capre e, and G. Hough. The ML Too oV en, animal sound recording eff orts, project included signifi cant contributions by S. especially those from published research, fail to Benson-Amram, C. Bloomgarden, B. Clock, R. be deposited in public archives. The professional Grotke, M. Fisher, J. Goetz, W. Hatch, G. Iacino, researcher should know that these recordings are J. Joseph, T. Levatich, M. Medler, M. Moskal, as important as traditional specimens and treat E. Olsen, A. Rahaman, M. Reaves, N. Rice, R. them accordingly (Kroodsma et al. 1996). An Rosen, W. Sandner, L. Serafi n, S. Smith, and archive’s internet site may, in the future, facili- C. Zan. For technical information about archi- tate the transfer of specimens from recordists to val requirements for opitical disk media, see archives by allowing contributors to enter data blb.biosci.ohio-state.edu/technical.htm. for their sounds online, to be used as a database for their collection and to be transferred to the Lnyjwfyzwj Cnyji archive database as specimens are deposited, possibly also over the internet. Other benefi ts Bwfsi, A. R. 1935. A method for the intensive to research should follow. For example, ML and study of bird song. Auk 52:40–52. Totally Hip Technologies have developed a Dfsypjw, M. 2004. Preserving visual record- QUICKTIME browser component that will allow ings at a library of animal behavior. Part any internet client to preview waveforms and 1: From submi ed media to archival mas- spectrograms of any digital sound called up by a ters. RGB DigiNews. [Online.] Available at search. They have also developed annotation soV - www.rlg.org/en/page.php?Page ID=17661& ware that will allow demarcation and retrieval of Printable=1&Article ID=1221. specifi c segments within archive recordings. Gfzsy, S. L. L., fsi D. A. McCfqqzr. 2004. Public sound archives will continue to serve Birdsong and conservation. Pages 341–362 in the public by producing synoptic series on Nature’s Music: The Science of Birdsong (P. animal sounds, an area in which the ML has Marler and H. Slabbekoorn, Eds.). Elsevier excelled, as has the BLA with its ARA record Academic Press, London. series. These are invaluable as aids in training Gwtypj, R. 2004. Digitizing the world’s largest col- students, ecotourists, native people, and others lection of natural sounds: Key factors to con- in identifi cation of animals by voice, especially sider when transferring analog-based audio in areas where visual contact is limited or materials to digital formats. RGB DigiNews. impossible, as in conservation and biodiversity [Online.] Available at www.rlg.org/preserv/ identifi cation eff orts in the tropics (Gaunt and diginews/diginews8-1.html#feature1. McCallum 2004).—Sfsiwf L. L. Gfzsy (e-mail: Kjqqtll, P. P. 1962. Bird-sound studies at [email protected]) fsi Dtzlqfx A. Njqxts, Borror Cornell. Living Bird 1:37–48. Laboratory of Bioacoustics, 1315 Kinnear Road, The Kjyyqj, R. 1989. Major wildlife sound libraries. Ohio State University, Columbus, Ohio 43212, USA; Bioacoustics 2:171–175. and Mfwh S. Dfsypjw, Gwjltw~ F. Bzisj~, Kthm, L. 1955. Memoirs of a Birdman. Scientifi c fsi Jfhp W. Bwfigzw~, Macaulay Library, Cornell Book Club, London. Laboratory of Ornithology, 159 Sapsucker Woods Kwttixrf, D. E., G. F. Bzisj~, R. W. Gwtypj, Road, Ithaca, New York 14850, USA. J. M. E. Vnjqqnfwi, S. L. L. Gfzsy, R. Rfsky, July 2005] Commentary 987

fsi O. D. Vjuwnsyxj{f. 1996. Natural sound resources management, and even public health archives: Guidance for recordists and request also draw insights from avian data (Rappole et for cooperation. Pages 474–486 in Ecology al. 2000). This situation thus calls for an effi cient and Evolution of Acoustic Communication in system serving accurate ornithological informa- Birds (D. E. Kroodsma and E. H. Miller, Eds.). tion broadly, both to meet such varied needs Cornell University Press, Ithaca, New York. and to demonstrate the critical importance of Njqxts, D. A., fsi S. L. L. Gfzsy. 1997. The the resource that underlies them. Borror Laboratory of Bioacoustics (BLB) Presently, such a system does not exist. For and the Bioacoustics Research Group at example, recent eff orts to assemble a list of all The Ohio State University. Bioacoustics 8: specimens of Red Junglefowl (Gallus gallus) in 281–286. natural-history museums in North America Njqxts, D. A., S. L. L. Gfzsy, C. L. Bwtsxts, fsi and Europe took six and a half months of le er- T. J. Kqtym, Jw. 2001. Database design for an writing and e-mailing to result in a list of 752 archive of animal sounds. IEEE Engineering specimens (Peterson and Brisbin 1998). Similarly, in Medicine and Biology 20:76–80. eff orts to assemble large-scale data sets on migra- tory bird breeding and wintering areas, neces- sary for modeling the future distribution of West Nile Virus in North America, were stymied by The Auk 122(3):987–990, 2005 ineffi cient access to information and took many © The American Ornithologists’ Union, 2005. months of eff ort and unnecessary tricks of data Printed in USA. manipulation (Peterson et al. 2003). The technology for such a biodiversity infor- Free and Open Access to Bird Specimen mation system nonetheless exists; it was, in fact, Data: Why?—Ornithology is in a unique posi- developed on the basis of avian data sets, with tion in systematics. Birds are the only major funding from the National Science Foundation. taxon for which more than 99% of species taxa at Subsequently, several eff orts have begun every point on the surface of the Earth are likely assembling such systems across many taxa (see to be known to science (Mayr and Vuilleumier Appendix). Most exciting is that developers of 1983, Peterson 1998). Scientifi c collections of these systems have collaborated to develop a birds document the distribution and diver- next-generation technology that will meld all sity of more than 10,000 species worldwide. these regional eff orts into a single, global bio- Although even these collections are in need of diversity information system—the technology, augmentation and improvement (Remsen 1995, termed “DiGIR” (distributed generic informa- Winker 1996, Peterson et al. 1998), data associ- tion retrieval), has won broad acceptance and ated with existing specimens constitute a rich has been incorporated into many eff orts. source of information about avian distribution Ornithology, with its large quantities of high- and diversity. This resource could serve as the quality information regarding an important basis for many exciting analyses and insights indicator taxon, has the opportunity to lead this into the natural history, ecology, systematics, new world of biodiversity informatics. Several and conservation of birds (Remsen 1995), and other taxonomic communities have already as a guide and motivation for further improve- advanced in integrating their data resources ment of the specimen basis and information via the internet (for examples, see Appendix), resources. and several institutions have already ventured The need for more effi cient access to orni- their ornithological data resources in a proto- thological data, however, is great. Systematic type internet-based distributed system (The eff orts to document and study avian diversity Species Analyst, now superseded by ORNIS). rely on the specimen record as a critical guide. Nevertheless, many computerized ornithologi- Biodiversity conservation eff orts depend heavily cal data sets remain either unavailable over the on avian information, as bird distributions can internet or available, but not integrated with inform conservation planning and prioritization data sets from other institutions. much more completely than other, less well- Free and open access and data value.— known taxa. Numerous other applications in Biodiversity information has traditionally been natural history, biogeography, ecology, natural concentrated in Europe and North America, 988 Commentary [Auk, Vol. 122 even though biodiversity is focused in tropical data gain value. Furthermore, open access to and subtropical regions. This contrast results specimen data results in feedback that leads from the complexities of the history of scien- to higher quality, again increasing the value. tifi c exploration, economics, and educational By contrast, data for which access is restricted and scientifi c opportunities. Like biodiversity do not benefi t to the same extent from analysis, itself, access to information about biodiversity scrutiny, feedback, and interest. is unbalanced. Distributed, not centralized.—A key feature of Modern internet technologies make feasible a the information systems under discussion is system in which information resources can be their distributed nature. Distributed databases accessed by anyone, anywhere on Earth. The may be sca ered across regions and countries, internet provides a medium of information fl ow but are integrated via the internet. This struc- that is limited only by internet access, a barrier ture off ers distinct advantages: (1) data remain that is rapidly disappearing over much of the at the owner institution and are usually not planet. Hence, the regional imbalances that centralized; (2) data served can be updated as characterize the current situation can be largely oV en as desired, keeping information up-to- alleviated. the-minute; (3) data ownership is never in ques- The key point of most debates on the subject of tion; (4) owner institutions can restrict or limit free and open access has been the value of speci- access as desired (e.g. to limit precision of data men data (Graves 2000). Museum curators know regarding distributions of endangered species, that the information associated with the speci- to protect rights of investigators regarding pub- mens they curate is valuable, and for that rea- lication of works in progress, etc.); and (5) the son they have oV en guarded such information collaborative nature of the eff ort is emphasized. carefully—the limited budgets at most collec- Hence, although it required several years of tions, many of which are in serious fi nancial dedicated activity to develop and distribute, situations (e.g. recent problems at the Academy this “architecture”makes the idea of providing of Natural Sciences, Philadelphia), demand that free and open access to information much more any resource be used wisely. Moreover, resources palatable in a number of ways. dedicated to computerization and broad data Value added.—Serving ornithological infor- provision may occur at the expense of specimen mation is not a one-way interaction, not just care and building the collection itself. However, a service to the broader community. Rather, “valuable” data that are not used yield nothing uniting data resources into a single pool to the owners or curators of those data. allows for several ways of adding value to the By contrast, data that are used increase mark- primary data. First and foremost, georeferenc- edly in value. Biodiversity information is too ing locality information becomes much more oV en derived from secondary sources (range feasible—because of the redundant nature of maps, fi eld guides, etc.), which both reduces localities (specimens from single localities scat- data quality and denies credit to those institu- tered across multiple collections, effi ciency of tions that house the primary data (oV en natural- georeferencing work on more densely collected history museums). A system with free and open landscapes), such an eff ort on a collection-by- access to data, however, permits users to access collection basis is very ineffi cient. The success the primary, vouchered information as close to of eff orts for georeferencing mammal specimen its source as possible. Similar to the marketing data (Stein and Wieczorek 2004, Wieczorek et al. strategies of Netscape and Adobe Acrobat, in 2004) is an excellent example. Several additional which providing free and open access is instru- possibilities—use of ecological niche modeling mental in building a market share and making to detect identifi cation errors, standardization a product, such access is key to establishing of taxonomic information, and use of collector natural-history museum collections as the pre- itineraries to detect date–locality errors—are mier source of information about biodiversity. being developed. All these improvements to In this sense, the value of data does not decline, data can be repatriated to the owner institutions but rather increases, as a result of free and open to improve the base quality of their data sets access. That is, as primary ornithological data and information content of the specimens. from specimens become the primary source of Funding potential of community eff orts.—A par- information on the distribution of birds, those ticular advantage of community collaborations is July 2005] Commentary 989 their excellent potential to leverage funding. The not already the primary information resource appeal of funding an eff ort in which all institu- for birds? The answer lies in the diffi cult and tions in a community participate is much greater ineffi cient access that has characterized this than that of funding an initiative that is based at resource. Simply, the data are not used because a single institution. Clear evidence of this poten- they are hard to access. As ornithology provides tial is the success that several taxonomic groups be er and more effi cient access to specimen have had in ge ing funding for community data resources—via ORNIS and related solu- eff orts to integrate data: ichthyology, funded by tions, and their descendents—the user base will the National Science Foundation and the Offi ce grow. Only in this way can avian collections get of Naval Research; mammalogy, funded by the the key recognition and support they deserve National Science Foundation; and herpetology, and need.—A. Tt|sxjsi Pjyjwxts, Natural funded by the National Science Foundation— History Museum and Biodiversity Research Center, summing to more than $4.5 million in new fund- University of Kansas, Lawrence, Kansas 66045, ing for informatics eff orts in scientifi c collections. USA (e-mail: [email protected]) and Cfwqf Cnhjwt These resources would likely not exist without fsi Jtms Wnjhtwjp, Museum of Vertebrate their community basis. Zoology, University of California, Berkeley, ORNIS and the future.—A fully integrated California 94720, USA. ornithological information infrastructure has enormous potential, and has now been Lnyjwfyzwj Cnyji funded by the National Science Foundation. Approximately 4–5 × 106 bird specimens are Gwf{jx, G. R. 2000. Costs and benefi ts of web held in North American museums, and ~80% access to museum data. Trends in Ecology of those specimens have been commi ed to and Evolution 15:374. participation in ORNIS. Perhaps yet another Mf~w, E., fsi F. Vznqqjzrnjw. 1983. New spe- 4 × 106 bird specimens are held in European cies of birds described from 1966 to 1975. museums, and an unknown quantity are held Journal für Ornithologie 124:217–232. in museums elsewhere in the world (2–3 × 106 Pjyjwxts, A. T. 1998. New species and new spe- more?). Hence, a rough estimate is that on cies limits in birds. Auk 115:555–558. the order of 10–12 × 106 bird specimens exist Pjyjwxts, A. T., fsi I. L. Bwnxgns. 1998. Genetic worldwide. If this resource were fully comput- endangerment of wild Red Junglefowl Gallus erized and integrated into a distributed “world gallus? Bird Conservation International 8: museum” of ornithology, the resource would be 387–394. enormously useful in a broad diversity of appli- Pjyjwxts, A. T., A. G. Nf{fwwt-Snl×jsf, fsi cations. Integrating specimen-based data with H. Bjs¤yj-D¤f. 1998. The need for con- observational data is enriching the specimen- tinued scientifi c collecting: A geographic based information still more: a recent addition analysis of Mexican bird specimens. Ibis to the ORNIS network included 15 × 106 obser- 140:288–294. vational records from several projects based at Pjyjwxts, A. T., D. A. Vnjlqfnx, fsi J. the Cornell Laboratory of Ornithology. Asiwjfxjs. 2003. Migratory birds as criti- At present, much information about birds is cal transport vectors for West Nile Virus in drawn from secondary sources. Conservation North America. Vector Borne and Zoonotic organizations prepare secondary information Diseases 3:39–50. resources (lists of endangered species, distribu- Pmnqqnux, A. R. 1986. The Known Birds of North tional summaries, etc.). Field guides synthesize and Middle America. Part I. Privately pub- information into range summaries and distribu- lished, Denver, Colorado. tion maps. Other resources are assembled solely Rfuutqj, J., S. R. Djwwnhpxts, fsi Z. Hzg€qjp. on the basis of observational information, which 2000. Migratory birds and spread of West lacks vouchering and can be unreliable in some Nile virus in the Western Hemisphere. circumstances (Phillips 1986). These secondary Emerging Infectious Diseases 6:319–328. resources are too oV en used as the basis for Rjrxjs, J. V., Jr. 1995. The importance of answering important questions about birds. continued collecting of bird specimens to Why are specimen data—the ultimate “library ornithology and bird conservation. Bird of life” information resource for biodiversity— Conservation International 5:145–180. 990 Commentary [Auk, Vol. 122

Syjns, B., fsi J. Wnjhtwjp. 2004. Mammals both traditional and nontraditional uses of of the World: MANIS as an example of these shared research resources. For example, data integration in a distributed network advances in analytical chemistry have enabled environment. Biodiversity Informatics, no. researchers to obtain data on heavy-metal con- 4. [Online.] Available at jbi.nhm.ku.edu/ taminants and diets from a single feather. Future viewarticle.php?id=11. technological advances will increase nontradi- Wnjhtwjp, J., Q. Gzt, fsi R. J. Hê rfsx. 2004. tional use of specimens, and two areas of rapid The point-radius method for georeferenc- growth at present are in contaminant and stable- ing locality descriptions and calculating isotope studies. We address these developments associated uncertainty. International Journal and their implications for bird collections. of Geographical Information Science 18: Contaminants.—Retrospective contaminant 745–767. studies of the 1960s and 1970s premiered a Wnspjw, K. 1996. The crumbling infrastruc- new and important use of specimens. One of ture of biodiversity: The avian example. the fi rst studies to use bird specimens in con- Conservation Biology 10:703–707. taminant research documented a 10- to 20-fold increase in feather mercury among seed-eaters Auujsin} and raptors aV er the introduction of alkyl- mercury seed dressings (fungicides) in Europe The following are websites for eff orts to in the 1940s (Berg et al. 1966). That research assemble biodiversity information systems: led to the banning of those seed treatments, MaNIS (elib.cs.berkeley.edu/manis); HerpNet and subsequent retrospective analyses using (www.herpnet.org); Global Biodiversity Infor- specimens confi rmed the eff ect of alkyl-mer- mation Facility (www.gbif.net); Red Mundial cury fungicides by documenting the decline para la Información de la Biodiversidad of mercury concentrations in feathers aV er (www.conabio.gob.mx/remib/doctos/remib_ the ban (Westermark et al. 1975). Probably esp.htm); Virtual Australian Herbarium the best-known use of museum specimens in (www.rbgsyd.gov.au/HISCOM/Virtualherb/ retrospective research documented eggshell virtualherbarium.html); SpeciesLink (www.cria. thinning in raptors following the introduction org.br/projetos); European Natural History of DDT in 1947 (Ratcliff e 1967, Hickey and Specimen Information Network (www.nhm.ac.uk/ Anderson 1968). These studies and others (see science/rco/enhsin/). For information on distrib- Kiff 2005) contributed to the eventual ban of uted generic information retrieval (DiGIR), go to DDT in many countries. digir.sourceforge.net/. Researchers have documented high levels For examples of taxanomic data resources on of contaminants in the biota of undeveloped the internet, see www.speciesanalyst.net/fi shnet/ regions, citing the global distribution of pol- (ichthyology); elib.cs.berkeley.edu/manis/ (mam- lutants as the cause (Arctic Monitoring and malogy); www.herpnet.org (herpetology). On Assessment Programme 1998). As global eff orts for georeferencing mammal specimen contaminant burdens increase, spatially and data, see elib.cs.berkeley.edu/manis/. The ORNIS temporally distributed biological samples are website is at ornisnet.org. needed to document changing contaminant levels. Archived avian specimens can document levels of heavy metals, because heavy metals bind to feather keratin at the time of growth The Auk 122(3):990–994, 2005 (Crewther et al. 1965). Archived specimens were © The American Ornithologists’ Union, 2005. used to document increases in mercury pollu- Printed in USA. tion in several avian food webs (Appelquist et al. 1985; Thompson et al. 1992, 1993). Time Use of Bird Collections in Contaminant and series of archived seabirds were also used to Stable-isotope Studies.—Preserved biological document increases in feather mercury concen- specimens are increasingly providing source trations in two avian food webs over the past material for research that is moving beyond 100 years, which were correlated with anthro- traditional questions in collections-based stud- pogenic inputs (Monteiro and Furness 1997, ies. Technological advances are facilitating Thompson et al. 1998). July 2005] Commentary 991

Although most specimen-based retrospective locations, and document diet shiV s. Stable-iso- contaminant analyses have dealt with mercury, tope ratios, like heavy metals, are incorporated all heavy metals can be measured in feathers. into feathers at the time of growth and remain Feathers are useful indicators of elemental body inert, providing a record fi xed in time (Mizutani burdens at the time of growth, because feathers et al. 1990) that enables researchers to monitor provide a route for elimination of contaminants long- and short-term changes in ecosystems (Goede and de Bruin 1984). Contour feathers using avian specimens. seem to have the least variation among feather Feather stable isotopes from archived Atlantic types, allowing comparison among studies Northern Fulmars (Fulmarus glacialis) docu- (Furness et al. 1986). Despite a general lack of mented broad-scale diet shiV s during the 20th information on toxicity thresholds in feathers century, probably a ributable to the whaling (e.g. what concentrations in feathers indicate industry (Thompson et al. 1995). The ability to negative organismal eff ects), feathers from detect ecosystem-scale shiV s in food webs with museum specimens represent a powerful tool stable isotopes is also proving useful in ecosys- for comparing temporal and spatial distribu- tem monitoring. Studies using archived whale tions of heavy metals in the environment. baleen suggested long-term changes in oceanic Birds are useful biomonitors of their envi- primary productivity in the Bering Sea, one of ronments, and they off er an opportunity to the world’s most important fi shing grounds sample at diff erent trophic levels. Contaminant (Schell 2000). This hypothesis is being tested studies generally use tissues and organs not using archived specimens of Bering Sea birds normally preserved by museums. But, with (G. J. Divoky pers. comm.). Similarly, specimen- planning, use of birds as biomonitors can be based isotopic analyses suggest historical coupled with standard museum processing dietary changes in federally listed popula- and preservation to simultaneously achieve tions of the Marbled Murrelet (Brachyramphus very diff erent scientifi c gains. Enhancing marmoratus), providing insight into possible working relationships between museums and reasons for their decline (S. Beissinger pers. contaminants biologists benefi ts both groups, comm.). This type of research is increasing and and this is an important direction of future highlights the value of historical specimens in growth for collections. Increasingly refi ned documenting change. analytical abilities will continue to enhance Stable isotopes such as deuterium, oxygen, the usefulness of museum specimens for con- strontium, and sulfur also show regional varia- taminant studies as new techniques reduce the tion. This variation can be enlisted to address amount of sample required for analyses. Small classic questions of population biology (i.e. amounts of muscle tissue now preserved for spatial and temporal distributions) in highly genetics, for example, may also prove valuable mobile organisms, such as migratory birds. In in future contaminants research. individuals and populations that move among Stable isotopes.—Stable isotopes are increas- isotopically distinct regions, multiple stable-iso- ingly being used in ecology, population biol- tope analyses have the potential to track organ- ogy, and ecosystem monitoring. Isotopic ratios isms throughout their annual cycle, and this is among many naturally occurring elements vary another growing research area (Hobson 1999). geographically and are incorporated into local Presently, these markers do not provide suffi - food chains. Diff erent tissues (e.g. feather, bone, cient resolution to monitor regional movements liver, kidney, and muscle) have diff erent isoto- among habitats or to assess population mixing, pic turnover rates, and the tissues of archived and more work is needed to understand links specimens can be used to provide clues regard- between abiotic and biotic isotopic signatures ing seasonal ecological processes in, for exam- within the systems (geographic and taxonomic) ple, migratory birds. Isotope ratios of carbon being studied. However, research on the physi- are oV en distinct among terrestrial, freshwater, ological processes governing stable-isotope and inshore and pelagic marine food webs, and ratios in consumer tissues (e.g. Gannes et al. nitrogen shows predictable trophic enrichment 1997, Pearson et al. 2003), coupled with local (e.g. Hobson 1999, Kelly 2000). Analyzed in environmental studies, will likely enhance our concert, these widely studied isotopes have understanding of the relationship between been used to delineate food webs, infer foraging biotic isotopic ratios and the environment and 992 Commentary [Auk, Vol. 122 improve the ability to track organism move- compromising the value of the preserved mate- ments. rial. Nondestructive sampling (e.g. feather or The proliferation of stable-isotope research blood collection) is sometimes used by research- has signifi cant implications for specimen use ers because it is seen as saving space, time, and and is an important direction of growth for col- money. However, it does not yield the long- lections. These studies have shown that speci- term scientifi c strengths of preserved whole mens are a valuable resource for understanding animals and therefore is not supported by most populations, diets, and changes over time museums. Broad issues of quality control exist in populations and their environments. This in samples that are small in quantity, destined research requires destructive sampling of small largely for destruction, and unvouchered pieces of specimens, and such use is certain to (Winker et al. 1996, Payne and Sorenson 2003, increase. This increase should be coupled with Smith et al. 2003). For example, using feathers expanded participation in building the resource. plucked during banding for stable-isotope, con- As Winker (2005) noted, the multidimensional taminant, and genetic research (e.g. Smith et al. benefi ts gained through whole-organism sam- 2003) may increase sample sizes for some stud- pling suggest that this is the most eff ective com- ies, but sample destruction makes replication mon ground on which to focus such expansion. problematic. Stable isotopes and contaminants With planning, this approach would also pro- can vary within a single feather and among vide the widest possible array of tissue types for feather types (Furness et al. 1986, Bearhop et contaminant and stable-isotope studies. al. 2002, Dauwe et al. 2003); without vouchers, Value of archived specimens.—Archived analyses may not be verifi able or repeatable. specimens provide important baselines for In short, unvouchered subsamples of birds do comparison with modern counterparts. This not have the high scientifi c value of the modern is especially true when a empting to docu- museum specimen. ment environmental contamination. Previous Preserving specimens for retrospective retrospective studies in birds highlight the research is clearly important and requires a need for, and general lack of, good temporal dynamic vision of how collections will be used series. At present, museums generally do not in the future. Continued technological advances have adequate time series to answer temporal ensure that specimens will continue to produce questions with rigor. Unlike genetic samples, answers to unanticipated questions (Suarez and some isotopic ratios and contaminant concen- Tsutsui 2004). Birds are important monitors trations from the same location can change of ecosystem health. New uses for specimens rapidly and exhibit large variation within and demonstrate an important and growing role among years. Documenting trends and histori- for collections in population and ecosystem cal changes at useful geographic scales with sta- management. It is important that these “new tistical power requires continued sampling and uses” be documented and publicized to make proper archiving. the scientifi c community and public aware of Birds are oV en sacrifi ced in food habit, ener- the increasing user base and the dynamic role getic, physiological, population, and pollution that museums play in conservation and envi- studies. Although these studies provide valu- ronmental sciences. Public, political, and fi nan- able information on avian biology, archiving cial support is necessary if museums are to meet the specimens can provide important long-term their obligation to anticipate “new uses” and data. We strongly encourage researchers to ensure that collections archive material that will deposit sacrifi ced birds into collections and to meet the needs of future research.—Djgtwfm A. off set costs to repositories by providing funds Rthvzj, University of Alaska Museum and the for preservation of this resource (see Winker Department of Biology and Wildlife, 907 Yukon 2005). Archived bioindicators become biomoni- Drive, Fairbanks, Alaska 99775, USA (Present tors that can be used to establish baselines for address: U.S. Fish and Wildlife Service, 1011 East future retrospective research. Tudor Road, Anchorage, Alaska 99503, USA; e-mail: As museums and new partners continue to [email protected]) and Kj{ns Wnspjw, build specimen series, it should be recognized University of Alaska Museum and the Department that time- and space-saving techniques, such as of Biology and Wildlife, 907 Yukon Drive, Fairbanks, preparing fl at skins, can oV en be used without Alaska 99775, USA. July 2005] Commentary 993

Lnyjwfyzwj Cnyji Knkk, L. F. 2005. History, present status, and future prospects of avian eggshell collec- Awhynh Mtsnytwnsl fsi Axxjxxrjsy Pwtlwfrrj tions in North America. Auk 122:994–999. (AMAP). 1998. AMAP Assessment Report: Mnzyfsn, H., M. Fzpzif, Y. Kfg~f fsi E. Arctic Pollution Issues. Arctic Monitoring Wfif. 1990. Carbon isotope ratio of feath- and Assessment Programme, Oslo, Norway. ers reveals feeding behavior of cormorants. Auujqvznxy, H., I. Dwfgfjp, fsi S. Axgnwp. 1985. Auk 107:400–403. Variation in mercury content of guillemot Mtsyjnwt, L. R., fsi R. W. Fzwsjxx. 1997. feathers over 150 years. Marine Pollution Accelerated increase in mercury contami- Bulletin 16:244–248. nation in North Atlantic mesopelagic food Bjfwmtu, S., S. Wfqiwts, S. C. Vtynjw, fsi R. W. chains as indicated by time series of seabird Fzwsjxx. 2002. Factors that infl uence assimi- feathers. Environmental Toxicology and lation rates and fractionation of nitrogen Chemistry 16:2489–2493. and carbon stable isotopes in avian blood Pf~sj, R. B., fsi M. D. Stwjsxts. 2003. Museum and feathers. Physiological and Biochemical collections as sources of genetic data. Bonner Zoology 75:451–458. zoologische Beiträge 51:97–104. Bjwl, W., A. G. Jtmsjqx, B. So¾xywfsi, fsi T. Pjfwxts, S. F., D. J. Lj{j~, C. H. Gwjjsgjwl, Wjxyjwrfwp. 1966. Mercury content in fsi C. M. Mfwynsj ijq Rnt. 2003. Eff ects of feathers of Swedish birds from the past 100 elemental composition on the incorporation years. Oikos 17:71–83. of dietary nitrogen and carbon isotopic Cwj|ymjw, W. G., R. D. B. Fwfxjw, F. G. Ljsst}, signatures in an omnivorous songbird. fsi H. Lnsiqj~. 1965. The chemistry of Oecologia 135:516–523. keratins. Advanced Protein Chemistry 20: Rfyhqnkkj, D. A. 1967. Decrease in eggshell 191–303. weight in certain birds of prey. Nature 215: Dfz|j, T., L. Bjw{tjyx, R. Pns}yjs, R. Bqzxy, 208–210. fsi M. Ejsx. 2003. Variation of heavy metals Shmjqq, D. 2000. Declining carrying capacity in within and among feathers of birds of prey: the Bering Sea: Isotopic evidence from whale Eff ects of molt and external contamination. baleen. Limnology and Oceanography 45: Environmental Pollution 124:429–436. 459–462. Fzwsjxx, R. W., S. J. Mznwmjfi, fsi M. Srnym, T. B., P. P. Mfwwf, M. S. Wjgxyjw, I. Wttigzws. 1986. Using bird feathers to Lt{jyyj, H. L. Gnggx, R. T. Htqrjx, K. A. measure mercury in the environment: Htgxts, fsi S. Rtm|jw. 2003. A call for Relationships between mercury content and feather sampling. Auk 120:218–221. moult. Marine Pollution Bulletin 17:27–30. Szfwj, A.V., fsi N. D. Txzyxzn. 2004. The value Gfssjx, L. Z., D. M. O’Bwnjs, fsi C. M. of museum collections for research and soci- Mfwy¤sj ijq Rnt. 1997. Stable isotopes ety. BioScience 54:66–74. in animal ecology: Assumptions, caveats, Tmtruxts, D. R., P. H. Bjhpjw, fsi R. W. and a call for more laboratory experiments. Fzwsjxx. 1993. Long-term changes in mer- Ecology 78:1271–1276. cury concentration in Herring Gulls Larus Gtjij, A. A., fsi M. ij Bwzns. 1984. The use of argentatus and Common Terns Sterna hirundo bird feather parts as a monitor for metal pol- from the German North Sea coast. Journal of lution. Environmental Pollution 8:281–298. Applied Ecology 30:316–320. Hnhpj~, J. J., fsi D. W. Asijwxts. 1968. Tmtruxts, D. R., R. W. Fzwsjxx, fsi S. A. Chlorinated hydrocarbons and eggshell Lj|nx. 1995. Diets and long-term changes changes in raptorial and fi sh-eating birds. in δ15N and δ13C values in Northern Fulmars Science 162:271–273. Fulmarus glacialis from two north-east Htgxts, K. A. 1999. Tracing origins and migra- Atlantic colonies. Marine Ecology Progress tion of wildlife using stable isotopes: A Series 125:3–11. review. Oecologia 120:314–326. Tmtruxts, D. R., R. W. Fzwsjxx, fsi L. R. Kjqq~, J. F. 2000. Stable isotopes of carbon and Mtsyjnwt. 1998. Seabirds as biomonitors nitrogen in the study of avian and mam- of mercury inputs to epipelagic and meso- malian trophic ecology. Canadian Journal of pelagic marine food chains. Science of the Zoology 78:1–27. Total Environment 213:307–315. 994 Commentary [Auk, Vol. 122

Tmtruxts, D. R., R. W. Fzwsjxx, fsi P. M. males; 5 females) active at some point during Wfqxm. 1992. Historical changes in mercury the period from 1850 to 1970. Egg collecting concentrations in the marine ecosystem of was justifi ed on both scientifi c and recreational the north and north-east Atlantic Ocean as grounds (Grinnell 1906), and many of the great indicated by seabird feathers. Journal of lights of American ornithology, including Ellio Applied Ecology 29:79–84. Coues, Robert Ridgway, and Grinnell himself, Wjxyjwrfwp, T., T. Oixo¾, fsi A. G. Jtmsjqx. collected bird eggs in their early years. T. Gilbert 1975. Mercury content of bird feathers Pearson, a co-founder of the National Audubon before and aV er Swedish ban on alkyl mer- Society, Audubon biologist Alexander Sprunt, cury in agriculture. Ambio 4:87–92. Jr., and even Guy Bradley, the Florida Audubon Wnspjw, K. 2005. Bird collections: Development warden whose shooting death by an egret and use of a scientifi c resource. Auk 122: plumer in 1905 sparked the modern Audubon 966–971. movement, were all egg collectors. Wnspjw, K., M. J. Bwfzs, fsi G. R. Gwf{jx. Serious oologists collected and stored eggs as 1996. Voucher specimens and quality con- entire clutches, or “sets,” beginning in about the trol in avian molecular studies. Ibis 138: 1870s. The contents were removed through a sin- 345–346. gle blowhole in the middle latitudes of the eggs. Thus, a museum “egg specimen” consists only of an empty eggshell with its associated egg- shell membranes. Each specimen was inscribed in permanent black ink with a collector-specifi c The Auk 122(3):994–999, 2005 “set mark,” typically consisting of such essen- © The American Ornithologists’ Union, 2005. tial information as species identity (indicated Printed in USA. by AOU number), collecting year, and number of eggs in the set. Details on collecting locality, History, Present Status, and Future Pros- collecting date, location of the nest, and col- pects of Avian Eggshell Collections in North lector name were recorded on a “data slip,” a America.—Bird egg collecting was formerly card that oV en contained the printed name and a popular pastime in North America, having address of the collector. Oological preparation originated as a cultural import from England and curatorial techniques are discussed more during the great Victorian era of natural history. fully in Kiff (1989b) and Limbert (2003). Although Audubon took a few eggs in the 1830s The fi ndings, mostly anecdotal and descrip- and 1840s, widespread hobbyist egg collecting tive, of oological studies were published in did not really take hold in North America until an astonishing array of small journals, many the 1860s, following the issuance by Stephen short-lived (Underwood 1954). The best of this Fullerton Baird, Secretary of the Smithsonian lot in the late 19th century was the Ornithologist Institution, of a “call to arms” (Baird 1861). His and Oologist, and later important oological circular listed numerous as-yet-undescribed journals included The Warbler, The Nidiologist, eggs needed by the Smithsonian and detailed The Osprey, and The Journal of the Museum of instructions on how to preserve them. The study Comparative Oology, all of which contain solid of eggs, or “oology,” as its adherents termed it, descriptive information still useful to contem- was at its zenith on this continent from about porary ornithologists. The Oologist was the 1885 through the 1920s. Owing to changes in longest-lived journal of the genre, though it social a itudes and regulation, hobbyist egg was not the best. It was published monthly collecting had declined markedly by the start from 1884 to 1941 and was discontinued only of World War II and had completely faded from when the hobby ran out of enough practitio- the American scene by 1970. Thus, the “oologi- ners to keep it going. Other than amassing cal chapter” of North American natural history their collections, the most lasting contribution lasted about a century (Kiff 1989a). of the oologists was the A.C. Bent life histories The vast majority of collectors were adoles- series, which relied heavily on their fi eld obser- cents who took only eggs of the common species vations. Indeed, the egg measurements from in their neighborhoods, but I have compiled bio- the Bent volumes still survive largely intact graphical data for 1,200 adult collectors (1,195 (albeit rounded off to whole millimeters) in July 2005] Commentary 995 the modern egg fi eld guides by Harrison (1978) there are probably no more than 300 major and Baicich and Harrison (1997). bird egg collections, including some in private Present extent of North American egg obscurity, and they contain a disproportionate collections.—Unlike avian study skins, skel- representation of taxa from western Europe, etons, and spirit specimens, egg collecting was North America, southern Africa, and Australia. always primarily an endeavor of amateurs, Uses of collections.—Although most “oolo- and even the largest institutional collections gists” were amateurs, they generally recorded are amalgamations of multiple private collec- useful and reliable data with their sets. Egg tions. A process of consolidation of collections specimens and their associated data have prob- was begun in the late 19th century by several ably been used in a greater variety of biological wealthy collectors, including John Lewis Childs studies than any other type of avian specimens. (Floral Park, New York), J. Parker Norris, Sr. At the WFVZ alone, the egg collection was used and Jr. (Philadelphia), and Edward Arnold in more than 4,000 research projects from 1956 (Ba le Creek, Michigan), and was continued to 1994 (Kiff 2000), and many egg specimen in the mid-20th century by Wilson Hanna (San uses were discussed by Green and Scharlemann Bernardino County Museum, California), Ed (2003) and Limbert (2003). Thus, it is all the Harrison (founder of the Western Foundation more curious that there never seems to have of Vertebrate Zoology [WFVZ] in Los Angeles), been a time when egg collecting was primarily and Nelson Hoy (Media, Pennsylvania). The a scientifi cally oriented activity, despite the pre- WFVZ continued to focus on collections tenses of its main practitioners. acquisition later than all other institutions; by The most traditional lines of specimen-based 1994, its holdings included ~180,000 egg sets, egg studies involve their external morphol- representing the combined assets of more than ogy, including mass, length and breadth, shell 300 separate collections (Kiff 2000). thickness and texture, color, and shape. These Surveys of North American institutions and characters have signifi cance in studies of tax- the few living oologists confi rmed the exis- onomy (Zelenitsky and Modesto 2003), ecol- tence of ~463,000 specimens in 72 collections ogy (Svensson 1978), evolution (Moksnes and (Kiff 1979, Kiff and Hough 1985) by the early RøskaV 1995), physiology (Rahn et al. 1985), 1980s. In the intervening two decades, I have and genetics (Tryjanowski et al. 2001). Almost become aware of only a handful of additional every issue of the major ornithological journals collections, mostly of modest size. I estimate now contains at least one paper on one of these that there are presently around 80 egg collec- egg collection-related topics. tions of research importance in North America Many contemporary lines of research and that, in aggregate, they contain <500,000 involve the use of eggshell fragments, oV en egg sets. Judging from comparisons of collec- in connection with conventional whole egg- tor fi eld catalogues with existing collections, shell collections. The study of eggshell ultra- I have the impression that the majority of the structure by scanning electron microscopic scientifi cally useful egg specimens collected in techniques is a fi eld of interest not only to North America survived the move from private poultry scientists, but also to taxonomists (e.g. closets to institutions. Mikhailov 1997). Bird egg collections are valu- Nearly all the largest egg collections in North able reference tools for archaeologists (Sidell America are located in the largest natural- 1993) who search through Indian middens history museums, as might be expected. By and Grand Canyon caves. Additional uses of now, the trend toward consolidation of collec- eggshell fragments include studies of x-ray tions has slowed. The WFVZ has acquired a diff raction (Gould 1972), pigmentation stud- few minor collections in the past decade, as has ies (Kennedy and Vevers 1976), and isotopic the University of Kansas Museum of Natural analyses (Hobson 1995). Studies of the eff ects History, but orphaned egg collections may now of environmental acidifi cation on passerines, go unclaimed. A large, previously unreported especially in Europe (e.g. Graveland 1998), collection was purportedly sold to a European rely heavily on baseline information provided collector by a private high school, to which it had by egg collections. Using museum collections, been inappropriately donated, and it is probably Green (1998) and Scharlemann (2003) docu- now lost to the research community. Globally, mented declines in eggshell thickness among 996 Commentary [Auk, Vol. 122

Turdus species that probably resulted from Egg collection data have oV en been used to reduced calcium availability caused by envi- document the historical distribution of bird ronmental acidifi cation. species for conservation management and other Egg collections were critically valuable in purposes (Kiff 1989c, Houston 2002). However, documenting widespread eggshell thinning reconstructing entire historical ranges of bird (Ratcliff e 1967, Hickey and Anderson 1968) species from egg specimen data (or any kind caused by DDE, a breakdown metabolite of of specimen data) oV en requires the same sorts the pesticide DDT. By now, hundreds of egg- of extrapolations that plague paleontologists, shell-based studies of this contaminant have owing to the patchy distribution of collectors. appeared in all major regions of the world, and In North America, egg collectors tended to be the present ban on DDT use in all but a handful concentrated in the most populous states and of countries is a direct result of this research. provinces, and large portions of the continent Specimen-based data documenting severe egg- are unrepresented in existing collections. Of the shell thinning among Brown Pelicans (Pelecanus 1,200 egg collectors for whom I have at least occidentalis) and other seabird species and DDE- some data, 146 lived in California, almost all of caused extirpations of Bald Eagles (Haliaeetus them south of the San Francisco Bay area, and leucocephalus) and Peregrine Falcons (Falco only 3 lived in neighboring Nevada. peregrinus) on the California Channel Islands Indeed, egg collections rarely, if ever, pro- (Kiff 1980) provided critical evidence in the vide random samples at any level, and certain “Montrose Case,” a U.S. Department of Justice information from them should be used with suit against the last U.S. manufacturer of DDT. caution. From my personal acquaintance with The case lasted for a decade and culminated in about 30 now-deceased egg collectors, I have a natural-resource damages award of $140.2 concluded that the most obvious collector biases million in 2001. Historically, that episode prob- involved egg size and color selection (the odd ably represents the most important use, from ones were more desirable and are thus overrep- an economic standpoint, of any avian speci- resented in collections), collecting date (the start men type. Studies of the eff ects and extent of of the breeding season is be er represented than other contaminants, particularly heavy metals the end for common species), clutch size (larger (Grandjean 1976), have also involved the use of clutches were considered more desirable), and eggshells and museum egg specimens. Becker parasitized clutches (some collectors thought (2003) recently summarized the many advan- that sets with cowbird eggs were “ruined” tages of bird eggs and other avian specimens as and did not collect them or simply threw out biomonitoring tools. parasite eggs). In addition, many (perhaps 5%) The data associated with eggshell collections of the specimens in North American collections are almost as valuable as the specimens them- are misidentifi ed or represent deliberate frauds. selves, and they may prove to be particularly Perhaps techniques will be devised in the future valuable in climate change studies. For exam- that will allow us to confi rm the identity of ple, Crick and Sparks (1999) recently showed questionable specimens, but it is unlikely that that passerines in the United Kingdom are they will involve DNA, a substance that only initiating egg-laying earlier in the year, prob- poorly prepared eggs contain. ably in response to global-warming trends, and Future prospects and recommendations.—Most egg collection data could be used to detect such of the following suggestions were discussed a trend in North America. The classic study by more fully in Kiff (1978) and are as relevant now Väisänen (1969) showed how egg collections as then: can be used to document changes in the histori- (1) Preserve traditional oological knowledge. cal distribution of a species paralleling changes There has been almost a complete loss of oologi- in climatic conditions. Studies of this type are cal expertise within the museum community, hampered by myopic regulatory a itudes, par- not only fi rsthand familiarity with proper ticularly in the United Kingdom, that have led preparation and curatorial techniques, but to the virtual cessation of scientifi cally based also a loss of knowledge about such factors egg collecting in recent decades, so only a par- as the reliability of individual collectors, col- tial record of broad environmental changes can lector biases, interpretation of set marks, and now be reconstructed from egg collections. species whose eggs were oV en misidentifi ed. July 2005] Commentary 997

Few contemporary collection managers know The two main challenges to such a concept are how to blow an egg properly, or even have the fi nding the requisite funds and persuading tools to do so. As a consequence, almost no egg institutions to contribute their collections to specimens are being preserved, owing to the the cause. Perhaps the most eff ective funding loss of knowledge of proper preparation tech- strategies will emphasize the many applications niques, lack of institutional support, regulatory of egg collections for conservation and biomoni- restrictions, and unfavorable funding trends. toring purposes. To my knowledge, there are no “working oolo- (4) Collect eggs and eggshell fragments for gists,” though the studies of many research- environmental monitoring purposes. Without ers involve egg- or nest-related questions. No continued, well-planned preservation of egg North American egg collection is staff ed by an specimens, many useful opportunities for docu- individual with traditional oological interests menting environmental changes will be squan- or knowledge, and even the WFVZ collection, dered. As discussed above, several exciting lines which contains nearly 40% of all egg specimens of research rely on eggshell fragments, and they in North American institutions (Kiff and Hough do not necessarily involve the use of conventional 1985), has not had a trained biologist directly museum specimens. In the future, egg collecting supervising it for the past decade. Thus, the and egg collections will likely take a diff erent body of traditional oological knowledge may form, and the exquisite preparation techniques vanish, except on the browned pages of extinct of the Victorian era may give way increasingly to journals, and existing egg collections may eggshell fragments stored in vials. gradually become objects of greater interest to (5) Compile a global database of basic data on historians than to biologists. egg size, color, shape, eggshell thickness, and (2) Provide funding for egg collection conser- similar parameters. The most important major vation and growth. Funding prospects for the reference source for egg measurement and support of egg collections remain bleak. There color data for birds of the world remains the are no longer “angels” from the private sector monumental “Handbuch der Oologie,” begun who will invest in egg collections, primarily by Max Schönwe er and brought to fruition by because there are no longer any self-serving Wilhelm Meise (Schönwe er 1960–1992). There motives for doing so. Government agencies, is no equivalent work in the English language, particularly those at local and state levels, are and these types of data are widely dispersed in fi nancially stressed, and museums of all types the literature and fi eld notebooks. The creation have suff ered as a consequence. Federal fund- of a comprehensive database of basic egg data ing can be found for the computerization of egg would be a tremendous time-saver for research- specimen data, but not for their interpretation. ers and, like all such compilations, be helpful in With a few commendable exceptions, tradi- revealing the many gaps in our knowledge and tional university natural-history museums are in existing egg collections.—Lqt~i F. Knkk, 9999 also fading from the scene, and many of the West Star Acres Drive, Star, Idaho 83669, USA. collections now housed at the WFVZ were relin- E-mail: lkiff @aol.com quished by institutions with an ever-waning interest in organismic biology. Lnyjwfyzwj Cnyji (3) Consolidate egg collections. Fred Lohrer suggested to me that a Nature Conservancy- Bfnhnhm, P. J., fsi C. J. O. Hfwwnxts. 1997. A type organization devoted specifi cally to saving Guide to the Nests, Eggs, and Nestlings of and housing natural-history collections might North American Birds, 2nd ed. Academic be formed. This could promote the consolida- Press, San Diego, California. tion of collections, perhaps in regional centers, Bfnwi, S. F. 1861. Circular in reference to collect- and leverage funding opportunities. I think it ing nests and eggs of North American birds. is a good idea in theory, especially in regard Smithsonian Miscellaneous Collections. to consolidation of collections. Most substan- Bjhpjw, P. H. 2003. Biomonitoring with birds. Pages tive research on eggs and other avian specimen 675–736 in Bioindicators and Biomonitors: preparations relies heavily on large sample Principles, Concepts and Applications (B. A. sizes. Therefore, the smaller the collection, the Markert, A. M. Breure, and H. G. Zechmeister, more limited the possibilities for research use. Eds.). Elsevier, Amsterdam. 998 Commentary [Auk, Vol. 122

Cwnhp, H. Q. P., fsi T. Sufwpx. 1999. Climate Symposium (D. M. Power, Ed.). Santa change related to egg-laying trends. Nature Barbara Museum of Natural History, Santa 399:423–424. Barbara, California. Gtzqi, R. W. 1972. Brown Pelican egg- Knkk, L. F. 1989a. Oölogy: From hobby to science. shells: X-ray diff raction studies. Bulletin Living Bird Quarterly 8:8–15. of Environmental Contamination and Knkk, L. F. 1989b. Techniques for preparing bird Toxicology 8:84-88. eggs and nests. Pages 111–117 in Notes from Gwfsiojfs, P. 1976. Possible eff ect of lead on a Workshop on Bird Specimen Preparation egg-shell thickness in kestrels 1874–1974. (S. P. Rogers and D. S. Wood, Eds.). Carnegie Bulletin of Environmental Contamination Museum of Natural History, Pi sburgh, and Toxicology 16:101–106. Pennsylvania. Gwf{jqfsi, J. 1998. Eff ects of acid rain on bird Knkk, L. F. 1989c. Historical breeding records populations. Environmental Reviews 6:41–54. of the Common Merganser in southeastern Gwjjs, R. E. 1998. Long-term decline in the United States. Wilson Bulletin 101:141–143. thickness of eggshells of thrushes, Turdus Knkk, L. F. 2000. A history of the Western spp., in Britain. Proceedings of the Royal Foundation of Vertebrate Zoology, 1956– Society of London, Series B 265:679–684. 1994. Pages 183–228 in Contributions to the Gwjjs, R. E., fsi J. P. W. Shmfwqjrfss. 2003. History of North American Ornithology, Egg and skin collections as a resource for vol. II (W. E. Davis, Jr., and J. A. Jackson, long-term ecological studies. Bulletin of the Eds.). Memoirs of the Nu all Ornithological British Ornithologists’ Club 123A:165–176. Club, no. 13. Gwnssjqq, J. 1906. Is egg collecting justifi able? Knkk, L. F., fsi D. J. Htzlm. 1985. Inventory Condor 8:155–156. of Bird Egg Collections of North America, Hfwwnxts, C. J. O. 1978. A Field Guide to 1985. American Ornithologists’ Union and the Nests, Eggs, and Nestlings of North Oklahoma Biological Survey, Norman, American Birds. Collins, Cleveland, Ohio. Oklahoma. Hnhpj~, J. J., fsi D. W. Asijwxts. 1968. Lnrgjwy, M. 2003. The Uses and Curation of Chlorinated hydrocarbons and eggshell Birds’ Egg Collections: An Examination changes in raptorial and fi sh-eating birds. and Bibliography. Peregrine Books, West Science 162:271–273. Yorkshire, United Kingdom. Htgxts, K. A. 1995. Reconstructing avian diets Mnpmfnqt{, K. E. 1997. Avian eggshells: An using stable-carbon and nitrogen isotope atlas of scanning electron micrographs. analysis of egg components: Pa erns of British Ornithologists’ Club Occasional isotopic fractionation and turnover. Condor Publications, no. 3. 97:752–762. Mtpxsjx, A., fsi E. RÃxpfky. 1995. Egg-morphs Htzxyts, C. S. 2002. Spread and disappearance and host preference in the Common Cuckoo of the Greater Prairie-Chicken, Tympanuchus (Cuculus canorus): An analysis of cuckoo cupido, on the Canadian prairies and adjacent and host eggs from European museum col- areas. Canadian Field-Naturalist 116:1–21. lections. Journal of Zoology (London) 236: Kjssji~, G. T., fsi H. G. Vj{jwx. 1976. A survey 625–648. of avian eggshell pigments. Comparative Rfms, H., P. R. Stymjwqfsi, fsi C. V. Biochemistry and Physiology 44B:11–25. Pflfsjqqn. 1985. Interrelationships between Knkk, L. F. 1978. Report on egg collections egg mass and adult body mass and metabo- in North America. In Final Report of lism among passerine birds. Journal für the Workshop on a National Plan for Ornithologie 126:263–271. Ornithology (J. R. King and W. J. Bock, co- Rfyhqnkkj, D. A. 1967. Decrease in eggshell directors). weight in certain birds of prey. Nature 215: Knkk, L. F. 1979. Bird egg collections in North 208–210. America. Auk 96:746–755. Shmfwqjrfss, J. P. W. 2003. Long-term declines Knkk, L. F. 1980. Historical changes in resi- in eggshell thickness of Dutch thrushes, dent populations of California Islands Turdus spp. Ardea 91:205–211. raptors. Pages 651–673 in The California Shm¾s|jyyjw, M. 1960–1992. Handbuch der Islands: Proceedings of a Multidisciplinary Oologie. Bands 1–4. Akademie-Verlag, Berlin. July 2005] Commentary 999

Snijqq, E. J. 1993. A Methodology for the American Minor Natural History Serials Identifi cation of Archaeological Eggshells. in the University of Michigan Libraries. Museum Applied Science Center for University of Michigan Press, Ann Arbor. Archaeology, University of Pennsylvania, Vƒnxƒsjs, R. A. 1969. Evolution of the Ringed Philadelphia. Plover (Charadrius hiaticula L.) during the S{jsxxts, B. W. 1978. Clutch dimensions last hundred years in Europe. A new com- and aspects of the breeding strategy of puter method based on egg dimensions. the Chaffi nch Fringilla coelebs in northern Annales Academia Scientifi ca Fennica, Europe: A study based on egg collections. Series A IV Biologica, no. 149. Ornis Scandinavica 9:66–83. Zjqjsnyxp~, D. K., fsi S. P. Mtijxyt. 2003. New Tw~ofst|xpn, P. L. Kzh~¶xpn, M. Asyhfp, M. information on the eggshell of ratites (Aves) Sptwfhpn, fsi M. Hwtrfif. 2001. Within- and its phylogenetic implications. Canadian clutch repeatability of egg dimensions in the Journal of Zoology 81:962–970. Jackdaw Corvus monedula: A study based on a museum collection. Biologia, Bratislava 56: 211–215. Received 30 June 2004, accepted 28 April 2005. Usijw|tti, M. H. 1954. Bibliography of North Associate Editor: R. M. Zink