Open Access
How to translate text using browser tools
1 July 2005 Bird Collections: Development and Use of a Scientific Resource
Kevin Winker
Author Affiliations +

Bird collections were founded and built during the heyday of global exploration. The mission of these collections through most of their history has been to document avian diversity and its distribution and to serve as a resource for research and education. As bird collections became established and grew, ornithology itself became a scientific discipline and broadly expanded its purview. Today, there are more professional ornithologists than at any time in history, and collections-related research represents only a small portion of the discipline. This is healthy. Collections are but one means through which we study birds. But we cannot be lulled into a view that the day of the collection is past—a decided risk when fewer ornithologists have direct experience either with collections or with the multidimensional strengths that a collections-based approach brings to science.

Too little attention has been paid to the role of bird collections in science. This role is changing. Because we understand avian diversity better than that of most other classes of organisms, the central goal for establishing bird collections would seem to be largely accomplished. As anyone studying avian diversity knows, however, much remains to be done: systematics, patterns and processes of differentiation, and geographic variation in birds remain vigorous areas of new learning. But while those of us close to collections remain connected to these important questions, to others what we do is increasingly arcane. After all (tongue in cheek), isn't there already a field guide? To an undiscriminating public, much of this work would seem to be done once a field guide appears. And, indeed, explorations at that level in most regions are largely complete. But in other respects the scientific strengths of collections-based research are blossoming—with no end in sight to the fruits that can be borne.

Bird collections are probably the strongest and most dependable shared resource in ornithology. Biological collections in general represent the original “big science” expenditures in the life sciences, antedating by centuries large contemporary endeavors such as genome projects. Long-term investments in the development and maintenance of collections have produced a resource equivalent in many respects to the mega-science facilities found in other disciplines (e.g. large telescopes or supercolliders). A key difference is that this resource has a useful lifespan that greatly exceeds that of mechanical facilities. Bird collections need to be viewed as a highly versatile and indispensable resource integral to the continued successful (and economical) pursuit of a wide range of subjects. Importantly, these subjects are no longer restricted to ornithology.

Unlike other, contemporary “big science” projects, biological collections establish an object legacy—continuing sources of data that are repeatedly tapped to provide answers to questions about birds and environmental conditions. Many of these questions were not even imagined by those who have built these collections. Indeed, as the ways in which museum specimens are used multiplies with conceptual, technological, and environmental changes, it is clear that we need to refocus on how best to continue developing this resource.

Specimens and Science through Time

Mismatches in temporal relevance.—

The classic analogy that natural history collections are like libraries invites direct comparisons between specimens and books. There is some material similarity to these objects; old books and bird skins are products of animal skin and plant materials. With care, books can easily last for many centuries, and a useful lifespan of more than a millennium is likely. Effective preservation of bird skins has been practiced for only about two centuries, but the skins will clearly last far longer. Beyond simple object comparisons, consider use: it is routine in collections-based work to use specimens 100 years old or older, and historical specimens are increasingly used to conduct retrospective studies—research asking questions about changes in birds and the environments they live(d) in. As time passes, preserved specimens increase in scientific value. By contrast, the value of journal articles and books we produce today will be relatively short-lived (see ISI Journal Citation Reports;; after only a few decades most of our papers will no longer directly contribute to science. Thus, in terms of utility and relevance, these two products a researcher can leave behind—specimens and publications—have strikingly different trajectories. Publications are important. But the longterm scientific value of specimens is widely underappreciated, and we probably place too much emphasis on producing publications, the more ephemeral of the two products that we might leave as our scientific legacy.

“Biological filter paper.”—

Just as historical anthropogenic objects like books reflect changes in human history, so too can preserved specimens enable us to appreciate and measure changes. One of the most important developments in collections-based science is the retrospective study, a very powerful approach for assessing changes in populations and environments. Specimens document life in three dimensions: geographic space (locality), biodiversity space (taxonomy), and time (date). The last dimension is becoming increasingly important, because historical samples enable us to enlist this strong analytic approach to measure and understand change. Probably the broadest reward this science brings to society as a whole is through the increasing use of specimens as “biological filter paper,” documenting “experiments” in the environments in which these animals lived. In many studies, a species or genus is chosen because it represents an important trophic level at which to measure bio-accumulation or magnification of contaminants. These measurements have profound implications for humans and the natural resources we manage, and these studies repeatedly demonstrate that historical samples are crucial. Birds are excellent bioindicators of environmental conditions, and bird specimens should continue to be at the forefront of this field. Continued acquisition of new specimens should be seen as a priority. The time dimension can be studied only with continued sampling of the avifauna.

“Arrogance of the present.”—

The scientific capabilities and accomplishments of today are truly impressive when compared with those of yesteryear. But this will be equally true in the future, when the accomplishments of today will be over-shadowed by those of our successors. Thus, to better serve our science, we should consider how we might contribute to a future 50 or 100 years from now. What do we have that they will not? A key asset of the present is access to a biota that is still probably half intact. There is strong evidence that passing samples of this biota forward to future researchers is one of the most effective ways to contribute to the accomplishments of future science, and that archiving specimens will enhance the effectiveness of future wildlife management and conservation (e.g. through the growing importance of retrospective studies).

Going Forward

Collections are helping to answer widely important questions about birds and our shared environments, and meeting the needs of users is a central purpose of a collection. As the user community grows, so too should support and participation in continued collections development. And the strengths of collections must be considered broadly, separately from individual research programs.

With increasing frequency, specimen loan requests ask different questions of preserved material than the questions for which that material was originally preserved (e.g. feather plucking of skins for genetic, isotopic, or contaminant analyses, and disease screening of genetic samples). In fact, one cannot predict what question, specimen type, or taxon the next loan request will hit upon. This is a double-edged sword: it reassures us that collections are broadly useful, but it suggests that collections growth is becoming increasingly out of touch with collections-based science. As the array of possible uses has increased, our ability to foresee what the specimen needs of tomorrow will be has declined. There is little doubt that there will be need, however. Important questions about changes in populations and environments will expand the need for specimens, and adequate sample sizes from today will be required.

So it is clear that the resource itself must continue to be developed. Collections were built on general acquisition policies, and their broad usefulness today reflects this. Continued growth should follow this course (in a guided manner), and care must be taken to foster this growth without exclusive reliance on focused research programs. Individual research programs are important components of collections growth, but healthy growth requires a broader diet. Wider user needs are not likely to be met by a focus on the scientific questions of today, which, although important at present, tend to be either taxonomically narrow or broad but shallow in sample size. Even the best-focused research of today will not meet tomorrow's specimen needs—unless, in aggregate, we work to increase that likelihood. Thus, most collections-based biologists and curators advocate a general acquisitions policy for our shared collection resources. But it is important to note that we recognize costs and commitments and that we work to maximize the gains that each archived specimen represents. Most effort goes to fill gaps (taxonomic, geographic, and temporal) and to increase sample sizes to enhance statistical power. The challenge is that there are a lot of gaps, especially in temporal and numerical dimensions.

The preserved objects themselves, as continuing sources of new information, have primacy over associated data. Recognizing that diverse, long-term scientific gains are achievable, more components of individual birds are being preserved, such as skin, partial skeleton, tissues, and stomach samples. Given the surprises that technology and science have derived from bird specimens thus far, it is not too outlandish to suppose that interdisciplinary teams (e.g. ornithologists, entomologists, parasitologists, virologists, isotope ecologists, computational and systems biologists, and community geneticists) will one day delve into a treasure trove of preserved avian stomach and tissue samples to extract complex network analyses of environments, communities, and biospheres. The fact that modern bird specimens are preserving so many attributes of today's biota speaks to the strengths of whole-organism sampling and preservation. Ornithology has yet to develop a more economical or efficient method than whole-organism sampling to accomplish such important, multidimensional documentation. This must be more widely understood.

Fashion trends are trumping strong science.—

The fashionable concept that the only good science is science that tests hypotheses has in some respects been detrimental to collections. Simple exploration, description, and comparison—the types of science upon which collections were established—are today “poor cousins” in terms of recognition, funding, and publication. This has occurred despite the obvious importance of collections in establishing the baselines of our understanding of biodiversity and its distribution in space and time. Specimens and collections can, with planning, continue to establish and provide the baselines from which hypotheses are developed and tested, in dimensions that are both traditional (e.g. biodiversity) and nontraditional (e.g. environments, contaminants, and diseases). Our science will be strongest when we recognize and support these multiple legitimate pathways and contributions to how knowledge is developed. Hypothesis testing is an important component of individual research programs. But hypothesis testing alone, except in the most general manner (e.g. “things will change”), is an inadequate basis for continued collections growth.

An altogether different phenomenon has emerged around humans killing birds for science. Fastidiousness in this regard wildly outstrips our responses to other, much greater sources of avian mortality, and collecting for science is singled out for astonishing levels of restriction and scrutiny. The contortions through which many permitting agencies are willing to go to squeeze scientific sampling (nearly out of existence in many cases) is truly amazing. Most of this regulation is not biologically defensible, and it is not conservation. Permitting is useful and necessary, but it often seems to be a political means of imposing belief systems on others, affecting government and wildlife management to the detriment of science and society. It is ironic that the very agencies that would benefit most from the knowledge that specimens would deliver to their management programs are so often hostile to this knowledge development. This is self-handicapping behavior.

Individual researchers also need to examine how their beliefs affect their work and their scientific commitment. The act of killing a bird for science is difficult for those who do it and seemingly a personal barrier for those who choose not to. Those not directly involved often have strong beliefs surrounding this act. Many have not sorted out their feelings and beliefs about scientific collecting, nor have they thought through the long-term consequences of their choices to either support or not support this activity. Two beliefs are often involved: belief in the sanctity of the life of an individual bird, and belief in an erroneous worldview of conservation in which every individual matters. I do not wish to demonize these beliefs, but it is wrong to impose them on others. Too often, permitting systems are being used to do so, which is detrimental to research and management. Equally problematic are choices by researchers that cause these belief systems to diminish the scientific effect of their own work. Too many workers go out of their way to avoid collecting birds, even if their studies would be better for it. Usually, science in general would be better served if they collected as part of their efforts. I long ago integrated collecting with banding, because I realized that I was releasing most of the data I was working so hard to obtain.

Except when noncollecting is necessary, owing to factors such as small, fragile populations or the requirements of a study to follow living birds, specimens should be an expected product of most field ornithology. By not delivering fully on the scientific promise of an effort, time and money are often inefficiently used and thus—in part—wasted. This can be viewed as misallocation of scarce resources. Some will argue that time does not allow them to both collect and accomplish their goals. If those goals are strong science, some change is warranted. Even banders should be preserving all accidental casualties. Researchers should understand that by choosing not to collect (or, if killing birds violates a personal belief, choosing not to preserve salvaged specimens), they are diminishing their own scientific legacy. One can retain a strong respect for life and conduct good science that includes preservation of specimens.

Everyone working with birds should consider the details of these issues. Conservation must focus on populations; every individual dies. In most populations, every individual does not matter. Study population biology. Do the math. Recognize that bird populations are a renewable resource, and that scientific collecting represents a practically insignificant (and non-additive) proportion of annual avian mortality. In turn, specimens provide multiple benefits to science, wildlife management, conservation, and society. The vast majority of bird populations can easily support the small amounts of collecting that will maximize scientific gain for the resources expended in avian research.

Specimen salvage.—

Many birds are killed inadvertently by humans (e.g. in collisions with towers, windows, or vehicles, or by pet cats) or die naturally. These are potential specimen “salvages,” and this source is greatly underutilized. On average, however, salvaged specimens are of less value than specimens actively collected for science. Concentrations in geographic and taxonomic space limit the effectiveness of salvage, as does the often mutilated or rotting condition of salvaged birds. An active salvage program can rapidly fill local gaps and become saturated in the areas (geographic and taxonomic) represented in the salvage stream. Thus, museum interest in salvage usually comes below the interest in actively collected specimens, because the latter have been taken with science specifically in mind and are thus more useful and in better condition. But salvage has value, and in one respect this value is underexploited. The relative concentrations of salvaged birds in geographic and taxonomic space represent an opportunity to obtain birds from areas not usually collected (e.g. cities and parks), to archive rare species from captivity, and to develop large sample sizes of some species. These opportunities seem to be rarely exploited, perhaps because the questions these samples can be used to address tend not to be in the realm of traditional museum studies. However, their value for biomonitoring, for example, is high, and this is an area where agencies, governments, and individuals unwilling to kill birds should be actively developing partnerships with museums.

Monitoring and surveillance.—

Birds play a prominent role in environmental monitoring, yet we often lack good baseline data against which to measure change. Too much avian monitoring and surveillance involves only counting animals. This is like taking health and disease statistics without addressing causation. Collecting, preparing, and archiving bird specimens through time is an economical way to enable implementation of retrospective studies when perceived changes occur, allowing detection of correlative changes that may have happened in such things as contaminants, food, and habitat use (e.g. through stable-isotope analyses), diseases, parasites, genetic diversity, sex and age structure, and traditional phenotypic parameters. This approach is effective for the monitoring and surveillance of populations, species, and environmental change. And the utility of these materials is not directly correlated with time. The other two dimensions that collections document (positions in biodiversity and geographic spaces) can produce important baseline information almost immediately. Specimens do not need to “ripen” for years to be useful. As a discipline, we have not effectively planned how best to do this job of using science for effective environmental stewardship. Nonlethal field work, counting, monitoring, and surveillance all remain important, but it is imperative to couple these approaches with a sample-based component if we are to maximize our success in bird conservation and management.

Leveraged research and partnerships.—

Collecting and sending specimens or specimen loans, especially to people at other institutions and in other countries, indirectly leverages research support (i.e. time and money) for the species being collected. Thus, through specimens, a management agency, an institution, a state, or a country can often get work done that they themselves could not do or could not afford to do, providing increased knowledge about that resource. (And I emphasize the importance of basing such efforts on vouchered specimens.) Indeed, it is surprising that more resource management agencies and their permitting personnel do not recognize this important and effective means of inexpensively learning about the resources they manage. The most effective managers will be those with the best science to apply to their management plans; courting the appropriate researchers with specimens and samples is highly effective in developing such partnerships.

Two other important partnerships are those between institutions (e.g. museum and state or federal agencies) and those between researchers and a repository. Agencies responsible for resource management are increasingly dissociated from the sample-based perspective of museums, a situation that is harming us all. With our shared goals of understanding and successfully managing and conserving wildlife, we are natural partners, and we need to bridge this divide and begin working more effectively together to obtain and archive the specimen resources that will enable the very best science and scientific management. The best “bridge” is general and specimen-based: changes will occur in many dimensions (many of them unpredictable), and adequate numbers of specimens can be highly effective in documenting these changes, for both contemporary and retrospective studies.

Individual researchers also have much to gain by partnering with a repository. Properly archiving your specimens and samples for perpetuity has widely recognized scientific benefits. Such partnerships are best arranged before writing proposals for funding and permits (and certainly before initiating fieldwork). Proposals are strengthened by an archival component, and most funding agencies support the added costs that ensure proper preparation and preservation. Those costs are generally a small portion of the total, and writing them into proposals, with input from the repository, is now routine. Reviewers are increasingly (and properly) expecting to see this. In their turn, repositories need to grant researchers specific rights to research priority, with a sunset clause (e.g. rights of refusal on potentially competing user requests being guaranteed for five years or as long as the researcher is conducting active research on the material). These partnerships are among the easiest to generate, because the immediate gains are apparent to all. Moreover, this approach ensures that important research material and associated data are not lost to science because of a local freezer failure or a lab or office cleaning.

Supporting the resource.—

Archiving specimens costs money and time, and museums cannot carry the burden alone. For example, just sending dead birds to a museum does not get the job done. Preparation capacities at museums are always saturated, and this is a bottleneck. Enhancing preparation capacity would have immediate benefits, and a broader distribution of preparation activities would work to the strengths of museums as repositories. This is a community resource, and it is time to develop community-wide solutions to collections development and maintenance. These solutions should be local, regional, and national, but a key basis is that those who are users or beneficiaries of specimens and related data need to become supporters and contributors.


Bird collections are a community research resource and provide broad benefits to our science, to the management and conservation of birds, and to society. In addition to strong contributions in traditional research, collections are making important, long-term contributions to issues that have little to do with the reasons for their establishment. These contributions are often more important to society than the original reason(s) for making the collections, and this needs to become part of the planning and reward processes for continued collections growth. Presently, most growth is focused on short-term gains; yet, as products of science, the specimens themselves have a much longer useful life than the publications generated from them. We recognize this, in part, simply by maintaining existing collections. The next step is to make new investments to enhance future gains. There is clear indication today of a need to collect, prepare, and archive specimens, and to do so in a way that increases the array of preserved components (i.e. animal parts), sample sizes, and dimensions (in biodiversity, geographic, and temporal spaces) available to present and future researchers. A lot of the world's biodiversity will ride the conservation coattails of successful avian management and conservation. As ornithologists, we have the opportunity and obligation to lead in this area. Specimens have been and should continue to be an integral part of the science behind monitoring, managing, and conserving our biological resources. Together we can direct collections growth to establish the baselines that we know will enhance our effectiveness in sound environmental stewardship.


I thank D. Causey, J. A. Cook, G. R. Graves, and J. V. Remsen, Jr., for discussions and T. Braile, D. D. Gibson, R. A. Z. Meier, R. M. Zink, and an anonymous reviewer for comments.


Kevin Winker "Bird Collections: Development and Use of a Scientific Resource," The Auk 122(3), 966-971, (1 July 2005).[0966:BCDAUO]2.0.CO;2
Received: 30 June 2004; Accepted: 28 April 2005; Published: 1 July 2005
Back to Top