Key points are not available for this paper at this time.
Herbaria contain an estimated 480 million specimens worldwide accumulated through the efforts of thousands of botanists for over four centuries and counting (Thiers, 2017). These institutions were primarily established to serve as (1) a central resource for taxonomic and floristic studies, including species discovery and description; (2) a comprehensive archive to document biodiversity for medicinal, economic, or scientific purposes; (3) a means to verify species identification; and (4) an accessible repository to formally deposit vouchers to ensure botanical research is both verifiable and repeatable. These conventional functions of herbaria remain essential to botanical research: new species continue to be described, systematic relationships are always subject to revision, nomenclature is routinely updated, and the documentation and study of biodiversity is increasingly important as we enter an anthropogenic era of mass extinction. However, emerging research also utilizes specimens in nontraditional, unanticipated ways (Fig. 1). This trend does not overshadow or invalidate past uses; rather, the utility of herbaria is expanding to include functions the original founders and collectors never imagined. But exploiting specimens for these unanticipated functions introduces the need for new methodological considerations. To advance collections-based science into this new era, we introduce the analogy of herbarium specimens as “exaptations”—that is, the current use of collections reaches far beyond their originally anticipated uses. The emergence of novel research uses for herbarium specimens in the last decade as illustrated by the number of relevant publications in the American Journal of Botany. These studies used specimens themselves (i.e., not just specimen label data for species distribution modeling) either for genetic analysis (e.g., systematics) or for quantification of intraspecific trait variation along taxonomic, temporal, or spatial gradients (often in context of global change issues). Note that none of these papers use herbarium specimens to integrate genotypic and phenotypic changes together. Gould and Vrba (1982) introduced exaptation to the evolutionary terminology to distinguish a biological trait that evolved through selection in the context of its present role (adaptation) from one that initially arose through natural selection in the context of a specific function (or had no function) but was later co-opted to serve a new role that increased fitness (exaptation). Feathers are a classic example of an exaptation, evolved first in non-avian dinosaurs and functioned in heat regulation or display and only later co-opted for flight (Gould and Vrba, 1982). Exaptations are thought to be prevalent across the tree of life at all levels of biological organization. However, because selection pressures are inherently multidimensional, inferring a trait's prior function with certainty is difficult. Therefore, the term was not widely adopted by evolutionary biologists. Ironically, the term itself has been exapted by other fields, especially in the context of human technological innovation (Larson et al., 2013). The CD-ROM is one example of a technological exaptation: developed specifically to replace vinyl records and only later was the technology applied for computer data storage (Dew et al., 2004). Here, we suggest that herbarium specimens can be viewed as exaptations. We use this analogy in the broad sense of economists and historians (Dew et al., 2004) who adopted the term from evolutionary biology to more loosely refer to any technological feature, object, or process that has been co-opted for a present role from a previously designed purpose. Recent studies leverage decades of herbarium specimens from across the globe in unanticipated ways to test hypotheses in fields such as ecophysiology, environmental chemistry, molecular systematics, population genetics, and global change biology (Table 1). These emergent uses fall well outside of what most collectors or the founders of herbaria specifically intended. Several pioneering botanists were aware of the potential of herbarium material for nontraditional uses (e.g., Merrill, 1916; Fosberg, 1946), but these perspectives were far from conventional through most of the history of plant collecting. Studies increasingly rely on herbarium specimens to answer otherwise seemingly impossible questions at vast temporal and spatial scales (Appendix S1, see the Supplemental Data with this article). For example, herbarium specimens have been used to test the prevalence of rapid evolution in introduced species (Buswell et al., 2011), quantify long-term phenological shifts as a result of climate change (Davis et al., 2015), track the evolution of secondary metabolite production in response to herbivory (Zangerl and Berenbaum, 2005), determine the phylogeography of the blight strain responsible for the Irish potato famine (Saville et al., 2016), and map the distribution of plant disease (Hood et al., 2010). Given that opportunistic studies test hypotheses for which the underlying data were not designed, the recognition of herbarium specimens as exaptations carries several implications. Progress has been made to validate opportunistic specimen uses (Appendix S1), but further statistical and curatorial considerations are needed. Most biological collections far predate most of our collective molecular, analytical, and computational knowledge. Of the top 100 largest herbaria in the world, 98 were established long before the structure of DNA was even published (Thiers, 2017). Collectors or curators a century ago did not consider the possibility of extracting DNA or measuring stable isotopes. As such, these and other unanticipated applications for herbarium specimens require the validation of various protocols specific to any post hoc uses, including the careful consideration of inherent biases in biological collections. Collection biases include pervasive taxonomic, phylogenetic, and temporal biases in sampling effort (Prather et al., 2004), as well as spatial biases (e.g., near roads). Further, as specimens are traditionally collected with identification uses in mind, collectors tend to focus on specific characters. For example, it is standard protocol to collect reproductive specimens that exhibit “typical” morphologies, with no herbivory damage or otherwise “distracting” features. Meanwhile, peculiar specimens are oversampled. Unknown motivations behind collections can blur inference in ecological or evolutionary studies that might assume a specimen characterizes a population. Similarly, the processes of drying, mounting, and storage can result in sample degradation or contamination that further bias the representation of species’ characteristics. Knowledge of bias cannot retroactively change collecting and preservation practices, but current practices can be carefully evaluated to maximize the utility of future collections. Herbarium “best practices” are perhaps implicitly agreed upon but not universal and far from enforced. These practices should be standardized to require that collections routinely include other explicitly defined metadata such as associated species, habitat type, light environment, edaphic conditions, size of population, plant height, digital images of field site and organism in situ, associated data sets, etc. Other considerations also need to be made, such as whether duplicate specimens represent the same individual and whether a given sheet includes multiple individuals, which is critical information for future genetic analyses. Lastly, destructive analyses are becoming more common yet the amount of herbarium material is finite. Infrastructure should be developed to facilitate this new wave of herbarium specimen use, such as including “secondary vouchers” (e.g., explicit curation of derived products from an organism for future genetic, chemical, or otherwise destructive analysis; Kageyama et al., 2006). Recent unanticipated uses provide powerful examples that validate the importance of biological collections. In an era of transparency and open data, herbarium vouchers should ideally become more prevalent as they serve as primary data points for nearly all botanical research (Schilthuizen et al., 2015). Although ongoing digitization efforts undoubtedly improve accessibility and enhance use, the continued collection and curation of the specimens themselves remains necessary because many uses depend upon the physical herbarium material. What unforeseen transformative technologies might be applied to herbarium specimens in the future? That is unknown, but the uses of herbaria will undoubtedly grow—if we cultivate the potential. As the wide-reaching effects of the Anthropocene are upon us, biological collections are vital to document biological and environmental change (Pyke and Ehrlich, 2010; Lavoie, 2013). Centuries of collections and the largely untapped information within are becoming widely accessible. We are now poised to answer scientific questions at levels previously not possible. This work was supported by a U.S. National Science Foundation Postdoctoral Research Fellowship in Biology to J.M.H. (1612079). We thank Pamela Diggle and four anonymous reviewers for comments that significantly improved the manuscript. We also thank Alison Hale, Anna Johnson, Susan Kalisz, Alan Prather, and Stephen Tonsor for discussion and comments on earlier versions of this essay. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Heberling et al. (Sat,) studied this question.