Natural history collections are critical resources for contemporary and future studies of urban evolution

Abstract Urban environments are among the fastest changing habitats on the planet, and this change has evolutionary implications for the organisms inhabiting them. Herein, we demonstrate that natural history collections are critical resources for urban evolution studies. The specimens housed in these collections provide great potential for diverse types of urban evolution research, and strategic deposition of specimens and other materials from contemporary studies will determine the resources and research questions available to future urban evolutionary biologists. As natural history collections are windows into the past, they provide a crucial historical timescale for urban evolution research. While the importance of museum collections for research is generally appreciated, their utility in the study of urban evolution has not been explicitly evaluated. Here, we: (a) demonstrate that museum collections can greatly enhance urban evolution studies, (b) review patterns of specimen use and deposition in the urban evolution literature, (c) analyze how urban versus rural and native versus nonnative vertebrate species are being deposited in museum collections, and (d) make recommendations to researchers, museum professionals, scientific journal editors, funding agencies, permitting agencies, and professional societies to improve archiving policies. Our analyses of recent urban evolution studies reveal that museum specimens can be used for diverse research questions, but they are used infrequently. Further, although nearly all studies we analyzed generated resources that could be deposited in natural history collections (e.g., collected specimens), a minority (12%) of studies actually did so. Depositing such resources in collections is crucial to allow the scientific community to verify, replicate, and/or re‐visit prior research. Therefore, to ensure that adequate museum resources are available for future urban evolutionary biology research, the research community—from practicing biologists to funding agencies and professional societies—must make adjustments that prioritize the collection and deposition of urban specimens.

journals were chosen to include representatives of broadly impactful journals, or journals that would reach a diverse audience (read by an audience from a broad taxonomic and/or subdiscipline background), and journals that were taxonomic-specific and included one journal for each major taxon. We limited the search to papers published between 2009 and 2019 to focus primarily on current practices of specimen use and deposition. The initial search results yielded 236 papers which were reduced to 84 papers after excluding reviews, projects solely focused on human or social changes, and projects which only relied on lab-reared organisms (Supplemental Table 1). We scored each study by whether they reported using museum specimens in their sampling (e.g., acknowledged a museum for specimens and/or provided specimen accession numbers; studies that could not have used specimens were marked as "not applicable"), whether they produced a resource that could be deposited in museums (voucher specimen, tissue sample, image, or recording), and whether they deposited any of these resources into a museum collection (any indication that some representative specimens were deposited).
We excluded fossil data by filtering all records with "fossil" in the preparation field, and excluded all records that had an entry in any of the following fields, indicating that they were include a stratigraphic position: "earliestorloweststage", "earliesteonorlowesteonothem", "esarliestepochorlowestseries", or "earliesteraorlowesterathem", "earliestperiodorlowestsystem".
We further refined the non-avian reptile data to remove any families with no extant members.
For birds, we excluded specimens from the Cornell Lab of Ornithology, given the large number of images and sounds that are part of the collection from the Macaulay Library. While these are important records, we wanted to focus our analyses on physical specimens given other digital records are a recent development and are limited in their utility.
To obtain the proportion impervious surface for each county, within Google Earth Engine To understand specimen deposition patterns for native and nonnative species, we analyzed differences between native and nonnative species deposition patterns in California.
We chose California as its large size and diverse habitat should be broadly representative and it contains several large museums, but is small enough to score each species appropriately. First, we extracted all records with "California" in the "stateprovince" field from the four VertNet datasets used above. Next, we classified each species as "native" or "nonnative", but excluded any species likely to be found only in zoos or ornamental gardens and marine mammals. We also excluded any species with fewer than five records in the entire dataset as these largely represented species that are not established in the area of interest. Next, for each species we counted the number of specimens deposited in museum collections in each decade, and compared the counts of native and nonnative species for each class of organisms for each decade. For each decade we excluded all species with no specimens collected so that we could compare rates of deposition between groups. There were not enough nonnative species of reptile or amphibian to statistically compare mean numbers of specimens collected per species between native and nonnative species, but for birds and mammals we compared average values with t-tests (counts of specimens and species per status category per decade available in Supplemental Tables 2-5).