Natural History Collections (NHCs) constitute the world’s largest repositories of long-term biodiversity datasets and are fundamental in tackling the questions of modern science, amongst which is species conservation. To do so, NHCs rely on specimen collection and voucher deposition, which, in recent decades, have seen a drastic decline. As part of an ongoing project to review the Portuguese zoological collections in the country’s NHCs, mammal data from its three major museums and smaller collections were compiled and analysed to understand the coverage and representation of the different taxa extant in Portugal. We found that the collections are not taxonomically, geographically or temporally complete. Approximately 86% of Portuguese mammalian species are represented in the country’s NHCs, but more than two-thirds of the taxa are represented by less than 50 specimens. Geographically, the collections cover little over 40% of the country’s territory and more than half the occurring taxa have less than 10% of their known distributions represented in the collections. A third of the taxa represented in the collections is considered threatened, yet only accounts for 15% of all specimens. A review on the status of Portuguese mammal collections is presented here, as well as a call to attention on the implications of incomplete collections towards fundamental research.

Collectomics, mammalia, natural history collections, Portuguese collections, taxonomy, specimen collecting

Natural history collections (NHCs) have been the backbone of scientific research and the basis for fundamental research in the last two centuries. They provide long-term repositories for specimen-based occurrence records that can be continuously reviewed and reassessed, thus playing a critical role as guarantors of the scientific nature of biological knowledge. Nowadays, the missions of NHCs have expanded from traditional realms to answering the needs of modern society (Miller et al. 2020), but, at the core, their goals remain the same – the collection, preservation and accessibility of biological specimens, its scientific study and the dissemination of scientific knowledge (Andreone et al. 2022).

Specimen-based research has changed throughout history, shifting from a mostly taxonomic and systematics approach to one encompassing other disciplines, such as ecology, population genetics and environmental pollution (Suarez and Tsutsui 2004). NHCs are unparalleled repositories of biodiversity data, from a taxonomic, geographic and temporal standpoint (Cook et al. 2014; Hedrick et al. 2020; Miller et al. 2020; Heberling et al. 2021; Hilton et al. 2021), with specimen collecting and voucher deposition being an essential step to the growth of these repositories (Rocha et al. 2014a; Malaney and Cook 2018; Monckton et al. 2020). Despite this importance, in recent years, there has been a steep decline in new accessions (Gardner et al. 2014; Troudet et al. 2018; Buckner et al. 2021; Thompson et al. 2021). Reluctance towards collecting has led to an upsurge in observation-based occurrences and non-invasive sampling, as well as the development of new methods (Gaiji et al. 2013; Minteer et al. 2014; Troudet et al. 2018; Balázs et al. 2020; Emami-Khoyi et al. 2021; Schilling et al. 2022; Byrne 2023). However, as replication is a core principle of the scientific method, without accessible voucher specimens research often becomes impoverished and non-replicable (Turney et al. 2015; Ceríaco et al. 2016; Monckton et al. 2020; Thompson et al. 2021; Nachman et al. 2023), leading to an incapacity to respond to the current biodiversity crisis (Krell and Wheeler 2014; Rocha et al. 2014a; Ceballos et al. 2015, 2017).

NHCs have been facing challenges like the lack of funding and training of new personnel (Vivo et al. 2014; Bezerra 2015; Kemp 2015; Orr et al. 2020; Wheeler 2020). Since the 1990s, financing cuts have hit collections all over the world (Cullota 1992; Dalton 2003; Froelich 2003; Andreone et al. 2014; Salvador and Cunha 2020) and, without proper structure and maintenance, irreplaceable collections have become at risk of being lost (Ceríaco et al. 2021). Andreone et al. (2014, 2022) and Gippoliti and Aloise (2016) provided the grim reality of the status of Italian natural history collections, which suffer from decreasing investment, lack of personnel and fieldwork assessments, threatening the knowledge of alpha taxonomy and conservation of mammal diversity. Vivo et al. (2014), in an overview of Brazilian natural history collections, argued that the status quo of the three major collections as part of universities hinders both the collections and its associated researchers.

In line with its international and European congeners, the challenges affecting NHCs in Portugal have become a topic of recent discussions (see Ceríaco et al. (2021); Donahue (2022); Santos et al. (2024)). Santos et al. (2024) showed that Portuguese NHCs fail to provide a good taxonomic, geographic and temporal coverage of Portuguese herpetofauna, noting a surprising low number of specimens of widely distributed and common species and an evident lack of national and institutional strategies to tackle the poor representation of local herpetofauna in Portuguese collections. The authors refer to the putative causes of this situation and conclude that, while they are in line with the international trend of neglect affecting NHCs and specimen collecting activities, these are also affected by the history of Portuguese natural history collections, as well as national and institutional policies (see Ceríaco (2021); Ceríaco et al. (2021)). Santos et al. (2024) also warn of the overall present and future scientific consequences caused by such a low number of specimens and geographically and temporally biased collections. This situation has yet to be addressed for collections of other taxonomic groups, such as invertebrates, fishes, birds and mammals, but similar trends are expected.

Mammals are amongst some of the most iconic groups of animals and are traditionally one of the best studied taxonomic groups. Over a quarter of mammals assessed by the IUCN Red List are considered threatened (IUCN 2024), facing several direct and indirect threats like habitat loss and displacement or hunting (Ferguson 2020). Cook and Light (2019) foresaw a continued interest in NHCs to future mammalogy, taking advantage of the modern technological advancements that can be applied to specimen data, such as genome scale data with spatial analysis, micro-CT scans and microbe studies. Dunnum et al. (2018) presented the most recent assessment of mammal collections from North, Central and South America and the Caribbean Region and found that spatial and temporal gaps exist within these collections. Similarly, Chiquito et al. (2021) assessed Brazilian mammal collections and reported that, while housing a considerable volume of specimens, the collections’ numbers are still insufficient to represent Brazilian mammalian diversity. The overall trend of decreasing collecting efforts is affecting the assessed collections, even the larger and historically active institutions, such as those in Europe, threatening long-term coverage (Bezerra 2015; Dunnum et al. 2018).

As part of the Iberian Peninsula biodiversity hotspot (Myers et al. 2000; Rosso et al. 2017), Portugal hosts a remarkable diversity of mammal species, especially when considering its considerable small area of 92,212 km². A total of 74 terrestrial mammal species have been confirmed to occur in the country (Table 1), of which 10 are endemic to the Iberian Peninsula and two are endemic to the Madeira/Azores Islands (Bencatel et al. 2019; Mathias et al. 2023). Most of these species are currently classified as Least Concern following the IUCN conservation guidelines, but 24 (one Critically Endangered, 11 Endangered, 12 Vulnerable) are threatened (see Table 2; Russo and Cistrone (2023a, 2023b); Mathias et al. (2023, 2024)). Three taxa are reported as extinct in the country: the Eurasian beaver Castor fiber, the brown bear Ursus arctos and an endemic subspecies of Iberian ibex, Capra pyrenaica lusitanica. The last known record of C. fiber dates from the 15^th century, while U. arctos has had no resident specimens in the country since the 19^th century (Bencatel et al. 2019). Capra pyrenaica lusitanica was extinct in the early 1890s (Almaça 1992). Nine mammals, mostly ungulates, are classified as game species according to the Portuguese legislation and are commonly hunted throughout their designated hunting seasons. Two of these, the European fallow deer, Dama dama and the European mouflon, Ovis aries musimon, were introduced exclusively for hunting purposes. Additionally, two species, Arvicola scherman and Pipistrellus nathusii, are of dubious presence in the country and are known from less than a handful of reports (Oliveira and Vieira 1896; Seabra 1910; Ayres 1914; Ramalhinho and Mathias 1988).

Despite its scientific, cultural and even economic importance, a complete survey of the Portuguese mammal collections remains to be done. Following the approach of Santos et al. (2024), we aim to review and illustrate the major patterns and trends related to the presence of Portuguese terrestrial mammals in national collections and, more specifically, answer the following questions: 1) Are all terrestrial mammal taxa occurring in Portugal represented in Portuguese collections? 2) How complete is the geographic coverage of such collections? 3) Do the current collections present temporal trends allowing them to be considered good temporal series? Furthermore, we want to understand how species of conservation concerns, as well as game species, are represented in collections.

To assess the current situation of Portuguese terrestrial mammals housed in Portuguese NHCs, we consulted the collections of the country’s three main Natural History Museums (Fig. 1) – the Museu Nacional de História Natural e da Ciência (MUHNAC, Lisbon), the Museu da Ciência da Universidade de Coimbra (MCUC, Coimbra) and the Museu de História Natural e da Ciência da Universidade do Porto (MHNC-UP, Porto) – all part of large public universities, as well the Aquário Vasco da Gama (AVG, Lisbon) collection. The collections of former Instituto de Investigação Científica Tropical (IICT) were integrated into the MUHNAC in 2015 and its Portuguese specimens are partially accounted for here. The current accessibility of the IICT collection, due to its conservation state and need for reconditioning, impeded the total inclusion of all its specimens in this study. The unaccounted IICT material totals around 150 specimens, exclusively represented by rodent species from the family Muridae (BSS pers. obs.).

Following the methodology already presented in a previous study on the Portuguese herpetology collections (see Santos et al. (2024)), for each collection, all available metadata associated with existing specimens was gathered, namely taxonomic identification, collecting date and locality data. Part of these data was already available on GBIF (Brites Soares et al. 2022), but most of it was compiled, based on the extant internal databases, entry catalogues and in situ review of the specimens. Whenever possible and especially for dubious identifications or difficult taxonomic groups, specimens were examined to confirm their taxonomic identification. All data were standardised following Darwin Core standards (Darwin Core 2022) and locality information was, when possible, georeferenced following the protocols of Chapman and Wieczorek (2020).

Taxonomy and nomenclature follow the Portuguese Atlas of Mammals (Bencatel et al. 2019), as well as updates from the Red Book of Mammals (Mathias et al. 2023, 2024) and the accepted occurring species in the country, including native and exotic species, are presented in Table 1. To assess geographic representativeness of the assessed collections, we mapped the extant specimens of the different taxa against their known distribution ranges. Species distributions were based on all records made available through the Portuguese Atlas of Mammals (Bencatel et al. 2019), as well as more recent distribution datasets, such as those by Grilo et al. (2022) and the Red Book of Mammals for Mainland Portugal (Mathias et al. 2023, 2024). Bat distributions are adapted from Rainho et al. (2002, 2013) and the Red Book of Mammals (Mathias et al. 2023, 2024). Species occurrence data in continental Portugal are projected, based on the UTM-29N (Universal Transverse Mercator) 10 × 10 km grid. For the autonomous regions of Azores and Madeira, species presence is registered by island following Bencatel et al. (2019) and projected based on the UTM-25N, UTM-26N and UTM-28N grids. Conservation status for each species follows Mathias et al. (2023, 2024), while the legal classification of game species was retrieved from relevant Portuguese legislation, namely Portaria n.° 100/2021, from the Ministry of Environment and Climatic Action (available at https://diariodarepublica.pt/dr/detalhe/portaria/100-2021-163012629). Only terrestrial mammals were included in this study and a later publication will focus on marine mammals (Santos et al., in prep).

A total of 7402 specimens with associated data were recorded (Table 1). The collections are home to additional specimens requiring full assessment and reconditioning that were not used in this study. Detailed results are provided in the following subsections.

Taxonomic representation

Of the 74 species of terrestrial mammals currently known to occur in Portugal, 64 are represented in the collections. One ungulate, the European mouflon Ovis aries musimon, one carnivore, the American mink Neovison vison, one rodent, the European snow vole Chionomys nivalis and seven bats, the Alcathoe bat Myotis alcathoe, the cryptic myotis Myotis crypticus, the whiskered bat Myotis mystacinus, the soprano pipistrelle Pipistrellus pygmaeus, the meridional serotine Eptesicus isabellinus, the common noctule Nyctalus noctula and the western barbastelle Barbastella barbastellus are not represented in the collections.

When considering taxonomic representation within individual collections, there are considerable differences amongst them. MUHNAC’s collections are the most complete and taxonomically diverse, with 61 out of 74 taxa represented (82.4%), while also holding 88% of the existing specimens in Portuguese museums (Table 1). MCUC’s collection covers 60.8% (45 of 74) of occurring taxa and its collections correspond to 3% of the existing specimens in the country. MHNC-UP’s collection contains 42 of 74 taxa (56.8%) and amounts to 9% of specimens in the country’s NHCs. The remaining analysed collection presents a scant (less than 10) number of specimens.

Two rodents, the house mouse Mus musculus and the Algerian mouse Mus spretus and one carnivore, the red-fox Vulpes vulpes, are the three most represented species in the collections, corresponding to 41.6% of all specimens. Almost two-thirds (44) of the taxa are represented by less than 50 specimens and 24 species are represented by 10 or less specimens (Table 1). Besides this, 18 taxa exist in only one of the assessed museums. Rodents are the most represented group, accounting for almost 60% of all specimens, followed by the Carnivora with 20% and the Eulipotyphla with 11% (Fig. 2). A more detailed analysis of each group is presented below.

Figure 1. 

Mammal collections of the Museu Nacional de História Natural e da Ciência, (MUHNAC, Lisbon) (A, B), the Museu de História Natural e da Ciência da Universidade do Porto (MHNC-UP, Porto) (C) and the Museu da Ciência da Universidade de Coimbra (MCUC, Coimbra) (D). Most of the specimens housed in the MHNC-UP and MCUC collections are taxidermy mounts. Photos by Bruna S. Santos (A, B, D) and Luis M. P. Ceríaco (C).

Figure 2. 

Proportion of extant Portuguese terrestrial mammal specimens per taxonomic order.

Table 1.

Download as

CSV

XLSX

Taxonomic diversity of terrestrial mammals occurring in Portugal and corresponding number of existing specimens for each species in the consulted museum/collection. Asterisks (*) denote introduced species. Museums: MCUC – Museu da Ciência da Universidade de Coimbra, MHNC-UP – Museu de História Natural e da Ciência da Universidade do Porto, MUHNAC – Museu Nacional de História Natural e da Ciência, IICT – Instituto de Investigação Científica Tropical, AVG – Aquário Vasco da Gama.

Taxa	MCUC	MHNC-UP	MUHNAC + IICT	AVG	TOTAL
Insectivores (Eulipotyphla)
European hedgehog Erinaceus europaeus	9	5	48	–	62
Spanish mole Talpa occidentalis	10	12	101	–	123
Pyrenean desman Galemys pyrenaicus	6	3	10	–	19
Eurasian pygmy shrew Sorex minutus	–	–	6	–	6
Iberian shrew Sorex granarius	–	–	30	–	30
Lesser white-toothed shrew Crocidura suaveolens	1	–	–	–	1
Greater white-toothed shrew Crocidura russula	3	8	567	–	578
Mediterranean water shrew Neomys anomalus	–	–	4	–	4
Etruscan shrew Suncus etruscus	–	–	4	–	4
Carnivores (Carnivora)
Iberian wolf Canis lupus signatus	13	4	92	–	109
Rex fox Vulpes vulpes	7	516	486	–	1009
Stoat Mustela erminea	–	–	5	–	5
Least weasel Mustela nivalis	6	8	42	–	56
European polecat Mustela putorius	7	5	26	–	38
American mink Neovison vison *	–	–	–	–	–
European pine marten Martes martes	–	–	1	–	1
Beech marten Martes foina	4	2	33	–	39
European badger Meles meles	4	7	22	–	33
Eurasian otter Lutra lutra	6	3	6	4	19
Egyptian mongoose Herpestes ichneumon	1	3	22	–	26
Common genet Genetta genetta	4	6	98	–	108
Wildcat Felis silvestris	3	3	13	–	19
Iberian lynx Lynx pardinus	3	2	7	–	12
Ungulates (Artiodactyla)
Wild boar Sus scrofa	3	2	11	–	16
Red deer Cervus elaphus	1	1	2	–	4
European fallow deer Dama dama *	–	2	2	–	4
Roe deer Capreolus capreolus	2	2	1	–	5
Iberian ibex Capra pyrenaica	1	–	–	–	1
European Mouflon Ovis aries musimon *	–	–	–	–	–
Rodents (Rodentia)
Red squirrel Sciurus vulgaris	1	5	2	–	8
Garden dormouse Eliomys quercinus	6	6	194	–	206
Southwestern water vole Arvicola sapidus	4	2	36	–	42
European snow vole Chionomys nivalis	–	–	–	–	-
Lusitanian pine vole Microtus lusitanicus	5	1	558	–	564
Mediterranean pine vole Microtus duodecimcostatus	–	–	454	–	454
Common vole Microtus arvalis	–	5	–	–	5
Cabrera’s vole Microtus cabrerae	–	–	10	–	10
Short-tailed field vole Microtus rozianus	–	7	17	–	24
Wood mouse Apodemus sylvaticus	3	2	585	–	590
Black rat Rattus rattus	7	3	273	–	283
Brown rat Rattus norvegicus	14	1	91	–	106
House mouse Mus musculus	3	3	1102	–	1108
Algerian mouse Mus spretus	3	–	958	–	961
Lagomorphs (Lagomorpha)
Granada hare Lepus granatensis	10	3	68	–	81
European rabbit Oryctolagus cuniculus	3	4	83	–	90
Bats (Chiroptera)
Greater horseshoe bat Rhinolophus ferrumequinum	7	3	35	–	45
Lesser horseshoe bat Rhinolophus hipposideros	4	3	37	–	44
Mehely’s horseshoe bat Rhinolophus mehelyi	1	2	51	–	54
Mediterranean horseshoe bat Rhinolophus euryale	6	1	16	–	23
Greater mouse-eared bat Myotis myotis	9	5	33	–	47
Lesser mouse-eared bat Myotis blythii	–	–	27	–	27
Escalera’s bat Myotis escalerai	1	2	3	–	6
Geoffroy’s bat Myotis emarginatus	4	–	1	–	5
Whiskered bat Myotis mystacinus	–	–	–	–	–
Bechstein’s bat Myotis bechsteinii	–	–	5	–	5
Daubenton’s bat Myotis daubentonii	1	–	13	–	14
Alcathoe bat Myotis alcathoe	–	–	–	–	–
Cryptic myotis Myotis crypticus	–	–	–	–	–
Common pipistrelle Pipistrellus pipistrellus	8	1	53	–	62
Soprano pipistrelle Pipistrellus pygmaeus	–	–	–	–	–
Kuhl’s pipistrelle Pipistrellus kuhlii	1	2	7	–	10
Madeira pipistrelle Pipistrellus maderensis	–	–	1	–	1
Savi’s pipistrelle Hypsugo savii	–	–	1	–	1
Serotine bat Eptesicus serotinus	4	3	24	–	31
Meridional serotine Eptesicus isabellinus	–	–	–	–	–
Common noctule Nyctalus noctula	–	–	–	–	–
Greater noctule bat Nyctalus lasiopterus	–	–	1	–	1
Lesser noctule Nyctalus leisleri	–	–	7	–	7
Azores noctule Nyctalus azoreum	–	–	10	–	10
Grey long-eared bat Plecotus austriacus	–	–	12	–	12
Brown long-eared bat Plecotus auritus	5	4	1	–	10
Western barbastelle Barbastella barbastellus	–	–	–	–	–
Common bent-wing bat Miniopterus schreibersii	7	4	104	–	115
European free-tailed bat Tadarida teniotis	9	–	1	–	10
TOTAL	220	666	6512	4	7402

Temporal patterns

Collecting date records are known from approximately 87% of specimens. Collection growth and specimen collecting have historically been inconsistent, both at national and institutional levels. Although all NHCs have specimens from the 19^th century, most of these older specimens are found in MCUC, as the other collections were mostly assembled in the 20^th century (Fig. 3). Nowadays, MUHNAC holds the largest collection out of the three museums (Fig. 4) and its specimens date in large part from the 1980s onward, as the historical collections (mid-18^th century onwards to 1978) were destroyed by a fire in 1978. Pre-1978 specimens extant today represent those from the IICT collections, as well as later incorporations from other research institutions and researcher’s personal collections. MUHNAC’s collecting efforts were particularly high through the 1980s and 1990s due to research groups associated with the Faculty of Sciences of the University of Lisbon. Specimens of the former IICT collections result in great part from pathology studies and were collected mostly during the 1960s, while the other incorporations are largely reflected in the 1970s collecting peak. From then on, new accessions were largely opportunistic and considerably rarer. Despite containing the oldest specimens in the country, with most of its specimens being from the late 19^th century, MCUC’s collection has had no new entries since 1935 (Fig. 3). MHNC-UP’s collection was amassed primarily in its founding days in the 1890s and 1900s, then spent most of the following century with less than five new records per year (Fig. 3). A recent singular incorporation of red fox specimens contributed to the collecting peak from the late 2000s to late 2010s (Fig. 3). Overall, these patterns are similar to those previously reported for the herpetological specimens in Portuguese collections (Santos et al. 2024).

Figure 3. 

Recorded collections of Portuguese mammals per year from the 1870s to 2020, for the MUHNAC + IICT, MCUC and MHNC-UP. Only specimens with available collecting date were used in this analysis. Note that the scales differ in each graphic representation and the most recent MCUC record dates from the 1930s. MUHNAC + IICT only reflects currently extant specimens and does not account for those existing in the MUHNAC collections before the 1978 fire.

Figure 4. 

Aggregated growth of the MUHNAC + IICT (pink), MCUC (yellow) and MHNC-UP (blue) mammal collections. Only specimens with available collecting date were used in this analysis. MUHNAC + IICT only reflects currently extant specimens and does not account for those existing in the MUHNAC collections before the 1978 fire.

Geographic coverage

Country-wide collecting effort for mammals is geographically biased (Fig. 5). Several regions have historically been more intensely surveyed, like the vicinities of the cities where the three main museums are located at (Lisbon, Coimbra and Porto), with other clusters found in the Parque Natural de Montesinho (PNM), Central Alentejo Region, Algarve and the autonomous regions of Madeira and the Azores. A total of 838 localities were recorded, although 42.8% of them (359) are represented by a single specimen.

When focusing on institutional collecting efforts, the biases become more apparent. MCUC specimens were collected in the surroundings of the Coimbra Region, with some scattered exceptions in northern Portugal, the Beiras Region and south towards Central Alentejo (Fig. 6). MHNC-UP displays the same regional bias, with most of its specimens from northern Portugal and localities adjacent to the museum, with other dispersed records mostly along the coastline. MUHNAC’s collection covers much more of the mainland, but also the Madeira and Azores Archipelagos. Corvo Island, from the autonomous region of the Azores, is the only insular region not represented in the collections. A considerable part of the more “modern” collection of MUHNAC is owed to the projects of museum-associated researchers and reflects their study regions (Mathias 1990; Mira and Mathias 1994; Libois et al. 1996a, 1996b, 1997; Collares-Pereira et al. 1997, 2000; Mathias et al. 2004).

A quantitative overview of geographic coverage can be obtained by dividing the country with a 10 × 10 km grid, as used on the Portuguese Atlas of Mammals (Bencatel et al. 2019). Overall, museum collections cover 42.4% of the country (Fig. 7). Individually, MCUC covers 4.1%, MHNC-UP 13.1% and MUHNAC 32.3%. Around 58.8% of squares are represented by a single collecting locality. Detailed accounts for each species using the same 10 × 10 km grid comparing museum specimens and known distributions are presented in Figs 8–21. Whenever museum specimen data fell outside of known distributions, we accounted for those records as additional distribution data and added them to the previously known in order to evaluate percentage of coverage.

No taxon has more than 50% of their known distribution represented in collections. A bat, Nyctalus azoreum and two rodents, Microtus arvalis and Rattus rattus, are the three best represented species in the collections (Table 2; Figs 20C, 14B, 15A, respectively), having, respectively, 42.9%, 40% and 20.2% of their known distributions represented in the collections. The Azores noctule, Nyctalus azoreum (Fig. 20C), is endemic to the Azores Archipelago and has a limited distribution, which explains the 42.9% covered by museum collections. The common vole Microtus arvalis (Fig. 14B) is still poorly known in the country and seems to have a limited distribution, hence the higher percentage of coverage. The black rat, Rattus rattus (Fig. 15A), established throughout mainland Portugal and the Madeira and Azores Archipelagos, was the target of several research projects that contributed to the 20.2% of coverage. Of the extant species, 59 have less than 20% of their distribution represented; of these, 38 have less than 10% and five species have less than 1%. Surprisingly, amongst the latter are three ungulate species – the wild boar Sus scrofa, the red deer Cervus elaphus and the roe deer Capreolus capreolus – all game species (see below), as well as the European pine marten Martes martes and the European free-tailed bat Tadarida teniotis.

Table 2.

Download as

CSV

XLSX

Conservation status and percentage of representation (%) of Portuguese terrestrial mammal species. Conservation Status: LC – Least Concern, NT – Near Threatened, VU – Vulnerable, EN – Endangered, NA – Not Applicable, NE – Not evaluated. * Denotes non-native species, § denotes species protected under the Habitats Directive Annex II. Number of represented squares column displays the number of 10 × 10 km squares for which there are museum specimens from a certain taxon versus the number of squares from which a species is known, following Bencatel et al. (2019) and Mathias et al. (2023, 2024).

Taxa	Conservation status	Iberian endemic	Game species	Number of represented squares	Percentage of representation (%)	Map
Insectivores (Eulipotyphla)
European hedgehog Erinaceus europaeus	LC	No	No	36/517	7	Fig. 8A
Spanish mole Talpa occidentalis	LC	Yes	No	47/377	12.5	Fig. 8B
Pyrenean desman Galemys pyrenaicus §	EN	No	No	9/219	4.1	Fig. 8C
Eurasian pygmy shrew Sorex minutus	EN	No	No	3/17	17.6	Fig. 8D
Iberian shrew Sorex granarius	VU	Yes	No	9/90	10	Fig. 8E
Lesser white-toothed shrew Crocidura suaveolens	EN	No	No	1/28	3.6	Fig. 9A
Greater white-toothed shrew Crocidura russula	LC	No	No	75/447	16.8	Fig. 9B
Mediterranean water shrew Neomys anomalus	VU	Yes	No	1/60	1.7	Fig. 9C
Etruscan shrew Suncus etruscus	LC	No	No	2/105	1.9	Fig. 9D
Carnivores (Carnivora)
Iberian wolf Canis lupus signatus §	EN	Yes	No	52/272	19.1	Fig. 9E
Rex fox Vulpes vulpes	LC	No	Yes	171/992	17.2	Fig. 9F
Stoat Mustela erminea	DD	No	No	5/44	11.4	Fig. 10A
Least weasel Mustela nivalis	LC	No	No	33/278	11.9	Fig. 10B
European polecat Mustela putorius §	EN	No	No	24/260	9.2	Fig. 10C
American mink Neovison vison *	NA	No	No	–	–	–
European pine marten Martes martes	VU	No	No	1/136	0.7	Fig. 11A
Beech marten Martes foina	LC	No	No	19/578	3.3	Fig. 11B
European badger Meles meles	LC	No	No	16/709	2.3	Fig. 11C
Eurasian otter Lutra lutra §	LC	No	No	13/956	1.4	Fig. 11D
Egyptian mongoose Herpestes ichneumon	LC	No	Yes	12/912	1.3	Fig. 11E
Common genet Genetta genetta	LC	No	No	35/645	5.4	Fig. 11F
Wildcat Felis silvestris	EN	No	No	10/294	3.4	Fig. 12A
Iberian lynx Lynx pardinus §	EN	Yes	No	10/146	6.8	Fig. 12B
Ungulates (Artiodactyla)
Wild boar Sus scrofa	LC	No	Yes	8/1000	0.8	Fig. 12C
Red deer Cervus elaphus	LC	No	Yes	3/427	0.7	Fig. 12D
European fallow deer Dama dama *	NA	No	Yes	2/157	1.3	Fig. 12E
Roe deer Capreolus capreolus	LC	No	Yes	3/426	0.7	Fig. 13A
Iberian ibex Capra pyrenaica	NT	Yes	No	1/18	5.6	Fig. 13B
European Mouflon Ovis aries musimon *	NA	No	Yes	–	–	–
Rodents (Rodentia)
Red squirrel Sciurus vulgaris	LC	No	No	6/448	1.3	Fig. 13C
Garden dormouse Eliomys quercinus	NT	No	No	8/134	6	Fig. 13D
Southwestern water vole Arvicola sapidus	VU	No	No	17/269	6.3	Fig. 13E
European snow vole Chionomys nivalis	DD	No	No	–	–	–
Lusitanian pine vole Microtus lusitanicus	LC	No	No	39/277	14.1	Fig. 13F
Mediterranean pine vole Microtus duodecimcostatus	LC	No	No	25/161	15.5	Fig. 14A
Common vole Microtus arvalis	DD	No	No	2/5	40	Fig. 14B
Cabrera’s vole Microtus cabrerae §	VU	Yes	No	7/240	2.9	Fig. 14C
Short-tailed field vole Microtus rozianus	VU	Yes	No	7/103	6.8	Fig. 14D
Wood mouse Apodemus sylvaticus	LC	No	No	57/462	12.3	Fig. 14E
Black rat Rattus rattus *	NA	No	No	42/208	20.2	Fig. 15A
Brown rat Rattus norvegicus *	NA	No	No	22/127	17.3	Fig. 15B
House mouse Mus musculus	LC	No	No	37/273	13.6	Fig. 16A
Algerian mouse Mus spretus	LC	No	No	54/383	13.6	Fig. 16B
Lagomorphs (Lagomorpha)
Granada hare Lepus granatensis	VU	Yes	Yes	21/794	2.6	Fig. 16C
European rabbit Oryctolagus cuniculus §	VU	Yes	Yes	18/1013	1.8	Fig. 17A
Bats (Chiroptera)
Greater horseshoe bat Rhinolophus ferrumequinum §	LC	No	No	11/260	4.2	Fig. 17B
Lesser horseshoe bat Rhinolophus hipposideros §	LC	No	No	16/274	5.8	Fig. 17C
Mehely’s horseshoe bat Rhinolophus mehelyi §	EN	No	No	9/61	14.8	Fig. 17D
Mediterranean horseshoe bat Rhinolophus euryale §	EN	No	No	10/78	12.8	Fig. 18A
Greater mouse-eared bat Myotis myotis §	VU	No	No	16/157	10.2	Fig. 18B
Lesser mouse-eared bat Myotis blythii §	CR	No	No	5/44	11.4	Fig. 18C
Escalera’s bat Myotis escalerai	VU	No	No	3/124	2.4	Fig. 18D
Geoffroy’s bat Myotis emarginatus §	EN	No	No	2/57	3.5	Fig. 18E
Whiskered bat Myotis mystacinus	VU	No	No	–	–	–
Bechstein’s bat Myotis bechsteinii §	DD	No	No	–	–	–
Daubenton’s bat Myotis daubentonii	LC	No	No	7/186	3.8	Fig. 19A
Alcathoe bat Myotis alcathoe	NE	No	No	–	–	–
Cryptic myotis Myotis crypticus	NE	No	No	–	–	–
Common pipistrelle Pipistrellus pipistrellus	LC	No	No	20/350	5.7	Fig. 19B
Soprano pipistrelle Pipistrellus pygmaeus	LC	No	No	–	–	–
Kuhl’s pipistrelle Pipistrellus kuhlii	LC	No	No	6/390	1.5	Fig. 19C
Madeira pipistrelle Pipistrellus maderensis	NE	Yes	No	1/9	11.1	Fig. 19D
Savi’s pipistrelle Hypsugo savii	LC	No	No	1/78	1.3	Fig. 19E
Serotine bat Eptesicus serotinus	LC	No	No	10/80	12.5	Fig. 20A
Meridional serotine Eptesicus isabellinus	LC	No	No	–	–	–
Common noctule Nyctalus noctula	DD	No	No	–	–	–
Lesser noctule Nyctalus leisleri	LC	No	No	4/226	1.8	Fig. 20B
Greater noctule bat Nyctalus lasiopterus	DD	No	No	–	–	–
Azores noctule Nyctalus azoreum	NE	Yes	No	3/7	42.9	Fig. 20C
Grey long-eared bat Plecotus austriacus	NT	No	No	6/121	5	Fig. 20D
Brown long-eared bat Plecotus auritus	DD	No	No	6/66	9.1	Fig. 21A
Western barbastelle Barbastella barbastellus §	LC	No	No	–	–	–
Common bent-wing bat Miniopterus schreibersii §	NT	No	No	21/156	13.5	Fig. 21B
European free-tailed bat Tadarida teniotis	LC	No	No	2/273	0.7	Fig. 21C

Figure 5. 

Map of collecting localities for all collections. Increasing circle size and colour intensity corresponds to the number of records for each location.

Figure 6. 

Map of collecting localities for all collections: AVG (green), MCUC (yellow), MHNC-UP (blue) and MUHNAC + IICT (pink).

Figure 7. 

Map of collecting localities for all collections, presented on a 10 × 10 km grid following Bencatel et al. (2019).

Figure 8. 

Distribution maps of A Erinaceus europaeus, B Talpa occidentalis, C Galemys pyrenaicus, D Sorex minutus and E Sorex granarius, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by William Harland (A), Diego González Dopico (B), David Perez (C) and Linda Lötjönen (D). Illustration by Maria João Santos (E).

Figure 9. 

Distribution maps of A Crocidura suaveolens, B Crocidura russula, C Neomys anomalus, D Suncus etruscus, E Canis lupus signatus and F Vulpes vulpes, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Roman (A), Sylvain Montagner (B), David Perez (C), Dario Taraborelli (D), zdc2021-garza (E) and Alexis Godin (F).

Figure 10. 

Distribution maps of A Mustela erminea, B Mustela nivalis and C Mustela putorius, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Laura Keene (A), vyatka (B) and bandwidthbob (C).

Figure 11. 

Distribution maps of A Martes martes, B Martes foina, C Meles meles, D Lutra lutra, E Herpestes ichneumon and F Genetta genetta, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Christoph Moning (A), mlouwsma (B), Valentino Traversa (C), Julia Moning (D), Victor Beccari (E) and Nuno Campos (F).

Figure 12. 

Distribution maps of A Felis silvestris, B Lynx pardinus, C Sus scrofa, D Cervus elaphus and E Dama dama, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by javielorriaga (A), Royle Safaris (B), Wendy McCrady (C), Tina Ellegaard Poulsen (D) and Frank Sengpiel (E).

Figure 13. 

Distribution maps of A Capreolus capreolus, B Capra pyrenaica, C Sciurus vulgaris, D Eliomys quercinus, E Arvicola sapidus and F Microtus lusitanicus, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by euqirneto (A), Yves Bas (B), toby (C), Robin Schmidt (D), David Perez (E) and Emilio Andújar (F).

Figure 14. 

Distribution maps of A Microtus duodecimcostatus, B Microtus arvalis, C Microtus cabrerae, D Microtus rozianus and E Apodemus sylvaticus, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Manuel Ruedi (A), Katrin Simon (B) and nbasargin (E). Illustrations by Maria João Santos (C), (D).

Figure 15. 

Distribution maps of A Rattus rattus, and B Rattus norvegicus, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Kai Squires (A) and Pavel Kungurov (B).

Figure 16. 

Distribution maps of A Mus musculus, B Mus spretus and C Lepus granatensis, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Andrea Adelfio (A), Julien Renoult (B) and Benoit Renaud (C).

Figure 17. 

Distribution maps of A Oryctolagus cuniculus, B Rhinolophus ferrumequinum, C Rhinolophus hipposideros and D Rhinolophus mehelyi, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Alex J. (A), Arnold Wijker (B), Johnnyrandom (C) and Nicola Steiner (D).

Figure 18. 

Distribution maps of A Rhinolophus euryale, B Myotis myotis, C Myotis blythii, D Myotis escalerai and E Myotis emarginatus, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Alfred Daniel J (A), Martin Grimm (B), batworker (C), joaopargana (D) and Manuel Ruedi (E).

Figure 19. 

Distribution maps of A Myotis daubentonii, B Pipistrellus pipistrellus, C Pipistrellus kuhlii, D Pipistrellus maderensis and E Hypsugo savii, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Damien Brouste (A), Yuzefovich Alexander (B), Alix Sanchis (C), Marc Domènech (D) and Vincent Rufray (E).

Figure 20. 

Distribution maps of A Eptesicus serotinus, B Nyctalus leisleri, C Nyctalus azoreum and D Plecotus austriacus, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Wesley Danes (A), Mateusz Ciechanowski (B), rjtizard (C) and Jakob Fahr (D).

Figure 21. 

Distribution maps of A Plecotus auritus, B Miniopterus schreibersii and C Tadarida teniotis, comparing museum records (black dots) with the species’ known extent in Portugal (pink). Respective collecting event chronograms are also presented. Live specimen photos retrieved from iNaturalist by Martin Grimm (A), Tyson Lee Holmes (B) and Manuel Ruedi (C).

Conservation

The collections hold 5991 specimens from the 32 species classified as “Least Concern”; 334 from the four species classified “Near Threatened”; 345 from the 12 species listed as “Vulnerable”; 286 from the 11 species listed as “Endangered” and 27 from the single species listed as “Critically Endangered” (see Fig. 22 and Table 2; Mathias et al. (2023, 2024); Russo and Cistrone (2023a, 2023b)). An additional 26 specimens are from seven species deemed “Data Deficient”, while 393 are from three species whose classification was considered “Non-Applicable”, which include non-native species (Table 2; see section below). The extinct Capra pyrenaica lusitanica is represented in the collections by a single specimen, a syntype extant in MCUC. A photo of the only living C. pyrenaica lusitanica specimen ever photographed is shown in Fig. 23. Additionally, 17 terrestrial mammal species extant in Portugal are part of the Habitats Directive Annex II (see Table 2), which aims to ensure the conservation of natural habitats and threatened fauna and flora species in the country. All bat species that occur in Europe are protected by the “Agreement on the Conservation of Populations of European Bats”, a treaty signed under the auspices of the Bonn Convention (Mathias et al. 2023).

Figure 22. 

Comparison chart of non-threatened (beige) vs. threatened (red tones) vs. no data (grey) specimens. Specimens are classified in categories of increasing danger, from “Least Concern” to “Critically Endangered”.

Figure 23. 

Only specimen of Capra pyrenaica lusitanica photographed alive. Photo adapted from Anonymous (1908), taken by Dr. Ricardo Jorge.

Figure 24. 

Comparison chart of non-game vs. game species specimens.

Despite most taxa being represented in the collections (64 of 74 species), over two-thirds of those accounted for have less than 50 specimens and only six species have over 500 specimens. As for geographic coverage, maps could not be produced for two species (Myotis bechsteinii and Nyctalus lasiopterus) due to lack of data quality and only three species have more than 20% of their known distribution represented in the collections. Species with threatened conservation statuses represent less than 15% of all specimens, despite corresponding to a third of the taxa represented in the collections. No museum collection is representative of all taxa extant in Portuguese NHCs and no species has a continued time-series of collected specimens.

This scenario reflects a tradition of lack of collecting in Portuguese biological sciences. Specimen collecting is a fundamental tool for present and future biological research (Prather et al. 2004; Rocha et al. 2014a; Hope et al. 2018; Miller et al. 2020; Fischer et al. 2021), but the global trend in current times has been a steep decline. Rohwer et al. (2022) recently uncovered a decline of over 50% in new amphibian, reptile, bird and mammal accessions in 245 institutions across the Globe. Several factors have contributed to the observed trend, from the emergence of legislation like the Wildlife Conservation and Protection Acts in the 1950s and the CITES Treaty, to global events such as the World Wars, economic crashes and pandemics and finally, an overall change in societal values that has led to a reluctance towards specimen collecting (Minteer et al. 2014; Rocha et al. 2014a; Byrne 2023; Nachman et al. 2023). The COVID-19 pandemic’s impact on collecting was the greatest recorded since before the 1900s, with collecting numbers hitting rock bottom (see data in Rohwer et al. (2022)).

In line with their international congeners, Portuguese NHCs are also experiencing a decrease in new recorded specimens, but in a more drastic way than its counterparts (Santos et al. 2024). As our results show, one of the main Portuguese NHCs, the MCUC, has had no new accessions of Portuguese mammal specimens since mid-last century, whereas the other two, although still receiving new material, have had their numbers severely slow down in comparison to previous time periods (Fig. 3). This paradigm is similar to that already observed for amphibians and reptiles (Santos et al. 2024), but also to other vertebrate and invertebrate groups (Santos et al. ongoing research, pers. obs.). According to Portuguese legislation, it is not mandatory that specimens sacrificed after capture be deposited in NHCs. As such, while some research may have required specimen collecting, it does not mean that those specimens made their way into publicly available collections. This decrease reflects the troubled modern history of Portuguese NHCs, which is severely impacted by both external factors and the internal struggles of each institution (Santos et al. 2024).

Historically, much of early research in Portuguese NHCs concerned the fauna and flora of the then overseas colonies (Almaça 1991; Ceríaco 2021), while neglecting, or at least leaving to second plan, the collection and study of mainland Portuguese specimens. This was, in part, left in the hands of amateur naturalists who would collect specimens and donate them to the nearest museum. This situation contributed to creating stronger collecting locality clusters, but still failed to aid the creation of time-series for species. The study of Portuguese fauna developed much later, especially from the 1970s to 1990s, when several teams associated with Portuguese universities started to dedicate their research efforts to several aspects of Portuguese mammalian species (Figs 3, 4).

The gaps and biases shown for Portuguese mammal collections have clear impacts on current and future scientific research, species conservation and teaching applications. Portuguese museums are poorly known by its own scientific community, creating a scientific chasm between museums and researchers. This disconnection begins in researchers’ early academic years, as students from the three universities with museums rarely visit or use the collections during their training. Like their international congeners, these institutions have been plagued by a lack of resources and trained personnel with knowledge of museum techniques and specimen preservation (Almaça 1993; Dalton 2003; Kemp 2015; Salvador and Cunha 2020), a situation strangely exacerbated by the fact that all museums belong to universities and mostly rely on them for funding. Ceríaco et al. (2021) and Santos et al. (2024) provided some hints on how being part of universities has negatively affected these collections. Vivo et al. (2014) had similarly reported that the university association is currently detrimental for the major Brazilian collections, as the working model needs to fit the needs of the collections and cannot be managed like faculty departments as they currently are. The disconnection may also stem from the fact that data are not easily accessible to researchers through museum online databases or international repositories like GBIF. For most Portuguese NHCs, mammal specimen data are not publicly available through accessible databases, with the sole exception of MUHNAC carnivores (Brites-Soares et al. 2022). On the contrary, Dunnum et al. (2018) reported that the 20 largest mammal collections in the Western Hemisphere are electronically databased and most are available via web data portals.

The absence of new accessions does not reflect the current landscape of studies on Portuguese mammals. In fact, recent years have seen great influx on publications regarding the phylogeography and biogeography of Portuguese mammals (Real et al. 2009; Vilaça et al. 2014; Boston et al. 2015; Barbosa et al. 2016; Bencatel et al. 2018; Querejeta and Castresana 2018; Queirós et al. 2019), phylogenetics (Paupério et al. 2012; Barros et al. 2016; Pires et al. 2017; Fletcher et al. 2019; Nóbrega et al. 2023), species conservation (Sarmento et al. 2008, 2017; Mathias et al. 2009; Palomares et al. 2010; Eggermann et al. 2011; Rocha et al. 2014b; Torres and Fonseca 2016; Silva et al. 2018; Encarnação et al. 2019; Proença-Ferreira et al. 2019; Barros et al. 2020; Torres et al. 2023), morphology and physiology (Oliveira et al. 2016; Rey et al. 2019; Pires et al. 2020; Ferreira et al. 2022), pathogen studies (Abrantes et al. 2013; Monterroso et al. 2016; Conceição-Neto et al. 2017; Carvalho et al. 2020; Abade dos Santos et al. 2022; Baptista et al. 2023; Lux et al. 2023; Santos-Silva et al. 2024) and even the resolution of long standing taxonomic problems and the discovery of new species (ICNB 2010; Paupério et al. 2012; Barros et al. 2015; Gallego et al. 2020; Rebelo et al. 2020). Bencatel et al. (2018) reported on the trends in mammalian carnivore research in Portugal, historically biased towards species of conservation concern and revealed knowledge gaps, in part due to the scattered and inaccessible nature of specimen data. Despite this plethora of studies, little to no specimens have been collected and deposited in Portuguese NHCs.

Very few of the studies mentioned above used specimens from the Portuguese NHCs and, more often than not, research was conducted without specimen collection, but rather from non-invasive sampling. As an example, bioacoustic surveys are an alternative technique in bat studies and were used by both Ferreira et al. (2022) and Nóbrega et al. (2023) to detect species in the Madeira and Porto Santo Islands. Additionally, Nóbrega et al. (2023) collected guano pellets and small wing tissue samples. Barbosa et al. (2016) conducted ear biopsies in live-caught and released Cabrera voles to study the species’ phylogeography, with additional sampling taken from barn owl pellets and a few museum specimens. Oliveira et al. (2016) captured live shrews to study how they coped with seasonal changes and released the specimens back into the wild after the completion of experimentation procedures.

Perhaps one of the most surprising results of this survey was the considerably low number of game species in Portuguese NHCs. With the exception of the red fox V. vulpes, the second most common overall, the remaining species were some of the least represented, especially the ungulates. The wild boar Sus scrofa is one of the most ubiquitous taxa in mainland Portugal and a staple big game species, yet is only represented by 16 specimens in the collections. Given the longstanding hunting tradition in Portugal, game species would, in theory, be some of the most abundantly found in museum collections. The early pre-scientific beginnings of Portuguese mammalogy can be traced back to the medieval ages, when publications on falconry and game species started to be produced in the country (Almaça 1985). Hunting was a privilege given by the Crown to the local nobility and some game species have been protected by law for centuries, going back as far as 1290, when royal decrees on the hunting of brown bears, hares, rabbits, red deer, roe deer and wild boars were published (Almaça 1985). Notably, two of the most common game species in the country, the Granada hare Lepus granatensis and the European rabbit Oryctolagus cuniculus, have a current conservation status of “Vulnerable”. European rabbit populations have been decimated by myxomatosis after the virus was introduced in France (Fenner and Ratcliffe 1965), with the disease becoming endemic to populations (Bertagnoli and Marchandeau 2015; Villafuerte et al. 2017) and, more recently, by the rabbit haemorrhagic disease (Kerr et al. 2009). Cases of myxomatosis spreading to the Granada hare have also been reported (Carvalho et al. 2020; Abade dos Santos et al. 2022). This led to an increase in pathogen research for both species, as it affects their conservation and that of their predators (Sobrino et al. 2008), including in Portugal (Delibes-Mateos et al. 2009; Abrantes et al. 2013; Monterroso et al. 2016; Carvalho et al. 2020; Abade dos Santos et al. 2022), yet these studies also did not contribute with new specimens for NHCs.

Specimens recovered from road-kill events or those that die while in the care of wildlife recovery centres could also be directed towards NHCs. Roads are features of modern-day landscapes with a known impact in species behaviour and biodiversity loss (Giulio et al. 2009), that has been the focus of several assessments for Portugal (Grilo et al. 2007; Carvalho and Mira 2011; Medinas et al. 2013; Carvalho et al. 2018; Silva et al. 2019). Grilo et al. (2007) surveyed some southern Portuguese national roads and highways bimonthly from July 2003 to December 2006 and recovered 801 carnivore specimens from eight species. This corresponds to more than half of the Carnivora specimens found in the assessed NHCs. Medinas et al. (2013) studied road effects on bat populations for a period of seven and a half months in southern Portugal and reported 154 road-killed bats from 11 species, including some rare ones like Barbastella barbastellus, a taxon not represented in any of the assessed collections. Although these specimens might not be found in the best preservation state to integrate into a collection, many of them could potentially have been salvaged and recovered to be accessioned to the country’s collections. To our knowledge, none of the specimens mentioned in the latter studies made their way to any NHCs.

Another source of specimens can be that of private collections and small museums. Bissoli-Silva et al. (2021) showed the relevance of regional collections, such as that of the Federal University of Espírito Santo, available in public databases, to better represent the biodiversity of a given region, the Atlantic Forest hotspot in Brazil. Similarly, Casas-Marce et al. (2012) presented evidence on the importance of private collections and small museums to obtain additional records to those available in larger institutions, concluding that, without access to such specimens, genetic sampling for conservation studies of threatened species like the Iberian lynx would have been biased. A similar point could be made for game species in Portuguese collections. Animal head trophies are traditionally kept from big game hunting and can be found displayed in local hunting associations, as well as traditional restaurants, especially in southern Portugal, while some high schools keep small natural history museums with several taxidermy mammal specimens (pers. obs.). Yet, these specimens are not “publicly available” as they would have been if they were in a NHC and no inventory of these exists. The current numbers of the country’s NHCs prevent population studies in these species and private collections like those mentioned above certainly account for more specimens than those extant in NHCs.

Internationally, the American Society of Mammalogists (ASM) has periodically done collection surveys to assess the status of both public and privately held collections from the Western Hemisphere, an effort done here for the first time for Portuguese collections, albeit only in public museums. The most recent assessment, published in 2018 (Dunnum et al. 2018), reported a decrease in annual growth rate, in line with the trend of the prior half century (McLean et al. 2016; Hope et al. 2018). Chiquito et al. (2021) provided a similar diagnosis for Brazilian mammal collections and found them to insufficiently represent the country’s biodiversity, despite the considerable number of housed specimens (ca. 372,200 specimens). It is worth noting that, according to the collections size classification used by Dunnum et al. (2018), the combined collections of the three main Portuguese NHCs would be considered a small collection (< 10,000 specimens). Dunnum et al. (2018) also reported 45 collections that were either closed or transferred to other institutions, the majority of which were university-based collections, in another example of the decreasing support for such resources in academic institutions (Gropp 2003; Schmidly 2005). Cook and Light (2019) classified some collections as “orphaned” due to the shifting of priorities regarding them, either by being created by individual researchers or being public collections unwanted or poorly supported by institutions, but argued that educating administrators on the importance of their collections can help reverse a poor decision (Bradley et al. 2012; Baker et al. 2014). Over half of the total specimens in the Western Hemisphere are held by the 20 largest collections and amongst those, half are university-associated museums (Dunnum et al. 2018). Unlike what has happened with Portuguese NHCs, some North American and Brazilian academic institutions like the Museum of Vertebrate Zoology (University of Kansas, Lawrence, USA), the University of Michigan (Ann Arbor, USA), the Museu Nacional (Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil) and the Museu de Zoologia (Universidade de São Paulo, São Paulo, Brazil) have found success in building large mammal collections and using them to foster graduate training programmes that prepared several new generations of mammalogists (Cook and Light 2019).

The present situation of Portuguese NHCs mirrors that of its international congeners, but in a much more drastic way. The current mammal collections and their historic shortcomings are conditioning fundamental research today and will condition research tomorrow. By failing to collect in the present, we are simply denying future generations the access to present day biological data, exacerbating the considerable time gaps that are already visible in Portuguese mammal collections. The observed gaps in knowledge will deepen if the collections stagnate and remain impoverished, something that is unfortunately expected due to the worldwide decline of specimen collecting and an apparent disconnection between the scientific community and Portuguese NHCs. One of the great plights of modern biological sciences is to deal with the changes brought on by human activities, including the on-going mass extinction of species (Ceballos et al. 2015). Traditional disciplines like taxonomy provide baseline conservation units in the form of species, but new approaches require past and present biodiversity data that are not being properly curated and collected (Wilson 2017). Fewer taxonomists are currently being trained, leading to a shortage of specialists which is causing a serious taxonomic impediment (Engel et al. 2021). In the same sense, modern studies in molecular genetics and phylogeography also depend on specimen derived data. This paper is not only a report on what can happen when institutions lose their main research and collecting missions, but also a call to attention for naturalists and museum professionals around the world.

An acknowledgment is owed to all collections managers and curators who provided access and data from the collections under their care: Ana Cristina Rufino from the Museu da Ciência da Universidade de Coimbra, Maria Judite Alves from the Museu Nacional de História Natural e da Ciência da Universidade de Lisboa and Francisco Calado from the Aquário Vasco da Gama. We also thank Riccardo Castiglia and Alexandra M. R. Bezerra for the commens provided and Maria João Santos for the small illustrations.