INTRODUCTION

Voucher specimens housed at biological collections and natural history museums are fundamental for the rigorous scientific study of biodiversity (Yates, 1985; ^{Nudds and Pettitt, 1997}; ^{Suarez and Tsutsui, 2004}; ^{Winston, 2007}; ^{Lavoie, 2013}; ^{Clemann et al., 2014}; ^{Holmes et al., 2016}). Biological collections of megadiverse developing countries have immensely contributed to the knowledge of life (^{Paknia et al., 2015}). For instance, the rate of species description (the sinequanum taxonomic practice that uses voucher specimens) in Colombia represents between 2.5 and 10 % of all the new species of terrestrial vertebrates described around the World (^{Arbeláez-Cortés, 2013a}). This result is particularly relevant for birds given that Colombia is the country with the largest number of bird species (^{Donegan et al., 2016}; ^{Avendaño et al., 2017}b), and there is an average of one new species described per year (Arbeláez-Cortés, 2013a; ^{Caycedo-Rosales et al., 2014}). The Colombian biological collections have a major role in the scientific study of the country's biodiversity (Arbeláez-Cortés et al., 2017), and several gaps in current knowledge (e.g., genetic diversity) can be filled up through collection of new specimens and by analyzing voucher material already catalogued in available collections (Arbeláez-Cortés, 2013b; Avendaño et al., 2017a).

The recognized value of biological collections for scientific knowledge of biodiversity contrasts with the scarce information published on the monetary costs involved in obtaining voucher specimens (Yates, 1985; ^{Blackmore et al., 1997}; ^{Mann, 1997}; ^{Bradley et al., 2012}; ^{Baker et al., 2014}; Bradley et al., 2014). Such costs can be easily quantified for field expeditions of funded projects. Knowing the monetary costs of collecting and curating voucher specimens can be instrumental for institutions and government to evaluate the effective use of the allocated budget to activities related to biodiversity documentation. It is particularly crucial for Colombia, a country that only invests between 0.24 and 0.3 % of its GDP in research and development (The World Bank, 2018; UNESCO, 2018), but which houses large biodiversity (Sistema de Información sobre Biodiversidad de Colombia [SiB-Colombia], 2018).

In 2016, the Colombian government began a national initiative known as Colombia BIO that included projects to conduct 20 expeditions to inventory biodiversity in unexplored localities. The program helped to articulate academic and scientific institutions, consolidating biological collections of specimens and tissues for molecular analyses, and making publicly available the data through online databases (^{Ayala et al., 2018}). The model set by Colombia Bio expeditions was followed at a department level (i.e., first administrative division of Colombia) in Santander (in northeastern Colombia) financed by funds from Sistema Nacional de Regalias with the management of Gobernación de Santander, and was conducted linking the scientific capacities from Universidad Industrial de Santander (hereafter UIS) and Instituto de investigación de recursos biológicos Alexander von Humboldt (hereafter IAvH). The project had the name "Investigación de la biodiversidad y los servicios ecosistémicos para la gestión integral del territorio - descubriendo los ecosistemas estratégicos para el fortalecimiento de la gobernanza en el departamento de Santander" (hereafter Santander-BIO). UIS aimed to obtain new biological material to consolidate its collections of vertebrates, entomology, herbarium, hydrobiology, microbiology, and tissue samples for molecular analyses of some taxa. We developed the collection of bird voucher specimens and their tissues, along with occasional observational records of some species.

Different scientific reasons support the necessity of continuing collection of birds in Colombia (^{Cuervo et al., 2006}). During the last decades several ornithological expeditions contributing with voucher specimens to biological collections have been conducted in the country (^{Stiles and Bohórquez, 2000}; Bohórquez, 2002; ^{Alvarez et al., 2003}; ^{Donegan et al., 2007}; Cuervo et al., 2008a; Cuervo et al., 2008b; Donegan et al., 2010; ^{López-Ordóñez et al., 2013}; Stiles and Beckers, 2016; ^{Izquierdo et al., 2017}; ^{Renjifo et al., 2017}; Stiles and Naranjo, 2017; ^{Avendaño et al., 2018b}; Avendaño et al., 2018a; ^{Córdoba-Córdoba and Sierra, 2018}), with some of them at Santander. The scientific documentation of Santander bird fauna encompasses more than 200 years (Avendaño, 2017), and at least 840 species have been recorded in this department according to eBird records (^{Rondón, 2017}). Public information indicates that there are at least 5299 voucher specimens, and around 100 tissue samples, representing 648 species for Santander (^{Arbeláez-Cortés et al., 2015}; GBIF.org, 2018), but these are not yet representative of the bird biodiversity of this Colombian region.

The aims of our study were: 1) to present results of three ornithological expeditions conducted in Santander, Colombia, during 2018 to collect and preserve bird voucher specimens by the UIS in the framework of Santander-BIO project, commenting topics of the biology of some taxa and discussing results in the historical context of bird collection in Colombia in general and of Santander in particular; and 2) to quantify the monetary costs related to the bird specimens obtained, as an example of the costs during regional expeditions in Colombia, depicting some features related to collecting methods.

MATERIALS AND METHODS

Study area and field work

Santander is located on the western slope of the Colombian Eastern Andes (between 72.28° W and 74.31° W, and between 5.42° N and 8.08° N), occupying 2.7 % of the national territory (30 537 Km²), and being the fourth department regarding population and economy in the country (Gobernación de Santander, 2016). There are two major physiographic units in Santander: 1) the Middle Magdalena valley (40 to 200 m.a.s.l) to the West that is characterized by a flat and undulated topography, and 2) the Colombian Eastern Andes (i.e., Cordillera Oriental) which occupies most of the department, and is characterized by a steep topography that reach up to 4345 m.a.s.l. This heterogeneous region includes orographic accidents like the Serranía de los Yariguíes, river canyons such as Chicamocha canyon, and plateaus like the Mesa de Los Santos or Jéridas (Fig. 1). The annual median temperature and precipitation vary locally from less than 8 °C to more than 28 °C and from 500 to 3000 mm/year (IDEAM, 2015). According to Holdridge classification, there are 15 life zones in Santander, including tropical humid, very humid forest, dry forest, pre-montane dry forest, very humid montane forest, and pluvial montane forest (Instituto Geográfico Agustín Codazzi, 1976). The forests of Santander are highly fragmented, where only 16.3 % of its area remains covered by natural forests (mainly secondary forests). Agricultural plantations of palm, coffee, and cacao crops are in the rural areas, as well as extensive cattle ranches, which cover approximately 60.2 % of Santander's area (Secretaría de planeación de Santander, 2014). Areas that are still ecologically conserved are mainly located within natural reserves, including a national natural park (Parque Nacional Natural Serranía de Los Yariguíes) encompassing around 59 000 ha (Secretaría de Planeación de Santander and Universidad Industrial de Santander, 2011).

Figure 1: a) Map of Santander department depicting some major orographic accidents and rivers. Municipalities where expeditions were conducted are indicated, as well as the capital city Bucaramanga (black star) The inset shows the place of Santander in Colombia. b) Landscape views of the forests where collections were conducted. From left to right: Cimitarra, El Carmen de Chucurí, and Santa Bárbara. 

In the framework of Santander-Bio project, we conducted three expeditions during February-March, July, and August of 2018, developing fieldwork focused on collecting birds.

Each expedition aimed to represent a different ecosystem of Santander. We named each expedition according to the name of the municipality (i.e., second-order politic and administrative division of Colombia) where the fieldwork was conducted and avoid referring to them as localities because in two expeditions, we performed collection of birds in more than one locality within the zone (Table 1). The expeditions were conducted at the municipalities of Cimitarra (humid tropical forest), El Carmen de Chucurí (very humid pre-montane forest), and Santa Barbara (pluvial montane forest), which had scarce biological records according to biodiversity databases.

Table 1 Expeditions of Santander-Bio. Geographic information and sampling effort are indicated for each locality where bird voucher specimens were collected during 2018 in three municipalities of Santander. 

Each expedition was two weeks long, with effective fieldwork of12 to 13 days. We focused our efforts on collecting mainly resident bird species inhabiting the interior of forests. Therefore, for each expedition, we placed ten to 13 mist nets (12 x 2.5 m) in the interior of the forests located around each campsite. Mist nets were mainly open at dawn (5:30 - 6:30 am) and closed before noon (10:00 -11:00 am), but some days mist nets were also opened at noon and during the afternoon (2:00 - 5:00 pm). The collection effort was calculated considering one mist net hour as a mist net of 12 m open for one hour (Table 1). In parallel, a fieldwork assistant with military training helped us conducting an occasional and opportunistic collection of birds around campsites during two expeditions by using an airgun pistol Crossman 5.5. Finally, on two opportunities, we recovered salvaged specimens (i.e., individuals dead by collision with human structures that were preserved as specimens). Details about the expeditions, the localities we worked, and the sampling effort are depicted in Table 1. We collected at least one specimen for each species captured except for the migrant Swainson's Thrush (Catharus ustulatus) of which we captured and released one individual at El Carmen de Chucurí. This project was authorized by Comité de Etica UIS (CEINCI) in the Acta N° 07 27 April 2018.

Two of us scientifically skinned birds in the field, to prepare voucher specimens; and for the majority of them we obtained and registered the standard biological information (i.e., body mass using a pocket electronic scale, size of the gonads, skull ossification, molt, brood patch, cloacal protuberance, presence of ocular ring, as well as the color of the iris, the bill and the tarsus). For some specimens, a few data were not taken because gonads were not found or data of molt, skull, or brood patch were forgotten to be registered. For some specimens, we preserved in diluted formol the stomach and the gonads. We also collected tissue samples of muscle, liver, and heart for every specimen, which were preserved in ethanol 96 % during the field work and then, after removing ethanol in laboratory, were cryo-preserved in a freezer at -80 °C as part of a new collection (UIS-CT= Colección de Tejidos del Museo de Historia Natural de la Universidad Industrial de Santander, Bucaramanga). Once in the bird collection (UIS-AV= Colección de Aves del Museo de Historia Natural de la Universidad Industrial de Santander, Bucaramanga) specimens were curated and their data digitalized following a format based on the Darwin Core standard (Biodiversity Information Standards TDWG, 2018). For taxa represented by several voucher specimens, some were cataloged at UIS-AV, and the remaining were exchanged for other specimens with the IAvH. To catalog specimens here, we used the catalog number of UIS-AV for most of them and the field number (consecutive SBIO-number that is attached in a label to each specimen) for the ones that were exchanged with IAvH.

Besides the collection of voucher specimens during the three expeditions, we also obtained occasional records of bird species based on visual and acoustic observations (50 observation hours), to complement our species list (Appendix) and discuss some issues about them. Therefore, these observational records were not included in our main analyses. We followed the taxonomy of ^{Gill and Donsker (2018)}.

Analysis of the monetary costs of collecting voucher specimens

As noted above, we had the aim of quantifying the monetary costs related to the bird specimens obtained. For this, we used the detailed budget approved for Santander-Bio project that includes the costs of materials and payments for services of every activity. To measure the costs involved in bird collection and curation, we calculated three kinds of costs but considering only a fraction of the allocated budget that we estimated was directly related to the collection and curation of bird specimens. First, we calculated the human resources payment by adding the budget allowance for activities related to bird specimens collection and curation such as 1) a fraction of the payment for senior and junior researchers in charge of field and biological collection work, 2) a fraction of the payment for a field assistant, 3) a fraction of payment for UIS students who helped with curational activities at UIS-AV, 4) a fraction of payment for other UIS students who helped with the digitalization of data. Second, we included a percentage of the costs of materials (i.e., mist nets and cryovials). The mentioned budget fraction was between 1/8 and 7/8 of the budget allocated for particular issues (see Table 2), considering that both the payment for human resources and the materials used were assigned to other activities related to the project and not just to obtain and curate the bird specimens. For instance, the total budget allocated to pay a junior researcher (Biologist) was around 23 500 000, but the payment included other activities not related to direct bird collection and curation (i.e., writing reports), and then we estimated the investment in specimens as 7/8 of such budget. The same was considered for the other human resources payments and the use of material. Third, we added our costs of transport, alimentation, and lodging during expeditions; and such value was considered as the cost of fieldwork logistics. The sum of the three kinds of costs divided by the number of specimens collected is the measure of the total cost of obtaining a bird voucher specimen and their tissues for molecular analyses, during an expedition of UIS in Santander-Bio project, as well as to place these voucher specimens and their tissues in a biological collection with curated and digitalized information ready to be publicly available. The costs were calculated in Colombian pesos (COP) as was the original budget but were converted to US dollars considering a change rate of $3249.75 COP = US$1 US, according to representative market rate from 31/12/2018.

Table 2 Costs of collecting and curating 300 bird specimens during expeditions of Santander-Bio. Costs are in Colombian pesos. The activities related with collecting and curation includes tissue samples and data digitalization. The "fraction" approximates the proportion of the specific budget invested for services or materials related directly with bird collection and curation. 

RESULTS

Bird records and voucher specimen collection

We captured 403 bird individuals, of which we collected 300 voucher specimens and their tissues, after a total sampling effort of 1290 mist net hours plus opportunistic collection using airgun (Table 1). These voucher specimens represent 117 species from 30 families and 12 orders (Appendix); the occasional records added 120 species and 19 families (Appendix). Therefore, our dataset includes 237 species from 49 bird families. The families represented by more than ten voucher specimens were: Trochilidae (83 specimens), Tyrannidae (35), Furnariidae (34), Thraupidae (28), Parulidae (16), Emberizidae (15), Thamnophilidae (14), and Pipridae (10). 62 species (53 %) were represented by two or more voucher specimens.

The number of species that we collected was 50, 26, and 49 for the expeditions to El Carmen de Chucurí, Cimitarra, and Santa Barbara, respectively. Eight species (6.8 %) were recorded in more than one expedition (Appendix): Green-fronted Lancebill (Doryfera ludovicae), Speckled Hummingbird (Adelomyia melanogenys), Olive-striped Flycatcher (Mionectes olivaceus), White-breasted Wood-Wren (Henicorhina leucosticta), Chestnut-capped Brushfinch (Arremon brunneinucha), and Three-striped Warbler (Basileuterus tristriatus) collected in Santa Bárbara and El Carmen de Chucurí, as well as Wedge-billed Woodcreeper (Glyphorynchus spirurus) and Ochre-bellied Flycatcher (Mionectes oleagineus) collected in Santa Barbara and Cimitarra. We obtained observational records of twelve species during one expedition, which were collected in a subsequent expedition (Appendix). The use of the airgun retrieved 13 species (11 %), of which only three species were also collected by mist nets (i.e., Whooping Motmot [Momotus subrufescens], Black-chested Jay [Cyanocorax affinis], and Blackburnian Warbler [Setophaga fusca]). The species observed but not collected were mainly from families of water birds, pigeons, raptors, swifts, and parrots (Appendix). Because of their biology (i.e., size, habitat, foraging strategy, or behavior), such species are difficult to capture using mist nets.

The number of families collected varied among expeditions (Fig. 2), being the expedition to El Carmen de Chucurí, the one providing the largest number (24). There was variation in the number of species collected per family (Fig. 2), but a large proportion of species captured during the three expeditions were from the Tyrannidae and Furnariidae, each one of them representing more than 11 % of the species collected in every expedition. It is noteworthy that Thraupidae represented a large proportion of the species collected in Cimitarra and Santa Barbara (15 to 20 %), and Trochilidae encompassed a large proportion (17 to 20 %) of the species in El Carmen de Chururí and Santa Barbara, but was just represented by one species (Pale-bellied Hermit [Phaethornis anthophilus], UIS-AV-2197) in Cimitarra. The capture rate also varied among expeditions (X²= 123, d.f. = 2, p < 0.05) being almost one order of magnitude lower in Cimitarra (0.08 individuals per mist net hour) than in the other two expeditions to El Carmen de Chucurí (0.3) and Santa Barbara (0.5).

Figure 2 The proportion of the number of species per Family collected during the three expeditions of Santander BIO 2018, Santander, Colombia. The category "other" includes families represented by less than seven species. 

The comparison of our ornithological collection with the historical list of bird specimens from Santander allowed recognizing that the 300 voucher specimens collected by UIS during Santander-BIO expeditions are the largest series obtained during the last five decades. Since 1965 no more than 250 specimens were collected per year, which marks the Santander-Bio project as an important contribution to the enrichment of the bird fauna collections from Santander. Along with the history of the ornithological collections in Santander, which began around 150 years ago according to GBIF records, there are only three years (1949, 1961 and 1964) where more than 300 voucher specimens were obtained. Such collections corresponded to expeditions conducted by the Smithsonian National Museum of Natural History U.S.A (1949), the Yale Peabody Museum of Natural History U.S.A (1961), the Museum of Vertebrate Zoology UCLA at Berkeley U.S.A (1964), and the Los Angeles County Natural History Museum U.S.A (1961 and 1964). The historical list of Santander voucher specimens also allowed identifying that 60 % of them are housed in foreign biological collections. It is only since 1970 that the Colombian biological collections began to be relevant in the documentation of bird diversity from Santander; currently, the two institutions with more voucher specimen records of Santander birds (GBIF.org, 2018) are UIS-AV at Bucaramanga and IAvH-A at Villa de Leyva (previously INDERENA). Comparing the historical list of voucher specimens from Santander with our data, we found that 16 species collected by UIS during Santander-BIO expeditions were collected for the first time by a voucher specimen for Santander (Appendix).

DISCUSSION

Historical ornithological expeditions in Colombia conducted by foreign researchers during the end of 1800's and beginnings of the 1900's provided thousands of specimens and hundreds of taxa records (e.g.,^{Allen, 1900}; ^{Chapman, 1917}; ^{Friedmann, 1947}), indicating that the country was one of the richest around the World. The current recognition of the Colombian bird fauna as the richest one is ultimately based on voucher specimens. The basic documentation of Colombian avian biodiversity has continued through collection during some expeditions and by the description of several new species, with a mean of one new bird species per year since 1990 (^{Córdoba-Córdoba, 2009}; ^{Arbeláez-Cortés, 2013a}).

During the last decades, several ornithological expeditions have documented bird biodiversity in different regions of Colombia, but some of these publications report occasional collection of voucher specimens. For instance: incidentally dead during captures in mist nets or specimens representing taxa considered noteworthy because represents range extensions (^{Bohorquez, 2002}; ^{Donegan et al., 2007}; Donegan et al., 2010; ^{Avendaño et al., 2018a}; b). One of these works included a large series of specimens for Santander (Donegan et al., 2010). There are also publications documenting results of systematic collection of bird specimens (^{Haffer, 1986}; ^{Stiles and Bohórquez, 2000}; ^{Alvarez et al., 2003}; ^{Cuervo et al., 2008a}; b ; ^{López-Ordóñez et al., 2013}; Stiles and Beckers, 2016; ^{Renjifo et al., 2017}; Stiles and Naranjo, 2017) for several departments (i.e., Antioquia, Boyacá, Caquetá, Cauca, Cesar, Chocó, Guainia, Guaviare, and Nariño), but not for Santander. The results of these publications indicated the collection of 25 to 287 specimens, representing between 25 and 150 species. Therefore, both the number of specimens and species we reported here during Santander-BIO expeditions are among the larger for recent ornithological expeditions in Colombia. As we noted above, the voucher specimen series obtained by UIS during Santander-Bio is one of the larger ones of the historical ornithological collections in Santander.

In Santander, there are several localities with ornithological records (^{Riaño, 2017}), and there are records for around 840 potential bird species (^{Rondón, 2017}). In this context, the total number of species we presented for 14 localities, at three municipalities, represented 28.2 % for those recorded in the department of Santander. Considering only records based on voucher specimens, we reported almost 14 % of the total number of the potential bird species for Santander. However, the number of Santander bird species represented by vouchered specimens in biological collections probably is around 650 (GBIF.org, 2018), and our collection series is ~ 20 % of such figure. We report for the first time a voucher specimen for 16 species in Santander. Among them we highlight the records of four species:

- The Red-legged Tinamou (Crypturellus erythropus) is a tinamou with a wide geographical range found in lowlands from Northern Colombia to Brazil, but it could be locally uncommon at some places (^{Hilty and Brown, 1986}; Hilty, 2003; ^{Restall et al., 2006}). Our voucher specimen is 230 Km from the closest Colombian voucher specimen from "Ayacucho, Cesar" (USNM 372312), and considering audio records this species has a few in Cesar (e.g., Xeno-canto, 2018, 45340, 45342) and to the North at La Guajira (e.g. Xeno-canto, 2018, 92405, 92576).

- The Cinnamon Screech Owl (Megascops petersoni) is a scarce owl that inhabits inside forests and is poorly known in Peru, Ecuador, and Colombia (^{Ridgely and Greenfield, 2001}; ^{Schulenberg et al., 2007}; ^{Cuervo et al., 2008b}). Our record (UIS-AV-2180) from El Carmen de Chucurí is one of the few voucher specimens for Colombia. Cuervo et al. (2008b) reported the first one in Anorí, Antioquia (ICN-34377), and commented on the existence of an old "Bogotá" skin. We also found, at UIS-AV, an uncatalogued specimen of this taxon without a label and from unknown origin, and IAvH-A reported another specimen from Cueva de Los Guacharos, Huila (IAvH-A 12391). The locality of UIS-AV-2180 is 190 Km from the nearest locality, but very close to an audio record (Xeno-canto, 2018, 143842) of this species from San Vicente de Chucurí, Santander.

- The Saffron-headed Parrot (Pyrilia pyrilia) is an uncommon local parrot in Colombia, which inhabits humid lowland forests (^{Hilty and Brown, 1986}; ^{Restall et al., 2006}). The localities of UIS-AV-1944 and IAvH-A (SBIO-57) are 120 Km from the nearest locality with a voucher specimen: Puerto Berrio, Antioquia (AMNH 154431), and farther from audio records in Antioquia (e.g., BSA 19070, 30144).

- The Chestnut-crowned Gnateater (Conopophaga castaneiceps) is a local uncommon species that inhabits the undergrowth of montane forests in the Andes, from Colombia to Peru (^{Ridgely and Tudor, 1994}). The locality of UIS-AV-1900, 2187, and IAvH-A (SBIO-62) is 138 Km from the nearest locality with a voucher specimen: Otanche, Boyacá (ICN 32833), but very close to an audio record (Xeno-canto, 2018, 401285) of this species from San Vicente de Chucurí, Santander.

Other relevant records were those for geographically restricted taxa or species of conservation concern such as Black Inca (Coeligena prunellei), White-mantled Barbet (Capito hypoleucus), Parker's Antbird (Cercomacroides parkeri), Ochre-breasted Antpitta (Grallaricula flavirostris), and Sooty Ant-Tanager (Habiagutturalis) (Appendix).

Reproductive periods of Neotropical birds expand several months, which are related to the abundance of food resources but are not well defined for several species (^{Echeverry-Galvis and Córdoba-Córdoba, 2008}). We collected juveniles and adults with physical indication of breeding condition, allowing us to determine that 70 - 80 % of the species from Cimitarra and Santa Barbara agreed with the breeding months previously reported, contrasting with El Carmen de Chucurí where only 50 % of the species agreed with published breeding periods (^{Hilty and Brown, 1986}; ^{Fjeldså and Krabbe, 1990}; ^{Del Hoyo et al., 2003}). Collecting detailed breeding data (^{Foster, 1975}; Echeverry-Galvis and Córdoba-Córdoba, 2008) must be considered in further expeditions aimed to obtain voucher specimens of Colombian birds to enhance the biological knowledge.

Regarding the mist net capture rate calculated from the data available in some publications for Colombian ecosystems (^{Stiles and Roselli, 1998}; Stiles and Bohórquez, 2000; ^{Polanco et al., 2015}; Stiles and Beckers, 2016; ^{Renjifo et al., 2017}; Stiles and Naranjo, 2017; ^{Córdoba-Córdoba and Sierra, 2018}), we found results that indicate a rate between 0.14 and 0.77 individuals captured per mist net hour (mean = 0.37, n = 7). Therefore, the capture rate we obtained during our expeditions to Carmen de Chucurí and Santa Barbara is within this range; however, the capture rate for Cimitarra (0.08) is very low and represents an issue for further study.

We acknowledge the bias of documenting bird diversity using only mist nets to document the whole bird fauna of a locality because they usually represent between 50 and 75 % of the species (^{Stiles and Roselli, 1998}; ^{Polanco et al., 2015}). Occasional observational records allow reporting a large proportion of species that are not easily captured by mist nets. Besides, for a montane forest in eastern Colombian Andes small birds (< 50 gr) are overrepresented in mist net captures (Stiles and Roselli, 1998). Here we obtained a similar result and found that collecting using airgun added species of larger sizes to the voucher specimen series. The use or airguns has been implemented in other expeditions in Colombia (Stiles and Beckers, 2016; ^{Renjifo et al., 2017}; ^{Avendaño et al., 2018a}), and we consider that their use, by qualified personnel, during ornithological expeditions can be implemented in further expeditions to optimize the documentation of bird taxa through voucher specimens.

Regarding the monetary costs published for the collection of voucher specimens, we found that one specimen costs between $17.8 and $767 US dollars for regional expeditions in other countries (Yates, 1985; ^{Blackmore et al., 1997}; ^{Mann, 1997}; ^{Bradley et al., 2012}; ^{Baker et al., 2014}; Bradley et al., 2014). The lower and higher costs are reported by Bradley et al. (2014), who measured monetary investment on mammals collected during 50 regional field trips. These authors included costs related to logistic and human resources services, but they did not include costs of materials. All reviewed studies reported costs for mammal or herbarium collections, and most of them calculated a mean of costs from multiple expeditions for several years. There is ample variation among different expeditions of the same institution (Bradley et al., 2012). We did not find published information on costs of bird collection, but the cost we calculated is within the range reported for mammals and plants. Institutions can check the performance of projects aimed at documenting biodiversity through voucher specimens by asking for estimation costs per specimen. Besides, these institutions and the government must consider these costs as an investment in scientific capital because these voucher specimens will render biological information for decades, or centuries, returning a large value in the form of scientific knowledge about the region or the country (Yates, 1985; Arbeláez-Cortés et al., 2017).

At this point, it is necessary to consider some issues regarding two possible ways to lowering collection costs. If we had collected the 403 bird individuals captured during the three expeditions, the cost per specimen could have decreased a 25 %. However, the kind and number of specimens collected per day was a tradeoff between the scientific relevance and the number of specimens that two of us (DV-E and EA-C) were able to process properly. For instance, this was the reason why we released a migrant species we captured in Carmen de Chucurí (i.e., we considered of more scientific relevance the resident birds captured that day). It is obvious that a larger series of specimens collected can decrease the costs per specimen, but researchers must always take rapid decisions in the field on what material to collect considering both scientific value and logistic issues. Probably, specimen numbers could be increased in further projects by including students as part of expeditions. It is because students are less costly as human resources and can increase the number of specimens processed per day. However, in the case of Santander-BIO project, students cannot participate in fieldwork but contributed to curatorial activities at UIS-AV. Besides, proper scientific skinning of birds seems to be a scarce competence among Colombian biology students, making it necessary to invest time to teach them.

Another issue that is necessary to comment on is that after collecting, curating, and cataloging specimens at UIS-AV, there are other costs involved in storing such material in the long term. For instance, ^{Blackmore et al. (1997)} consider such cost in their valuation of collecting (or obtaining specimens by other methods), curating, and storing material at the London Natural History Museum. We did not have a measure of such post-curational costs at UIS-AV, but they could be calculated considering three issues: 1) cost of mobiliary for storing specimens, 2) cost of the area (i.e., appropriate facilities) where UIS-AV is placed, and 3) costs of services, including scientific curation (usually as time of a curator and auxiliaries), costs of fumigate the collection, repair of some equipment, and electricity associated with the functioning of equipment for collection maintenance. The cost number 2 is difficult to calculate but could be approximated by costs per m² for rents (i.e., per year) or construction of habitational spaces in the same city and near the collection. The costs per specimen in this case also will diminish while the number of cataloged specimen's increases. Therefore, maintaining an active collection of specimens is a way to optimize the resources allocated to science by institutions and governments.

CONCLUSIONS

We conclude that it is possible to quantify the costs for the collection and curation of scientific specimens, considering the budget directly assigned to these activities. However, we cannot give a precise protocol to measure the cost of a specimen in other expeditions, but we highlight that any calculation of such cost must consider, besides the logistic costs of fieldwork, the human resources costs associated with time invested in direct collecting and curatorial work. The bird specimens obtained in Santander-BIO expeditions rendered the basic information of the birds inhabiting three ecosystems of this Colombian region, but also will allow further studies (e.g., analyses of genetic and phenotypic variation). Therefore, the costs involved in obtaining such material must be considered as an investment in scientific capital, and not as expenditure. We also conclude that voucher specimen collection based on mist nets can be complemented using airguns, or even low caliber shotguns, to have a representative series of the local avifauna. Finally, we recommend to biologists to optimize monetary resources allocated by the government to explore Colombian biodiversity developing, or improving, skills (e.g., skinning, proper capture and manipulation of birds in mist nets, proper sampling and storing of tissues for molecular analyses) which allow collecting large series of voucher specimens to fill scientific knowledge gaps, which will allow having a clear picture of the country biota.