The assessment of the conservation status of species is an essential preliminary step in wildlife management and should largely be based on accurate scientific knowledge of trends in their distribution ranges and/or overall abundances (Dayton 2003, Groves et al. 2002). It enables the prioritization of resources and actions in biodiversity conservation (Wilson et al. 2009, Arponen 2012). While the coverage of formal assessments of conservation status by the IUCN may be still insufficient for several invertebrate, plant and fungi groups, vertebrates have a much better assessment level, birds being the first group to achieve a complete evaluation of extinction risk for all known species (IUCN 2022). However, even for these groups, it is not always possible to achieve sound assessments, due to the lack of both scientific knowledge and good field data (Papeş & Gaubert 2007, Martin & Molina 2013, Rueda-Cediel et al. 2018). Bird groups with very secretive habits are particularly difficult to study. This is particularly apparent in species for which field data are very difficult to obtain (Drew & Collazo 2012) as is the case for some nocturnal, marsh or forest bird species such as owls (Strigidae), crakes and rails (Rallidae), pittas (Pittidae) or tapaculos (Rhinocryptidae), resulting in classification as Data Deficient (Butchart & Bird 2010). In this sense, the buttonquail (Turnicidae) family is a good example of a group of very cryptic species, which very often pass unnoticed to bird observers. Buttonquails are one of the most understudied bird families (Yarwood et al. 2019, Winkler et al. 2020), whose unobtrusive behaviour makes them truly difficult to detect by both sight and voice being most notably detected by their tracks and other signs of activity (Gutiérrez-Expósito et al. 2011, Lees & Smith 1998). An exhaustive review of specific bibliographic references regarding all buttonquail species resulted in only 184 items, published between 1844 and 2019, in which the main study subject was a buttonquail species, of which only 104 were field-based studies (Gutiérrez-Expósito 2020). Only 78 of these publications are peer-reviewed articles and were mostly focused on endangered species found in economically developed areas: South Africa (Fynbos Buttonquail Turnix hottentottus), Europe (Common Buttonquail T. sylvaticus) or Australia (Black-breasted Buttonquail T. melanogaster), while most of the basic biological knowledge remains virtually unknown for most of them, even for some still common and widespread species (Gutiérrez-Expósito 2020).

Following del Hoyo & Collar (2014), the family Turnicidae is composed of 18 bird species, all belonging to the genus Turnix, except the Quail-plover Ortyxelos meiffreni. The Turnicidae is an Old-World bird family with distribution ranging from the West Palearctic and Sub-Saharan Africa to Australia and New Caledonia through South and East Asia and the islands of Malaysia, Philippines and Indonesia. Turnicidae are strictly ground dwellers found mainly in grassland, steppe and low shrub land, with some inhabiting low understory of wet and sclerophyll subtropical and tropical forests (Madge & McGowan 2002, del Hoyo & Collar 2014). While most of the species are monotypic or consist of two subspecies, three have much greater subspecific variation, most of them being island endemics (Table 3; del Hoyo & Collar 2014). Most of the species are listed as Least Concern in the IUCN Red List. However almost all other categories can be found within the family, with Black-breasted Buttonquail and Sumba Buttonquail T. everetti listed as Vulnerable, Buff-breasted T. olivii and New Caledonia Buttonquail T. novaecaledoniae as Critically Endangered and Luzon Buttonquail T. worcesteri as Data Deficient (del Hoyo & Collar 2014, IUCN 2022).

In this context, citizen science platforms can provide an opportunity to overcome the scarcity of scientific-based information on buttonquails and other poorly known bird groups through the participation of a huge number of observers reporting worldwide (Wiersma 2010, Wood et al. 2011). Among others, eBird, developed by the Cornell Lab of Ornithology, is the largest worldwide citizen science birding platform (Sullivan et al. 2014). These data are based on the creation of birding lists for a given location and date (hereafter checklist). Every checklist gives the option to record the number of individuals or just presence under different protocols. Up to July 2022, eBird has nearly 8 hundred thousand users who have contributed to this database with nearly 70 million complete checklists around the globe, which are reviewed by local expert reviewers, ensuring the correct identifications of birds listed by users.

In this paper, we will use the elusive buttonquails as a study case to provide a simple index to prioritize the study and conservation actions for all species and subspecies. This is performed through the combination of available distribution information already composed by BirdLife International and the field data provided by the eBird citizen science platform. This concern index has been estimated for every taxon as a combination of two parameters: (1) the scarcity of the taxon and (2) the uncertainty of the available data for the taxon, while each parameter is based on two components respectively. Relative range size and frequency of occurrence have been used as an approximation to the scarcity of the taxon, while range accuracy and eBird effort have been used to calculate the uncertainty parameter.

Table 1.

Concern index parameters composition for each taxon with the definition of its respective components.

The cryptic nature of buttonquails, the lack of scientific knowledge together with the small number of species with a high subspecific variation with all IUCN Red List categories represented, make the Turnicidae family a very good model avian group to evaluate a synthetic approach like this. Although using the same principles as those used in IUCN Red List assessments (IUCN Standards and Petiotions Committee 2022), the resulting index is based on a single dynamic continuously updated public database which has the potential to allow to prioritize the conservation and research needs not only for every bird family or group but also for lower taxonomical levels such as subspecies or even ecological units such as populations.

METHODS

We obtained data (hereafter observations) for all buttonquail species from the eBird platform, together with a global database of effort events (eBird 2019) from 1913 to May 2019 (eBird 2019). Buttonquails are terrestrial birds, so we used all eBird data available except those logged under the pelagic protocol (Sullivan et al. 2009), which means complete checklists, partial checklists and incidental observations. With these data, we created a spatially explicit point GIS layer file for every buttonquail species. Even when common, buttonquails are seldom seen so any sighting is highly appreciated by birdwatchers. Therefore we assumed that all buttonquails are equally difficult to observe and that they are always registered when an encounter with any of the species takes place. Consequently, we assumed that whenever buttonquails were not mentioned in a checklist the bird had not been spotted and thus, we treated those checklists as buttonquail absences. We used the spatially explicit shape files of buttonquail distribution maps provided by BirdLife International and Handbook of the Birds of the World (2018) to define the study areas for every buttonquail species. We calculated its range size, as the BirdLife extent of occurrence in km², by using the ‘area’ algorithm of QGIS in the World Cylindrical Equal Area coordinates reference system.

With these data, we created four components to be paired to obtain two parameters, which will be the base for the calculation of the concern index. As a proxy for species abundance, we named the first parameter ‘scarcity’ which is a combination of the relative range size of the taxon and its frequency of occurrence in the eBird data set, while as an approximation to the accuracy of the existing knowledge, we called the second parameter ‘uncertainty’, which results from the combination of the accuracy of its distribution area, hereafter range accuracy and the eBird survey effort, hereafter eBird effort.

The relative range value results from scaling the extension of the range sizes (km²) of the species set, between 0 and 1, by min-max normalization, as:

where x is the range size of a given species, m is the range size of the most restricted species and M is the range size of the most widely distributed species. The frequency of occurrence is the proportion of eBird observations of a given species out of all the checklists made within its BirdLife distribution area.

We calculated the range accuracy of the distribution areas as the proportion of all observations for a given species which lay within its BirdLife distribution area. Finally, we calculated the eBird effort as the checklist density within the BirdLife distribution area.

The same process was done when assessing at subspecific level, but taking into account all 47 described races of every buttonquail species, plus the 12 monotypic species (del Hoyo & Collar 2014). We used the subspecific distribution areas described by del Hoyo & Collar (2014), to split the specific BirdLife distribution areas into subspecific spatial files to calculate their respective range sizes.

Each parameter has been calculated as the geometric mean of its correspondent components. To avoid the skewness of results we used as parameters the absolute values resulting from a logarithmic transformation. To allow operations, all zero and zero/zero values have been approximated to 10^–10. Finally, in a bidimensional Euclidean space defined by the uncertainty and scarcity parameters, we calculated, for every taxon, the concern index as the distance to the origin from the point defined by its respective uncertainty and scarcity values as defined by the Pythagoras theorem as:

The same data was extracted and calculated for every recognized buttonquail subspecies except for those of Barred Buttonquail T. suscitator in mainland Asia for which a clear barrier between subspecies cannot be defined. Consequently, the Indian Subcontinent subspecies taigoor and bengalensis have been treated together and the SE Asia races blackistoni, pallescens, thai and plumbiceps have also been treated jointly.

Excluding Data Deficient species, we assigned each IUCN category with a natural number, with lower values for non-endangered species and higher to critically endangered ones, where Least Concern = 1, Near Threatened = 2, Vulnerable = 3, Endangered = 4 and Critically Endangered = 5 (IUCN 2022). To validate our results, we did a Spearman correlation test between the obtained concern index and its parameters and the current IUCN Red List categories.

RESULTS

eBird data

We obtained a total of 21,511 buttonquail eBird observations from 8948 different localities worldwide (Table 2), comprising all buttonquail species except the Luzon Buttonquail and the New Caledonian Buttonquail, for which there were no sighting in the eBird database. Distribution maps and spatial distribution of eBird sightings of all buttonquails species can be consulted in  Figure S1 (s111-501-510.pdf). More than half of the observations correspond to a single species, the Barred Buttonquail (n = 11,314) a common and widespread species found in South East Asia. The proportion of observations of a given species that fell within its distribution area as reported by BirdLife International ranged between 1 in Spotted Buttonquail T. ocellatus and Sumba Buttonquail to none in the Buff-breasted Buttonquail. The sampling effort made by birdwatchers within the distribution areas of the different buttonquail species was highly variable ranging from 2 checklists/100 km² for the Black-rumped Buttonquail T. nanus to 60 and 86 checklists/100 km² for the Black-breasted Buttonquail and the Buff-breasted Buttonquail, respectively (Table 2).

Together with the monotypic Luzon and New Caledonia Buttonquails, up to 26 buttonquail subspecies have not a single observation in the eBird database. In addition to the monotypic Little T. velox and Painted Buttonquails T. varius, three subspecies or subspecies groups of Barred Buttonquail had more than 1000 observations: Taiwan T. s. rostratus (n = 4289, 19.9%), Indian T. s. taigoor and bengalensis (n = 4185, 19.5%) and Indochinese T. s. blackistoni, pallescens, thai and plumbiceps (n = 1061, 4.9%), together with the Kurrichane Buttonquail T. sylvaticus lepuranus (n = 1327, 6.2%) (Table 3). Up to 13 subspecies or monotypic species (23.6%) have all their eBird observations within the BirdLife International corresponding distribution area, while for the rest there exists much variation, ranging from 0% in the Buff-breasted Buttonquail to 99% in the Madagascar Buttonquail T. nigricollis. The highest density of checklists is found in the distribution area of the Taiwan Barred Buttonquail T. s. rostratus with up to 1323 checklists/100 km².

Table 2.

Number of total eBird observations (eBird total), buttonquail observations (eBird in BirdLife) and total checklists within the corresponding BirdLife International distribution area (eBird checklists). Extent of occurrence area in km² as defined by BirdLife (Range size), scarcity and uncertainty parameters and concern index for every buttonquail species, IUCN conservation status and number of subspecies by systematic order (del Hoyo & Collar 2014).

Figure 1.

Bidimensional representation of the scarcity and uncertainty parameters to create the concern index of all buttonquail species. IUCN risk extinction category is shown in brackets. For a better understanding, the lower monitoring concern index zone has been enlarged.

DISCUSSION

By using data from a single citizen science birding platform and the published distribution areas for every species and subspecies of buttonquails, we have been able to create a simple index to rank monitoring and conservation priorities.

The index correlates with the already published IUCN Red List categories for every species (IUCN 2022). Higher values match with the most endangered species: Buff-breasted, New Caledonia and Luzon Buttonquails. The Buff-breasted Buttonquail is listed as Critically Endangered by the IUCN (2022), and Endangered by the Australian environmental authorities (Mathieson & Smith 2009), remaining the only Australian bird species never photographed alive. Moreover, very recent research found that most probably all recent sightings should be better regarded as misidentifications with the commoner Painted Buttonquail (Webster et al. 2022) and thus proposing to change its Australian legal status to Critically Endangered. Caledonian Buttonquail is listed as Critically Endangered (IUCN 2022). The species is only known by the type specimen when the species was described in 1889 (Ogilvie-Grant 1889). In absence of intensive specific surveys (Elkstrom et al. 2002) and due to the absence of sightings, there is serious concern for the continued existence of this species (Butchart et al. 2018).

Figure 2.

Scatter plot of the concern index and the IUCN categories for buttonquail species.

Figure 3.

Concern status of (A) all buttonquail subspecies and monotypic species, with a detailed view of the (B) high and (C) very high concern index groups.

The Luzon Buttonquail is barely known from a few localities in the mountains of central Luzon, in the Philippines (Collar et al. 1999). Due to the lack of information, it is listed as Data Deficient (IUCN 2022). As expected, the uncertainty parameter is very high (Table 2). In 2009, after two nights of bird trapping in Dalton Pass, one bird was trapped and sold for food, being the only living specimen of this species ever photographed (Allen 2009), suggesting a viable population may still persist in this region. However, this is the only record in the 21st century. Historically the species is only known from six localities, and 79% of all 43 existing museum specimens were trapped at a single location: Dalton Pass, where there is very heavy hunting pressure by locals (Collar et al. 1999, Round & Allen 2010). Although more data are needed, the species should likely be better listed as Endangered or Critically Endangered. The South African fynbos endemic Fynbos Buttonquail, until recently, was listed as Endangered, However, Lee et al. (2018) found it to be more common and widespread than previously thought, thus proposing to down list the species to Vulnerable. This species ranks in our index with lower or similar values than other Least Concern species. Currently the species has been re-assessed and listed as Least Concern (IUCN 2022).

Due to the high diversity of living forms, these assessments are usually made only at the species level (IUCN 2022). Nevertheless, conservation status could also be evaluated at lower taxonomic levels, such as subspecies, and multiple ecologically or regionally defined units, such as populations (Gärdenfors et al. 2001, Phillimore & Owens 2006). In such cases, many taxa could be in decline before a formal assessment can be done or conservation measures and actions are taken. Based on the correlation between our concern index and the classification of the IUCN Red List we are confident that our concern index will work properly also at the subspecific level. We obtained very high values for highly unmonitored small island endemic subspecies such as Bawean Barred, Savu Red-Backed, Sulu Common and Banggai Red-backed Buttonquails (Table 3, Figure 3D). A possible explanation for these underreported taxa might be that birdwatchers visiting remote islands in search of endemic birds will pay low attention to somewhat less attractive habitats such as those usually preferred by buttonquails, and consequently those endemic buttonquails subspecies could be under-recorded in the eBird database. Although the lack of data could not confirm the scarcity of these subspecies, it does reveal a high uncertainty about their conservation status, consequently classifying them very high in our concern index. Together with New Caledonian and Buff-breasted Buttonquails, nine more subspecies have high concern index values and thus could be regarded as Endangered (Table 3, Figure 3C). Of these 15 taxa, the Buff-breasted Buttonquail is found in Australia and New Caledonian Buttonquail in New Caledonia, one of the overseas territories of France. Therefore, both species are under the protection of wealthy countries, which means that legislative and economic resources for monitoring and conservation are likely to occur. However, the remaining 13 taxa depend on just three countries: five in the Philippines and four in Indonesia and Papua New Guinea, respectively. In these countries, the conservation of those potentially highly threatened taxa is compromised by the lack of economic and human resources for population monitoring and the design and implementation of effective conservation measures. Among the subspecies with the lowest concern index values, three groups can be identified. Most of them show very low values for both uncertainty and scarcity, and consequently can be considered to be safe (Figure 3B). Among all taxa classified at the lower concern index values, those with higher values are Abrolhos Painted Buttonquail T. v. scintillans and the Andalusian Buttonquail T. s. sylvaticus (Table 3, Figure 3B). The first one is endemic of the Houtman Abrolhos Islands, a small archipelago in West Australia which is classified as Vulnerable by the Australian Government, while the second one is the only buttonquail species present in the Western Palearctic. Although there is no assessment of the entire taxon, the Andalusian Buttonquail has been classified as Critically Endangered at the European level (BirdLife International 2015). It has been recently considered extinct in Spain, its last stronghold in Europe, by the Spanish Government, with just a single population known to persist in west Morocco (Gutiérrez-Expósito et al. 2019).

Table 3.

Concern index and uncertainty and scarcity parameters for all buttonquail subspecies, subspecies groups, and monotypic species sorted by concern index values with the proposed English names, mostly based on Gutiérrez-Expósito et al. (2011) and del Hoyo & Collar (2014).

A similar assessment should be undertaken at the population level, many of which could be highly endangered locally or on the verge of extinction. This could be the case for the Taiwan population of the Indochinese Common Buttonquail, where despite the very high eBird sampling effort made in the island (345,594 checklists), not a single sighting of this species has been obtained, while with the same effort, up to 4290 observations are found for the endemic Taiwan Barred Buttonquail.

The secretive habits of the buttonquails make them easily overlooked. As a consequence, silent extinction processes can occur, as happened with the Andalusian Buttonquail (Gutiérrez-Expósito et al. 2020), and this could be a risk for those taxa with a high concern index. However, as research on them continues, more detailed knowledge is coming to light, allowing the downlisting of the Black-breasted Buttonquail and the Fynbos Buttonquail from Endangered to Vulnerable (Lees & Smith 1999, Smyth & Pavey 2001, IUCN 2022) and Least Concern (Lee et al. 2017, 2018, IUCN 2022), respectively, giving some hope for those understudied species and subspecies.

Our approach is a very simple method to evaluate the monitoring and conservation priorities for taxa lacking a formal extinction risk assessment within a given avian group. Furthermore, as far as the concern index is based on two components, uncertainty and scarcity, it is also possible to prioritize focusing on research or conservation, or both, depending on the component's values obtained. Rather than a substitute of the formal assessments, this index is especially useful as it can be applied not only to bird species but also to other taxonomical levels as subspecies or even to ecological units as populations of a given species.

Of course, a simple approach like this is has limitations. Results in areas where eBird is not the primary birding tool should be taken with caution as data in these areas can underestimate bird occurrence. However, as when we examine buttonquail observation data we control for observation effort, we think that the results obtained should be comparable among taxa. We applied our novel approach just to one bird family, but future work could expand it to other bird groups or adapt it to the information available from different data sources for other taxa.

In summary, in a conservation landscape of perpetually limited funds, we propose a straightforward approach that can give managers, policymakers and ecologists an easy tool to prioritize conservation and research resources.