The importance of addressing different Red Lists in conservation studies : an analysis comparing the conservation status of Brazilian mammals

The importance of addressing different Red Lists in conservation studies: an analysis comparing the conservation status of Brazilian mammals. Red Lists are important conservation tools because they attempt to estimate the extinction risks of species. We compared the conservation status of Brazilian mammals presented in the Brazilian Red Book with those presented in the IUCN Red List, highlighting the importance of each list and why they should be used jointly. Out of 636 species, 181 were considered endemic to Brazil and 121 were considered threatened by at least one of the lists. Considering the complete database, 86 % of the species had the same status on both lists, whereas only 48 % of the threatened species had the same status. Some possible factors responsible for variations are the period in which the evaluations were carried out, the evaluation process and the fact that a species threatened nationally may not be threatened globally. We recommend that communication should be improved, that lists should be kept updated, and that both the type of information and the data itself to be used in the assessments should be standardized.


Introduction
Biodiversity conservation is one of the biggest challenges facing the current generation (Vale et al., 2009). Megadiverse countries, such as Brazil, therefore have an enormous responsibility when it comes to protecting endangered species (Brandon et al., 2005). The richness of Brazil's mammal species, for example, is considered by some authors to be the highest in the world, with over 700 species and a high degree of endemism at the national level (Mittermeier et al., 1997;Costa et al., 2005;Lewinsohn and Prado, 2005;Quintela et al., 2020).
When species are assigned to categories (known as conservation status) that represent their degree of threat, their risk of extinction can be estimated, making it easier to infer which species need urgent conservation actions (Peres et al., 2011), evaluate the state of biodiversity, identify sites for conservation action, and inform policy and management (Rodrigues et al., 2006). Red Lists of threatened fauna are, from this point of view, important conservation tools. Having already assessed the global risk of extinction of more than 116,000 species (including more than 5,000 mammals), the International Union for Conservation of Nature (IUCN) has played a major role in making these lists known worldwide. Some of the criteria used in those assessments are restricted geographic distribution, small and declining population size, and, based on quantitative analysis, a high probability of extinction in nature. Its scheme of species classification according to threat status uses the following categories: Not Evaluated (NE), Data Deficient (DD) (when there is no adequate information to assess the risk), Least Concern (LC) (when the species is evaluated but does not fall into the other categories; usually encompassing abundant and widely distributed taxa), Near Threatened (NT) (when the species is close to qualifying as threatened or when it is expected to be classified as such soon), Vulnerable (VU) (when the species faces a high risk of extinction in the wild), Endangered (EN) (when the species faces a very high risk of extinction in the wild), Critically Endangered (CR) (when the species face an even higher risk of extinction in the wild), Extinct in the Wild (EW) and Extinct (EX).
In Brazil, the Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio), a Brazilian government institution from the Ministry of Environment, periodically publishes the so-called Red Books of Threatened Fauna. These Red Books have a similar role to the IUCN Red Lists, listing the species considered to be threatened nationally, classifying them according to their conservation status, and providing ecological information about them. The most recent Red Book was published in 2018, representing a huge effort to evaluate all described vertebrate taxa occurring in the country and listing 680 mammal species officially known to occur in Brazil. Of these, 108 (15.9 %) were considered nationally threatened (ICMBio/MMA, 2018).
In the present study, we compared the conservation status of Brazilian mammal species listed in the Brazilian Red Book of Threatened Fauna with those presented in the IUCN Red List, highlighting the importance of each list and why they should be used jointly in order to generate even more accurate assessments. We expected to find some differences in the status of species whose geographic distribution was broad and exceeded the country's territorial limits (i.e. non-endemic species). However, as the national distribution of species that occur exclusively in Brazil (i.e. endemic species) corresponds to their global distribution, we postulated that the status of those species would not vary between lists. In case some endemic species (especially those considered to be threatened) had a different conservation status in each list, we would emphasize the need for special attention the next time their conservation status is assessed.

Material and methods
We compiled a database (see table 1s in supplementary material) containing all Brazilian mammal species (regardless of subspecies) according to the Brazilian Red Book of Threatened Fauna (ICMBio/MMA, 2018), their national (obtained from the Brazilian Red Book itself) and global conservation status (obtained from the IUCN Red List of Threatened Species platform, 2019-3 version), and the year in which the species' status was assessed in each of the lists. In the Brazilian Red Book, however, the pampas deer (Ozotoceros bezoarticus) and some primates were evaluated only at the subspecific level. In these cases, to standardize our analysis, we chose to consider the status of the least threatened subspecies as the status of the species. Using the data found in the 'Geographic Range' section of the IUCN Red List and the Brazilian Red Book, we also added the information of whether a species was endemic to Brazil or not. The lists were then compared according to the conservation status of each species to observe which species differed in status between lists.
We also observed whether each species had the same status on both lists or if it had a lower conservation status (i.e. less threatened) on one of the lists. For example, if a species was assessed as not threatened (i.e. Least Concern or Near Threatened) by the Brazilian Red Book but as threatened (i.e. Vulnerable, Endangered or Critically Endangered) by the IUCN Red List, we considered it had a lower status in the national list. Similarly, if a species was classified as Critically Endangered in the Brazilian Red Book and as Vulnerable in the IUCN Red List, we considered that it had a lower status in the global list, despite being considered threatened by both lists. We made those comparisons considering four different scenarios: a) all species present in our database; b) only the endemic species; c) only the species considered to be threatened (i.e. species classified as either Vulnerable, Endangered or Critically Endangered) by at least one of the lists; d) species considered, simultaneously, as endemic and threatened. These analyses did not include species that were categorized as Data Deficient in either of the lists.
In order to better analyze the differences between the lists, we divided the analyzed species into eight groups based on taxonomy (Order rank) and/or ecological characteristics: Aquatic Mammals (comprising cetaceans and sirenians), Carnivora, Chiroptera, Didelphimorphia, Glires (comprising Rodentia and Lagomorpha), Primates, Ungulates (comprising Artiodactyla and Perissodactyla) and Xenarthra (comprising Pilosa and Cingulata). For each group, we compared the proportions of species classified in each conservation status with lists using Fisher's exact test (only the species classified as Near Threatened, Vulnerable, Endangered and Critically Endangered were considered). The analyses were performed in R version 4.0.2.

Results
According to the Brazilian Red Book of Threatened Fauna, 680 mammal species were known to occur in Brazil. Since we chose not to include taxa that were not evaluated by the IUCN Red List, as well as those that IUCN considers as subspecies (as opposed to full species), and the candango mouse (Juscelinomys candango), classified as extinct by IUCN, our database comprised 636 species. Additionally, 181 species present in our database (28.5 % of the total) were considered endemic to Brazil. Primates, Chiroptera and Glires made up most of the species, both when considering the complete list and when considering only endemic species ( fig. 1). As the tapeti (Sylvilagus brasiliensis) was the only member of the Order Lagomorpha in our database, the high representativeness of Glires in our analyses was due to the richness of rodent species. The Aquatic Mammals group, on the other hand, was the only group in which no species were considered endemic to Brazil, and three other groups presented a low number of endemic species: Carnivora (the hoary fox, Lycalopex vetulus, was the only endemic species), Ungulates (the small red brocket deer, Mazama bororo, was the only endemic species), and Xenarthra (the three-banded armadillo, Tolypeutes tricinctus, and the maned three-toed sloth, Bradypus torquatus, were the only endemic species).
One hundred and twenty-one species (19.0 % of the total) were considered threatened by at least one of the lists (table 1). Of these, 104 were considered threatened according to the Brazilian Red Book, with 54 (51.9 %) being classified as Vulnerable, 40 (38.5 %) as Endangered and 10 (9.6 %) as Critically Endangered. In the IUCN Red List, 40 species ( Primates and Glires made up most of the threatened species (table 1). Only one species classified as Critically Endangered according to the IUCN Red List did not belong to one of these two groups: the single-striped opossum (Monodelphis unistriata) (Didelphimorphia). The Brazilian Red Book, however, classified one didelphimorph (the black-shouldered opossum, Caluromysiops irrupta) and two cetaceans (the blue whale, Balaenoptera musculus, and the Franciscana dolphin, Pontoporia blainvillei) as Criti-   Regarding the conservation status of species by group, the Least Concern status was the one in which most of the species of any group were classified. The group Carnivora, however, presented the most significant difference between the lists, with 13 species considered threatened according to the national list but only three according to the global one (table 1). Statistically significant differences between the proportions of species classified in each conservation status (excluding Least Concern) between the two lists were only observed for the groups Carnivora (p-value = 0.001) and Chiroptera (p-value = 0.01). While the Brazilian list has more species classified as Vulnerable, IUCN classifies more species as Near Threatened. Considering only the endemic species (and also excluding species classified as Least Concern), on the other hand, no statistically significant difference was observed between lists for any group. Although the two lists are similar when considering the total number of species classified in each conservation status, further analysis shows that this equivalence may be apparent, since the status of many species varies between the two lists.
Considering the complete lists and excluding the species that are classified as Data Deficient in either assessment, 420 species (85.7 % of the total) had the same conservation status on both lists, whereas 27 (5.5 %) had a lower status according to the national assessment, and 43 (8.8 %) had a lower status on the global list (table 2). However, when only the endemic species were considered, we observed that 100 species (79.4 %) were classified with the same status on both lists, while 16 (12.7 %) had a lower status on the national list and 10 (7.9 %) had a lower status on the global list (table 2). Nevertheless, divergence between lists was even more pronounced when we restricted our analysis to threatened species. In this case, 51 species (47.7 %) had the same conservation status on both lists, 21 (19.6 %) had a lower status according to the national assessment, and 35 (32.7 %) had a lower status on the global list (table 2). Finally, considering the endemic species that are also threatened, 35 species (60.4 %) had the same status on both lists, whereas 13 (22.4 %) had a lower status on the national list and 10 (17.2 %) on the global list (table 2).
Considering the species analyzed by group, the Carnivora, once again, stands out: of the 13 analyzed species considered to be threatened, 11 (84.6 %) had a lower status on the global list and only one (7.7 %) had equal status on both lists (table 2). This pattern was also observed in the Aquatic Mammals group, where six of the 10 analyzed species (60.0 %) had a lower status according to the global assessment. Endemic didelphimorphs also presented a tendency of divergence between lists: only half of the six species considered had the same conservation status on both lists.
The average difference between the years in which species evaluations took place in each list was 2.73 years, with 21 species evaluated in the same year on both lists, six evaluated one year apart, 341 two years apart, 153 three years apart, 17 four years apart, 67 five years apart, 30 six years apart and one that was evaluated seven years apart.

Discussion
Although the two lists pursue the same goal (i.e. to evaluate extinction risks of species and classify them accordingly) and use the same categories of threat and the same criteria on their assessments, the conservation status of more than half of the threatened taxa differed between lists, and this variation was more marked in some mammal groups than in others. Since a species threatened nationally may not be threatened globally (Gädendorfs, 2001), one  of the main reasons why the conservation status of many non-endemic species differ between the two lists becomes clear. If we look at mammal groups composed mostly of species with wide geographic distributions, in which rates of endemism are low (such as the orders Carnivora and Cetacea), this becomes even more evident. Indeed, the groups Carnivora and Aquatic Mammals were those with the greatest proportion of threatened species having a lower status in the global list than on the national list. The puma (Puma concolor), for example, can be found across much of the American continent, from Canada to southern Argentina (Nielsen et al., 2015) and it is classified as Least Concern globally, even though it is considered Vulnerable in Brazil. Similarly, the southern right whale (Eubalaena australis) has a circumpolar distribution across the entire Southern Hemisphere (Cooke and Zerbini, 2018) and is also classified as Least Concern globally, but as Endangered in Brazil. In such cases, the IUCN recommends that national assessments evaluate species as if they were endemic or completely isolated from other populations to obtain a preliminary status. After taking this first step, the status of the species can either be changed or subsequently maintained, considering the possibility of migration of individuals into and out of the region under analysis (IUCN, 2012). Endemic species, on the other hand, present a more delicate situation. As an endemic species only occurs within a restricted area, its regional population also corresponds to the global one. Therefore, it was expected that the conservation status of endemic species would not differ between national and global lists. However, our analysis has shown that this was not always the case, as the conservation status of 20.6 % of the endemic species and of 39.6 % of the species that were both threatened and endemic varied between lists.
One possible factor responsible for variation in status between lists is the period in which the evaluation of the status of taxa was carried out. However, if we consider the time interval between the national and global evaluations of a given species, it is noteworthy that this never exceeded seven years. Furthermore, the Brazilian three-banded armadillo (Tolypeutes tricinctus) was classified as Endangered according to the Brazilian Red Book and as Vulnerable according to the IUCN Red List, despite being endemic to Brazil and both assessments taking place in 2013. This could indicate that the period in which the evaluation was carried out may not be the only reason for the divergences observed, nor the main reason for all of them. Nonetheless, we recognize that changes in the conservation status of a given species can occur within short periods of time, following new publications concerning reassessments of its geographic distribution and of major changes undergone by its habitat (e.g. Fernandes et al., 2007;Attias et al., 2009;Hirsch and Chiarello, 2012), and taxonomic revisions (especially in cases where a single species is divided into two or more, e.g. Agapow et al., 2004;Nascimento and Feijó, 2017;Ang et al., 2020). The Brazilian Red Book (ICMBio/MMA, 2018) also mentions that more recent and accurate information (especially regarding declines or recoveries of populations) and adjustments in the method itself may be responsible for changes in the conservation status of species, sometimes even resulting in their removal from the list of threatened taxa (i.e. when a species classified as Vulnerable, Endangered or Critically Endangered is re-classified as Least Concern or Near Threatened). This was the case of the humpback whale (Megaptera novaeangliae): previously classified as nationally threatened, the prohibition of whaling activities by the Brazilian government in 1987 resulted in an increase in the number of individuals in national waters (Andriolo et al., 2010;Bortolotto et al., 2016)  A possible additional cause of divergences may be the evaluation process itself. Although both lists are based on expert opinion and follow a strict process to have assessments performed as accurately as possible, it should be considered that there may be a subjective component in assessing the risk of losing species (especially if the methods are not strictly followed). Costa et al. (2005) stated that national lists could also benefit from scientific knowledge generated by unpublished data, including theses, dissertations, local journals, and personal field experience. However, we observed that global lists can also use this type of data to assess species extinction risks. Therefore, some divergences between lists may not be related to the type of publication used, but as we have mentioned, to the data and to the process itself. Some previous works have attempted to evaluate and compare Red Lists in a similar way to ours. However, contrary to what we expected, publications focusing on Brazilian mammals are not that common. Costa et al. (2005) briefly compared the conservation status of threatened Brazilian mammals using the 2003 national list. Nonetheless, in addition to the current list being much more comprehensive than the previous ones, those authors did not carry out as many analyzes as we did. The Brazilian national species list was also compared with the IUCN Red List by Brito et al. (2010) in a work that addressed various taxa from three other countries besides Brazil: Colombia, China, and the Philippines. Other relevant works dealing with vertebrate groups other than mammals are those of Garcia and Marini (2006), who focused on threatened Table 3. Endemic species considered to be threatened by at least one of the lists and whose conservation status varied between the assessments: * species with higher conservation risk according to the Brazilian Red Book (national assessment) than with the IUCN Red List (global assessment).  (2012), who focused on Brazilian reef fishes. As in our study, these studies found divergences between lists that needed to be resolved because they could raise doubts on the credibility and usefulness of these important conservation tools. Nonetheless, there is little point in debating whether one list is better than the other. The main goal of our study was to draw attention to the fact that differences in the conservation status of species may exist between global and national lists and that such differences do not necessarily represent errors or outdated information. The two lists are based on different spatial scales and, consequently, have distinct potential uses. The national list (i.e. the Brazilian Red Book), at least in Brazil, is the one used to define which species of Brazilian fauna are considered threatened, so that those species can be fully protected under the Brazilian laws, and actions such as their capture, transportation and commercialization be prohibited. The IUCN, on the other hand, aims to show what actions are needed to save species from extinction and where they should be directed (Rodrigues et al., 2006). The IUCN Red List therefore plays a fundamental role in guiding scientific research, influencing allocation of resources for conservation, and informing policies and conventions (especially international ones) (Rodrigues et al., 2006). Both lists also provide useful information about the assessed species, including their geographic range, ecology, natural history, and the main threats to their survival. It seems reasonable to assume that while regional lists are critical to decision makers within a given country, serving as a basis for the elaboration of national public policies and during the creation of conservation units and other legally protected areas, global lists, which can also guide such actions within a bigger scenario, may function as a 'barometer of life' (an expression the IUCN often uses to describe its own potential) at a global scale. The global list gains a greater visibility than national lists, since it is internationally recognized, and is fundamental for the conservation of species with wide geographic distribution.
Assessing the extinction risk of a species is not an easy task since there are uncertainties and predictions throughout the process. We thus recognize the quality of the work that is done by the authorities responsible for evaluations and recommend that communication and information exchange between authorities and researchers be improved. Perhaps the best way to avoid future divergences between lists (especially for endemic species) would be to undertake a joint assessment between the authorities responsible for the national and global assessments. It is also extremely important to keep the lists updated so that they always reflect the current status of each species. Standardizing both the type of information and the data itself to be used in those assessments would, if possible, also be of great value, as would be the presentation, by the Brazilian Red Book, of the conservation status of all species at the specific level (as we have mentioned, some species were evaluated only at the subspecific level).
We also recommend special attention when making future conservation status assessments of species that, although endemic, were classified with different status in each of the lists (see table 3). Additionally, it is important to focus on species classified as Data Deficient since the main reason that leads a species to be classified as such is the lack of adequate information about its distribution and/or its population (ICMBio/MMA, 2018). Thus, the possibility that a given species classified as Data Deficient is threatened should not be overlooked.
Finally, we would like to mention that, while we focused on two main lists in this article, several other lists could be similarly analyzed. The larger the scale, the harder it is to detect and identify eventual regional discrepancies. Thus, state and biome lists, for example, can also be important, especially in a country of continental dimensions like Brazil. Indeed, while a few Brazilian states have their own lists of threatened fauna, most states still lack these (see Brito, 2008). Analyses at smaller scales may allow more accurate conclusions and, when interpreted together, tend to promote a better understanding of how threatened a species really is. In this regard, some recent studies deserve to be highlighted because they have proposed novel approaches related to conservation status assessments using, for example, data on habitat preference and population abundance (e.g. Santini et al., 2019), or on ecological traits (e.g. Davidson et al., 2009). It is also important to highlight that endemism is a relative measure related to the idea of habitat restriction. Since all species end up being endemic to a certain area (although this area may be large enough to correspond to several countries, for example), care must be taken when using this concept. Still, we believe that national lists may be easier to incorporate into effective conservation strategies than international lists. Conflicts in conservation policy can be avoided if the evaluation process is not confounded by processes that do not operate within the study area. Nonetheless, we believe that the use of both global and national lists in a complementary way (or at least the mention, in the publications, of how threatened the studied species is, both at the global level and where the corresponding study took place) tends to make conservation studies and publications more comprehensible, providing readers with a better understanding of how threatened the studied species is. List Criteria at Regional and National Levels: