**4. Discussion**

Based on the assessments, the number of endangered species of *Brachycephalus* should increase from one to 21 (Table 2). This is a significant shift and poses the question regarding why only one species had been formally recognized as threatened until now (Table 2). There are two possible reasons: (1) a delay due to the fact that many species have only been recently described and (2) a resistance based on current policies of the MMA to change a species conservation status in a short period of time (see below). Twenty-two species have been described in the last 10 years, 14 of which are only known from their type locality, and there is a natural tendency to expect them to be more widely distributed; however, studies in recent years have been gradually revealing new species rather than new records of known species, and new records of already described species have not substantially altered their geographical distributions (e.g., [24,29,76,99]). For example, a new locality record for *B. fuscolineatus* published after its description increased its EO by just 0.19 ha [24], and this species still has the smallest estimated EO for any *Brachycephalus* species (Table 2). Two new records of *B. curupira* (Table 1, [29]) double its EO, which remains small (= 4751.4 ha; Table 2). A new record of *B. albolineatus* published after its description [76] and two new localities included in Table 1 substantially extend its EO, but as in the case of *B. curupira*, this remains small (= 2784.4 ha; Table 2). A new record for *B. nodoterga* [99] did not change the EO of the species because it is located within its EO polygon. *Brachycephalus* was not found in 13 localities from southern São Paulo to northeastern Santa Catarina with an altitude comparable to other localities where *Brachycephalus* populations were present. Overall, the reduced geographical distributions of *Brachycephalus* is the rule for the montane species of the genus, i.e., the *B. ephippiumsi* and *B. pernix* groups [23]. *Brachycephalus ephippium* is the only exception of a montane species group with a large EO, but it is expected that some, if not all, populations may be identified as distinct species in future studies [23].

With respect to the resistance to incorporate drastic changes into the o fficial number of endangered *Brachycephalus* species, this proposal is warranted despite the current policy of the MMA indicated. The MMA joined several international agreements that set targets for the conservation of the country's threatened biodiversity, and these e fforts have been implemented in the successful execution of National Action Plans for the Conservation of Brazilian Endangered Species (Planos de Ação Nacional - PANs). The national scientific community and the MMA have been working together to list threats and conservation actions to all threatened species of the country and to review and to monitor these actions annually. This is possibly the reason that the MMA tended to prefer moving forward with conservation strategies of species that are already listed as threatened rather than revising the list. The e ffort to prioritize conservation initiatives prior to substantial updates to the list of endangered species is recognized, but the need for MMA to revise the list and to recognize the species listed in this article as threatened is also acknowledged given that they are not ye<sup>t</sup> legally protected.

The most prevalent threat to *Brachycephalus* is deforestation, much of which is no longer done with heavy machinery and chainsaws. Recently, deforestation in the Atlantic Rainforest has become more subtle and involves the selective removal of trees and shrubs, particularly through inconspicuous strategies, such as bark girdling, which leads to the opening of the canopy and an increased tendency for wind to knock down additional trees. These actions are deliberately conducted a few meters into the forest edge to avoid detection by environmental inspectors. This type of deforestation has been carried out at an alarming rate in Paraná and in the northeast of Santa Catarina for at least 25 years to cultivate bananas, and more recently, to cultivate palm trees (*Archontophoenix alexandrae*). Deforestation for agricultural activities could also result in soil contamination, a ffecting species that depend on specific microhabitats and that have permeable skin [121]. Finally, deforestation could exacerbate edge effects, altering microhabitats and microclimatic conditions, which changes sunlight exposure, soil moisture, and plant species composition in the edges [15]. Indeed, *B. fuscolineatus* was not encountered in forest edges but only in more nuclear vegetation [24].

Deforestation in lowlands can lead to a decrease in the altitudinal distribution of cloud forests [122], potentially shifting the distributions of montane species of *Brachycephalus* to higher altitudes. This

possibility of altitudinal species displacement could also be driven by climate change [23]. In tropical forests, temperatures can vary from 0.4 ◦C to 0.7 ◦C per 100 m altitude variation [123]. The thermal variation in the altitudinal gradient in a site with the occurrence of *Brachycephalus* (*B. pernix*) was determined as 0.56 ◦C of the reduction every 100 m of altitude [124]. A di fference in the precipitation levels at this site was also evaluated, with an increase of 40 mm in mean annual precipitation every 100 m at elevation [124]. Studies on litter anurans of the Atlantic Rainforest, including *Brachycephalus*, have shown that population densities are particularly a ffected by air humidity, air temperature, and altitude [52,55,78,90]. This climatic dependence and its relationship with the altitude gradient raises concerns for the long-term conservation of *Brachycephalus* species that occur in mountains with a restricted altitudinal amplitude.

Lowering the category of threat for *B. pernix* from CR [45] to VU (Table 2) is proposed. The e ffects of trampling and timber harvests by tourists in the type locality of the species are likely to be minor, which is entirely distributed within a protected area (Parque Estadual Serra da Baitaca; Table S1). There is a threat of fire in part of the EO of the species, but the vegetation cover is regenerating well in this area after many years of managemen<sup>t</sup> by volunteer mountaineers, reducing fire susceptibility.

It is recognized that there is some level of subjectivity to apply EO and the number of locations of threatened species. This is because each parameter shows some overlap between EN and VU categories. *Brachycephalus hermogenesi* and *B. nodoterga* fit the EN category, but both are recommended to be considered for the VU category because part of their EO is in protected areas (Table S1).

The presence of threatened *Brachycephalus* in protected areas is a useful tool to rank the species for which conservation actions are more urgent. In Table S1, 10 species without records in protected areas are recognized with three classified as CR (*B. boticario*, *B. mirissimus*, and *B. quiririensis*), two as EN (*B. actaeus* and *B. leopardus*), one as VU (*B. albolineatus*), and four as DD (*B. atelopoide*, *B. auroguttatus*, *B. bufonoides*, *B. leopardus*, and *B. verrucosus*). There are no known living populations of two species (*B. atelopoide* and *B. bufonoides*). The remaining eight species belonging to the *B. pernix* group occur in southern Paraná (*B. leopardus*) and Santa Catarina (*B. actaeus*, *B. albolineatus*, *B. auroguttatus*, *B. boticario*, *B. mirissimus*, *B. quiririensis*, and *B. verrucosus*). Also, it is argued that DD species need special attention to direct further studies to complete adequate assessments of their conservation status as soon as possible.

Santa Catarina stands out as the state in which emergency conservation actions should converge. Creating protected areas is an important way to protect species, however, the conservation of the top three priority species would require the creation of three new protected areas. A protected area for the CR *B. quiririensis* could already house other species of *Brachycephalus* that are not in any reserve, namely *B. leopardus* (EN) and *B. auroguttatus* (DD). Nonetheless, to be e ffective, a protected area would first require the expropriation of the land in addition to managemen<sup>t</sup> actions aimed at recovering forest quality. Given that there are dozens of protected areas waiting for expropriation, this path to conservation does not seem likely at the moment. Private protected areas are an alternative (e.g., Reserva Particular do Patrimônio Natural—RPPN), and some of them already protect two species of *Brachycephalus* (*B. mariaeterezae* and *B. tridactylus*; Table 1 and Table S1). This is the most stable category of protected areas in Brazil and cannot be undone; however, one aspect that does not stimulate the creation of more private protected areas is the lack of governmen<sup>t</sup> incentives to private owners, except for exemption from territorial taxes. There is an impediment to transferring public financial resources to private persons, even if they are addressing conservation measures.

The conservation of *Brachycephalus* should also include alternatives to the creation of protected areas. One approach would be to lease land with the occurrence of threatened *Brachycephalus* at a percentage of the regional value of production per hectare of mountainous lands, which would be an incentive for landowners to leave their land intact. This must be governed by a renewable contract. For this strategy to be put into practice, it is vital to attract international resources. It would also be interesting to attract additional resources of the lease value to promote environmental recovery. The managemen<sup>t</sup> of invasive alien species, both plants and animals, is unfortunately incipient in Brazil due

to the high involved costs, thus discouraging the proposition of new managemen<sup>t</sup> projects. The state of Santa Catarina has emphasized its concern with the conservation of microendemic anurans in its region, creating a specific program for this purpose (Portaria Instituto Estadual do Meio Ambiente - IMA N◦ 283/2018 - 19/12/2018). This is an interesting e ffort that can put actions discussed into practice and can also result in other e ffective and innovative actions for the conservation of *Brachycephalus* in Santa Catarina. In the long term, successful practices can be replicated in other regions of Brazil.
