**5. Conclusions**

In this study, we used Rabinowitz's rarity categories to classify New World pitvipers and our results show that although most NW pitvipers are not rare, several species have a combination of features (narrow habitat breadth and small geographical range) that may make them vulnerable to extinction in the near future. We detected regions where a concentration of rare pitvipers occurs (e.g., mountainous regions of Mexico, southern Central America, and Transandean South America), which may become target regions for new protected areas. Our results also show that rarity in NW pitvipers tends to decrease with increasing latitude and hence latitude can be used as a proxy in the lack of data, helping the e fforts of conservation in some situations. Lastly, we compared rarity patterns with IUCN assessments and showed that among species in our rarest category, some were not listed as threatened. Thus, we propose the use of our rarity analyses as additional criteria for setting conservation priorities, which can contribute e ffectively to the conservation of NW pitvipers. Indeed, combining complementary approaches like Rabinowitz's rarity method (this study) with those presented in the previous paragraphs (Table 2), may substantially improve our knowledge on the conservation needs of NW pitvipers.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/1424-2818/11/9/147/s1, Figure S1: Relief maps showing the distribution of rare vipers in the New World (North and Central America above, South America below). Note that most rare pitvipers occur in mountainous areas. Grid details and richness as in Figure 1. Figure S2: Comparison of the size of viper geographical distribution among regions of the New World. Table S1: Sources of locality data used to calculate geographical range for New World vipers. Table S2: Habitats of New World vipers. Major habitats cited in the literature, number of major habitats, and the category considered in our analyses for habitat breadth. Table S3: Species abundance of New World pitvipers. The column "Abundance" (rare = 1, frequent = 2 and abundant = 3) is based on literature information for each species and LA is the category considered in our analyses based on the previous column (low local abundance for one in previous column, high local abundance for two and three). Table S4: Geographical range (GR1, in km2), extent of occurrence (EOO, cf. IUCN), region of the New World (CEAM = Central America; CISA = Cisandean South America; ISLA = island endemics; TRSA = Transandean South America; USNM = United States and northern Mexico), IUCN category (LC = least concern; DD = data deficient; NT = near threatened; VU = vulnerable; EN = endangered; CR = critically endangered), source of IUCN category (BRZ = Brazilian assessment by ICMBio; IUCN = IUCN Red List; SIS = IUCN Species Information System; TS = this study), rarity variables (GR2 = geographical range; HB = habitat breadth; LA = local abundance), rarity cell in Rabinowitz "seven forms of rarity" (see Figure 1), and rarity category (NR = not rare; LI = low intermediate; HI = high intermediate; RT = rarest) for all species of New World pitvipers. Table S5: Body size (maximum length, in milimeters) and absolute latitude used in the analysis for predicting rarity. The column "references" refers to the data on maximum length. Table S6: Parameter estimates of Bayesian phylogenetic mixed-models analyses. An asterisk indicates a statistically significant result (*p* < 0.01). CI = Confidence intervals.

**Author Contributions:** Conception of the work, I.B.-B., M.M., and L.R.V.A.; acquisition, analysis, and interpretation of data, I.B.-B., M.M., L.R.V.A., M.B., and P.I.P.; writing—original draft preparation, I.B.-B. and M.M.; writing—review and editing, I.B.-B., M.M., L.R.V.A., M.B., and P.I.P.; acquisition of funding, M.M.

**Funding:** I.B.-B. L.R.V.A. and M.M. thank São Paulo Research Foundation (FAPESP) (fellowships #2012/15398-7 and #2012/22197-8 to I.B.-B., #2012/02038-2 and #2016/14292-1 to L.R.V.A., grants #2011/50206-9, #2015/21259-8, and #2018/14091-1 to M.M.). M.M. and P.I.P. thank the Brazilian Research Council (CNPq) for research fellowships (306961/2015-6 and 310885/2017-5). M.B. receives funding from The Ru fford Foundation.

**Acknowledgments:** We thank two anonymous reviewers for providing useful suggestions, C. Spencer for providing unpublished data on the distribution of *Crotalus atrox* and C. Nogueira and the group of collaborators of the database on the distribution of Brazilian snakes for providing updated unpublished distribution data.

**Conflicts of Interest:** The authors declare no conflict of interest.
