Venomous snakes of medical importance in the Brazilian state of Rio de Janeiro: habitat and taxonomy against ophidism

Abstract Snakebite envenoming is a major global health problem that kills or disables half a million people in the world’s poorest countries. Identifying the biting snake and its habitat use is key to understanding snakebite eco-epidemiology and optimizing its clinical management. To prevent and combat the neglected snakebite disease, we characterize the morphology, geographic distribution, habitat use, and snakebites of medically important venomous snakes in the state of Rio de Janeiro (Brazil). Despite Philodryas spp. not being considered of medical importance by the Brazilian Ministry of Health, we also explore their data once the bites may require medical intervention, may cause death, and their consequences are underestimated. Methods: We assessed taxonomy and geographic data from specimens housed in scientific collections, the literature, and the Notifiable Diseases Information System. Our data revealed fragility in the morphological characters recommended to distinguish Bothrops jararaca from B. jararacussu, identify the subspecies of Crotalus durissus and distinguish the species of Philodryas. To help identify these species, we present an identification key to the venomous snake species from Rio de Janeiro based on the morphological data collected. We record the genera Bothrops and Micrurus in all mesoregions of the state. Here, we provide the first record of C. durissus in the Serrana region, supporting the hypothesis of geographic expansion of the species in the state. The crotalic antivenom must not be missing in Médio Paraíba, Centro-Sul Fluminense, and Serrana, where the rattlesnake C. durissus occurs. Bothrops bilineatus and Lachesis muta have historical records presented for the first time herein. However, these species are likely endangered or extinct in the state. There were 7,483 snakebites reported between 2001 and 2019, with an annual average of 393.8 cases. The Bothrops genus is responsible for the majority of accidents. The highest number of cases occurred in the Serrana region, the largest pole of family agriculture in Rio de Janeiro. We improve the identification of venomous snake species, better delimit their distribution, and update the number of cases of snakebites, thus providing greater precision in the attention to this problem in Rio de Janeiro. We emphasize the importance of clinical studies to test using bothropic-crotalic antivenom and heparin in all mesoregions to treat B. jararacussu envenomation; and mechanical ventilation, atropine, and anticholinesterases in the emergency health centers in the Metropolitana and Norte Fluminense regions due to the occurrence of the coral M. lemniscatus in these areas.

On average, 27,120 snakebite cases are reported annually in Brazil, and these cases are strongly associated with environmental and socioeconomic factors (Lira-da- Silva et al., 2009;Schneider et al., 2021).The state of Rio de Janeiro harbors the following medically important venomous snake species: the lanceheads Bothrops jararaca, B. jararacussu, B. fonsecai, B. bilineatus, B. alternatus, and B. neuwiedi; the coral snakes Micrurus corallinus, M. decoratus, and M. lemniscatus; the rattlesnake Crotalus durissus; and the bushmaster Lachesis muta (Souza and Machado, 2017).Some bites caused by the genus Philodryas have been categorized as mildly dangerous, and one death has been reported (Salomão and Di Bernardo, 1995).The Lichtenstein's green racer (P.olfersii) and the Patagonia green racer P. patagoniensis also occur in Rio de Janeiro state (Souza and Machado, 2017).Due to the medical attention required in some cases [e.g., Salomão and Di Bernardo (1995)], Philodryas have also been considered of medical importance in the past [e.g., Brasil (2001); Puorto and França (2003)], but not anymore (see Brasil, 2023).

Palavras
Therefore, snakebite cases are also underreported and neglected (Vaiyapuri et al., 2013).For example, in India, deaths from snakebites are about 20 times higher than the official records (Mohapatra et al., 2011).Evidence shows that the same pattern frequently occurs in Brazil (Machado, 2018), making the fight against snakebites even more urgent.At the clinical level, the identity of the biting snake can help healthcare professionals anticipate victims' syndromes and support decision-making when treating the patient (i.e., whether or which type of antivenom to administer).This decision is important because antivenoms are effective against a limited number of venomous snakes and have potentially lethal side effects, such as fatal allergic reactions, and should not be used to treat nonvenomous snakebites (Silva et al., 2016).This is especially important because antivenom vials are scarce and expensive in many countries.Therefore, we aim to facilitate and improve the identification of venomous snake species by mapping and gathering data on habitat use in Rio de Janeiro state.Despite not being considered a snake of medical importance by the Brazilian Ministry of Health, we also updated the database on snakebites and included data on Philodryas species because their bites may require medical intervention (Brasil, 2001), may cause death (Salomão and Di Bernardo, 1995), and their consequences are underestimated (Puorto and França, 2003).We expect to improve the identification of snake species by reporting new morphological and habitat use data and detailing their geographic distribution across Rio de Janeiro.

Study area
The Rio de Janeiro state (in southeastern Brazil) has an area of 43,750,425 km 2 , 92 municipalities, and a population of 17,463,349 inhabitants (IBGE, 2022).The population density is 399.16 people per square kilometer, with 15,454,239 people living in urban areas, of which 11,838,752 (74% of the state's total) inhabit the Metropolitana region.More than 1.2 million people are starving, and 22% of the population lives below the poverty line (IBGE, 2010;Neri, 2022).The state lies within the threatened Atlantic Forest biome (Colombo and Joly, 2010).The habitat heterogeneity in the state provides an extraordinary variety of habitats and species (Veloso and Góes-Filho, 1982;INEA, 2011;Martins et al., 2012;Silva and Dereczynski, 2014).Geopolitically, it is subdivided into regions (Baixadas Litonâneas, Centro-Sul Fluminense, Costa Verde, Médio Paraíba, Metropolitana, Nordeste Fluminense, Norte Fluminense, and Serrana), which are used by the government to adopt health policies (IBGE, 2018).This geopolitical division directly influences the distribution of antivenoms, which is based mainly on snakebite data from the Notifiable Diseases Information System (Brasil, 2022).

Mapping
Geographic coordinates were obtained from the Instituto Brasileiro de Geografia e Estatística (IBGE, 2018) and gazetteers (USBGN and CidVil).Records lacking precise locality data were georeferenced using municipality centroids.We extracted historical climatic data on temperature, phytophysiognomy, and elevation for each point record using a 30 arc-second spatial resolution from WorldClim version 2.1 for the period between 1970 and 2000 (Fick and Hijmans, 2017) to construct an environmental profile of geographical occurrence for each species.Geographic distribution maps were produced in QGIS version 3.26 (QGIS Development Team, 2020).

Morphological data
Pholidosis, morphometric, and color pattern data were collected from preserved voucher specimens (Supplementary file I).We examined 202 specimens from 12 species and five genera.Terminology for scale counts and qualitative traits follows Campbell and Lamar (2004) for Viperidae, Thomas (1976) for Colubridae, and Silva Junior et al. (2016) for Elapidae.We measured head width (HW) using a digital caliper to the nearest 0.1 mm.We measured the snout-vent length (SVL) and tail length (TL) using flexible ruler (Dowling, 1951;Thomas, 1976;Francini et al., 1990).We determined the sex of specimens through a small incision at the base of the tail or by the presence of an everted hemipenis (Yuki, 1994).

Snakebites in Rio de Janeiro
We obtained data on snakebites in Rio de Janeiro state from 2001 to 2019 by accessing the Notifiable Diseases Information System (Brasil, 2022).We accessed, organized, and computed the data in an electronic spreadsheet and included cases of patients who recovered (with or without sequelae) and patients who died from the snakebites.
This study used exclusively public data provided by the Notifiable Diseases Information System (Brasil, 2022).

Results
We recorded 2,123 specimens collected in Rio de Janeiro state between 1820 and 2021, with a richness of five genera and 13 species distributed in three families (Figures 1-8).

Taxonomy
We found that the distinction between B. jararacussu and B. jararaca and between P. patagoniensis and P. olfersii Crotalus durissus subspecies were impossible to identify due to the large overlap of diagnostic characters available to date.We were also unable to find characters that would allow us to identify the subspecies of C. durissus.A few specimens fit the diagnosis of C. d. terrificus, most of them showing diagnostic characters for two or more subspecies, such as the presence or absence of contrasting diamond-shaped spots on the last third of the body and short versus long white paravertebral lines on the first third of the body (Figure 5).
Philodryas patagoniensis has a brown ventral scale border (vs white in P. olfersii); a brown head (vs.bronze in P. olfersii); a blackened dorsal shield border on the first third of the body (vs.whitish in P. olfersii); the background color of the belly darkening towards the tail (vs.uniform in P. olfersii); postocular stripe absent (vs.present in P. olfersii); dorsal background color with black spots (vs.uniform in P. olfersii) (Figure 7).We had no issue identifying the other species (Figures 4-6).

Discussion
Identifying the causative agent of envenomation and knowing its geographic distribution and habitat use are helpful to understanding human-snake conflict dynamics and providing adequate health facilities (Molesworth et al., 2003;Kasturiratne et al., 2008;Gutiérrez, 2012;Gutiérrez et al., 2017).Bothrops and Micrurus genera were recorded in all eight Rio de Janeiro mesoregions.Bothropic and elapid antivenoms can be found in antivenom care centers in all eight mesoregions; however, bothropic antivenom is found in 22 centers, while elapid antivenom is found in 13 centers.
Our data revealed a weak combination of characters recommended to distinguish Bothrops jararaca from B. jararacussu, mainly due to the wide overlap of the diagnostic character "area occupied by interspaces" about 1.5 to 2 times larger than (in B. jararaca) or subequal (in B. jararacussu) to dorsal blotches (Campbell and Lamar, Snakes and ophidism in Rio de Janeiro 2004).Thus, we provide a more apparent distinction between Bothrops species from Rio de Janeiro state.Bothrops jararaca was the most recorded species in Rio de Janeiro, which agrees with other studies [e.g., Pontes et al. (2009)].However, we found B. jararacussu in all mesoregions, and species-level identification is indicated.Therefore, we recommend that clinical studies with good experimental design be conducted to evaluate the use of bothropiccrotalic antivenom serum and heparin in the treatment of bites by B. jararacussu (dos-Santos et al., 1992;Rostelato-Ferreira et al., 2010).
Crotalus durissus is restricted to the mesoregions of Médio Paraíba, Centro-Sul Fluminense, and Serrana, and therefore, crotalic antivenom must not be lacking in these mesoregions.Crotalus durissus collilineatus venom is predominantly positive for crotamine, while C. d. cascavella is negative and C. d. terrificus is mainly negative (Tasima et al., 2020).Considering that the Brazilian crotalic antivenom may be deficient, we recommend that further studies be conducted to better characterize the geographic and morphological boundaries of rattlesnake subspecies in Brazil to thereby help ensure the recommended use of both crotamine-positive and crotamine-negative entire venoms to produce the antivenom (Tasima et al., 2020).
Crotalus durissus population may have first expanded towards the state's coast during the 1950s (Duarte and Menezes, 2013) (Figure 5).The Médio Paraíba region, where the species is now most abundant, housed coffee production during the 19 th century and later became a pasture area, which may help explain the expansion of Crotalus (Bastos et al., 2005).After all, the significant floods between 1950 and 1967 in the Rio Preto region may have contributed to the dispersal of rattlesnakes in deforested areas of some other municipalities, such as Valença (Bastos et al., 2005).Here, we provide the first record of rattlesnakes in the Serrana mesoregion, supporting the hypothesis of the geographic expansion of the species in the state.Assigning individuals of C. durissus to any subspecies is impossible because of the high level of overlap of the following diagnostic traits: absence or presence of contrasting diamond-shaped spots on the last third of the body and short or long paravertebral white lines on the first third of the back [see Harris and Simmons (1976); Amaral (1977)].A deep taxonomic review to set the Crotalus durissus limits is required.Moreover, the epidemiological surveillance by monitoring and evaluating the geographic expansion of the species needs to consider the first records of the species for the Serrana region, which would facilitate interventions against snakebites.
We improved the distinction between live and preserved specimens of P. olfersii and P. patagoniensis previously reported in the literature [e.g., Peters and Orejas-Miranda (1970); Thomas (1976)].Philodryas olfersii venom has proteolytic, hemorrhagic, fibrinogenolytic, and edematogenic activities (Assakura et al., 1992).Bites by specimens of Philodryas may manifest pain, heat, erythema, edema, and ecchymosis and are sometimes treated with bothropic antivenom (Correia et al., 2010;Medeiros et al., 2010).Philodryas spp.were recorded in all mesoregions, except the Noroeste Fluminense.Despite the previous occurrence of some severe cases, including one death (Salomão and Di Bernardo, 1995), no further concern is currently necessary.
Given the geographic distribution of M. lemniscatus, the antivenom service centers in the coastal, Metropolitana, and Norte Fluminense regions should be equipped with artificial respiration and elapid antivenom.The elapid antivenom produced in Brazil only with M. frontalis and M. corallinus venoms can neutralize those of M. frontalis, M. corallinus, and M. spixii but not well M. altirostris and M. lemniscatus (Tanaka et al., 2010).Considering the need for improvement of elapid antivenom, the complementary use of anticholinesterase to treat envenomations by M. lemniscatus carvalhoi but not those of M. corallinus (Coelho et al., 1992;Vital Brazil and Vieira, 1996) should quickly become part of clinical studies for future uses in snakebite therapy.In severe bites by M. lemniscatus, in which the venom acts postsynaptically, anticholinesterases may be useful as an ancillary measure (particularly when antivenom is unavailable or insufficient) and for patients treated with high doses of antivenom but with no improvement of paralysis or with delayed recovery (Bucaretchi et al., 2006).Ciscotto et al. (2011) also highlight the importance of using M. frontalis venom to produce heterologous elapid antivenom and suggest replacing M. corallinus venom with M. ibiboboca venom to enhance antivenom since M. ibiboboca antivenom was able to cross-react more efficiently with other venom proteins.Despite these results, additional experimental and clinical studies are needed to identify an eventual deficiency of the Brazilian Coralsnake Antivenom to neutralize Micrurus spp.envenoming efficiently.Nevertheless, considering the interspecific variation of the venoms, the discussion to produce Micrurus antivenoms (Ciscotto et al., 2011) and that we found M. corallinus and M. lemniscatus in sympatry, their morphological distinction becomes relevant for species-specific attention.
Lachesis is historically recorded in six Rio de Janeiro mesoregions, with no record in the last 18 years.Therefore, Lachesis antivenom is unavailable in the state.Epidemiological data indicate that Lachesis is responsible for 4% of snakebites in Brazil (about 1,450 accidents per year), with an average mortality rate of 1.04%, i.e., 15 people per year (Silva et al., 2015).Considering the rarity of L. muta in Rio de Janeiro, the question remains whether it is worthwhile for the state of Rio de Janeiro to house Lachesis antivenom in some mesoregions where the species may occur.
We surveyed 7,483 records of snakebites in Rio de Janeiro between 2001 and 2019, with an annual average of 393.84, therefore updating the data on snakebites in the state (Machado and Lemos, 2016;Bochner and Struchiner, 2004.The highest number of cases occurred in the Serrana region (n = 2,905), followed by Metropolitana (n = 1,948) and Costa Verde (n = 1,615), and the lowest number of cases occurred in the Norte Fluminense (n = 473) mesoregion (Supplementary figure II).The Serrana mesoregion is the largest pole of family agriculture in Rio de Janeiro and the central pole of olericulture in the state, responsible for supplying the metropolitan region (Moreira et al., 2002).In this region, intensive family farming with little mechanization predominates, which may increase the encounter between humans and snakes.This observation alone should not be considered sufficient to understand the serious problem and the failure to combat snakebites, but it can be a starting point for devising environmental education strategies to mitigate the problem.
It should also be noted that snake populations are affected by multiple factors, such as habitat destruction, decrease or increase in prey populations, changes in climate variables, and the use of agrochemicals (Gutiérrez, 2020).We recommend further epidemiological research to address the causes that preclude the effective combating of snakebites.
We improve the identification of venomous snake species, better delimit their geographic distribution, and update snakebite cases, thus providing greater precision in snakebite care in Brazil.Mapping comprehensive datasets is imperative for understanding human-animal conflict dynamics (such as vulnerability to medically important snakebites) and for providing background information needed to enable adequate health facilities as well as the provision of antivenom and other therapeutic innovations (Molesworth et al., 2003;Kasturiratne et al., 2008;Gutiérrez et al., 2010;Gutiérrez, 2012).The established centers against ophidism must provide antivenom supplies considering the snake genus distribution, epidemiology data on snakebites and professionals with expertise in animal venom.A good relationship between humans and snakes requires public awareness and education (Gouveia et al., 2015).Accordingly, a booklet disseminating our findings is in preparation to be distributed in snakebite care centers in Rio de Janeiro state.

Figure 1 .
Figure 1.Distribution map (A) and pictures of live specimens of Bothrops jararaca showing a range of colour patterns from Niterói (B), Miguel Pereira (C-D), State of Rio de Janeiro (E), Petrópolis (F), siblings with different dorsal colour patterns same mother from Guapimirim (G, H).Photos by Breno Hamdan (B, C, F, G, H), Gustavo Cunha (D, E).

Figure 2 .
Figure 2. Distribution map (A) and pictures of live specimens of Bothrops jararacussu showing different colour patterns from the State of Rio de Janeiro (B, D-E), Cachoeiras de Macacu (C), and Rio de Janeiro (F).Photos by Claudio Machado (B, D, E) and Breno Hamdan (C, F).

Figure 6 .
Figure 6.Distribution map (A) and pictures of preserved specimens of Lachesis muta from Santa Maria Madalena (B, C), Campos (D).Photos by Gustavo Cunha (B, C, D).