Ecology of ticks in a taxocenosis of snakes from the Serra do Mendanha , Rio de Janeiro , Brazil , with new host records

We studied the ecology of ticks found in different species of a taxocenosis of snakes from the Serra do Mendanha, an area of Atlantic Rainforest located in the state of Rio de Janeiro, southeastern Brazil. Snakes were sampled monthly in the field during a period of 48 months. The specific identity of the hosts and their parasites, the number of parasites, and and snout-vent length and body mass of each host were recorded. A total of 25% of the species of snakes in the area were parasitized by ticks (larvae, nymphs and adult females) Amblyomma rotundatum Koch, 1844 was the predominant parasite species. The infestation parameters varied among the species of snakes sampled, with the highest prevalence of A. rotundatum being observed in the viperid Bothrops jararaca (Wied, 1824) (71.4%), followed by the colubrids Xenodon neuwiedii Günther, 1820 (33%), Chironius laevicollis (Wied, 1824) and Spilotes pullatus (Linnaeus, 1758) (both with 22%). The latter three species also showed the highest rates of infestation by A. rotundatum. The results of the present study suggest that a combination of skin shedding, habitat of the host, type of scale and pattern of scale distribution on the body of the host can influence the degree to which a given species is parasitized by ticks

The ticks found in sampled snakes belong to two species: A. rotundatum and one unidentified species of Amblyomma Koch, 1844.The Brazilian Amblyomma are characterized by the presence of a conspicuous scutum with marginal eyes present; anal fault posterior to the anus and gnathosoma long (palps as long as the hypostome and cheliceras larger than base of gnathosoma, and second palpal article much longer than remaining articles) (ARAGÃO & FONSECA 1961, SERRA-FREIRE 2001).Even though most Acari are apometabolic, some characters used in their diagnosis may vary depending on sex and developmental stage.Consequently, in order to identify adult females of A. rotundatum, we evaluated the presence of two short and rounded spurs on each of coxa II-VI; base of gnathosoma without prominent posterolateral angles; scutum with small to medium-sized punctations, more abundant laterally; hypostome with dentition 3/3 (with three rows of teeth on each side).Nymphs of A. rotundatum have two spurs on each of coxa II and III, internal spur small; hypostome rounded on anterior portion, hypostomal dentition 3/3; palps elongate; dorsal base of gnathosoma subquadrangular with rounded (not pointed) lateral margins; scutum with deep punctations along lateral margins (KEIRANS & DURDEN 1998).The larvae are characterized by having coxa I with two spurs and each of coxa II and III with one spur; eyes with convex cornea; lateral portion of gnathosomal base rounded in dorsal view; sixth festoon wider than long; opisthosoma without dorsal rows of spines from the scutum to festoons III and VI; capsule of Haller's organ with axial and paraxial setae subequal, and antiaxial seta small, all of non-porous structure (AMORIM & SERRA-FREIRE 1999, AMORIM et al. 2001).
Amblyomma rotundatum was the most common tick on parasitized snakes, corresponding to 99.1% (n = 108) of the total sample size.Out of all ticks found on snakes, 55% (n = 60) were nymphs, 2.2% (n = 46) were larvae, and 1.8% (n = 2) were adult females (Figs 1-3); the life stage of one (n =1) individual (0.9%) could not be determined.The latter specimen seems to belong to a different species which we were not able to identify, and is herein referred to as Amblyomma sp.Only the terrestrial and arboreal snakes were parasitized.Parasites were not found on fossorial, cryptozoic, semi-aquatic or aquatic snakes (Tab.I).Most ticks (89%, n = 97) were found between the dorsal scales of the snakes; 5.5% (n = 6) were found between the scales on the top of the head; and 5.5% (n = 6) were found underneath the ventral scales.The viperid B. jararaca showed the highest incidence of parasitism (prevalence = 71.4%),followed by the colubrids X. neuwiedii (33%), C. laevicollis and S. pullatus (both 22.2%) (we did not consider B. jararacussu here because of the small [n = 2] sample size).The latter three species also showed the highest mean values of infestation (X.neuwiedii:16.7 ± 11.2; C. laevicollis:14.5 ± 10.6; S. pullatus:11.0± 15.6).One individual of C. laevicollis which was maintained in captivity for four days had all of its ticks (n=21) eliminated after shading the old skin.
We observed the highest rates of parasitism between the months of July and October.This period coincides partially with the lowest rates of snake findings and the lowest incidence of rainfall in the region (Figs 4 and 5) (PONTES & ROCHA 2008).

DISCUSSION
Our results showed that relatively few species of snakes sampled from the Serra do Mendanha were parasitized by ticks (only 25%), and in the great majority of the cases, by only one species (A.rotundatum).This suggests a broad utilization of different host species by only one species of ectoparasite along the year in this area.A. rotundatum is broadly distributed in the country, and has been found parasitizing snakes in different regions of Brazil (e.g. ARAGÃO 1936, DANTAS-TORRES et al. 2005, EVANS et al. 2000, LABRUNA et al. 2005).
Because the only additional species of Amblyomma found in this study was represented by a larva, we were not able to determine whether it belongs to Amblyomma dissimile Koch, 1844 (a species similar to A. rotundatum).Amblyomma rotundatum and A. dissimile are common parasites of amphibians and reptiles, and are both distributed in the Nearctic and Neotropical regions.Amblyomma dissimile, however, has a wider range of hosts (e.g.DOOS et al. 1974, SERRA-FREIRE 2001, BOTELHO et al. 2002).Data by ARAGÃO (1936) indicates that, even though these two species may have sympatric distribution, A. dimissile favors higher temperatures, what seems to be confirmed by reports from EVANS et al. (2000).The biological characteristics of these two species allow for high adaptability to anthropic environments, but also indicate some variation in their ability to complete their respective biological cycles, particularly as a function of the kinds of hosts available (TEIXEIRA et al. 2003).Even though there are records of males of A. rotundatum (KEIRANS & OLIVER 1993, LABRUNA et al. 2005), the low number of specimens recorded (three) and the teratogenic characteristics of at least one of them, lead us to question the importance of this stage in the life cycle of this species, believed to be parthenogenetic since 1912 (e.g.ARAGÃO 1912, ARAGÃO 1936, OLIVER 1989, SERRA-FREIRE 2001, FREITAS et al. 2002, HANSON et al. 2007).
The results of this study suggest some variation in the levels of infestation displayed by the six species of parasitized snakes.These differences are probably a function of two factors: the differing patterns of scale distribution on the bodies of these species (which may or may not favor tick attachment); and the microhabitat utilized by the hosts.Studies performed on lizards suggest that the patterns of scale distribution are an important factor determining the degree to which different species are parasitized by the larva of Eutrombicula alfreddugesi (Oudemans, 1910) (Trombiculidae) (CARVALHO et al. 2006, CUNHA-BARROS & ROCHA 2000, CUNHA-BARROS et al. 2003, ROCHA et al. 2008).On the other hand, it has also been shown that the type of habitat more frequently utilized by different species of lizards of the same genus (Mabuya) affect the degree to which each species is parasitised by E. alfreddugesi (CUNHA-BARROS & ROCHA 1995).Three of the species parasitized in this study (C.fuscus, C. laevicollis e S. pullatus) are semi-arboreal, whereas three others (X.neuwiedii, B. jararaca e B. jararacussu) are predominantly terrestrial (MARQUES et al. 2001, CAMPBELL & LAMAR 2004, PONTES & ROCHA 2008).The latter two species have keeled scales, contrasting with the remaining species, which have smooth scales.Despite these considerations, however, we cannot rule out the possibility of an association between the observed patterns of infestation and the seasonal patterns of snake activity in the region, which is higher in periods when precipitation is higher (Fig. 5).This could explain the high indexes of infestation ob-

5
served for X. neuwiedii, which is most active during cold and dray periods, coinciding with the time when the larvae and nymphs of the ticks are dispersing (PONTES & ROCHA 2008).
The absence ticks on snakes with cryptozoic and fossorial habits (Echinanthera affinis (Günther, 1858) (Colubridae), Echinanthera cephalostriata Di Bernardo, 1996 (Colubridae) and Micrurus corallinus (Merrem, 1820) (Elapidae) could be in part explained by the friction between the body and the substrate when the snake is burying, together with the smooth and juxtaposed scale pattern, what seems to make it difficult for the ticks to attach and remain attached.The absence of ticks on aquatic and semi-aquatic snakes may be explained by the fact that these parasites are terrestrial, and thus not adapted to aquatic environments.
The results of the present study suggest that a combination of skin shedding, type of scale, pattern of scale distribution on the body, and habitat can influence the degree to which a given species is parasitized by ticks.