Body length of Hylodes cf . ornatus and Lithobates catesbeianus tadpoles , depigmentation of mouthparts , and presence of Batrachochytrium dendrobatidis are related

A fungal pathogen Batrachochytrium dendrobatidis (Bd), which can cause morbidity and death of anurans, has affected amphibian populations on a worldwide basis. Availability of pure cultures of Bd isolates is essential for experimental studies to understand the ecology of this pathogen. We evaluated the relationships of body length of Hylodes cf. ornatus and Lithobates catesbeianus tadpoles to depigmentation of mouthparts and determined if dekeratinization indicated an infection by Batrachochytrium dendrobatidis. A strong association existed for both species, one from South America (Brazil: São Paulo) and one from North America (USA: Maine). We believe it prudent not to kill adult amphibians if avoidable, thus obtaining tissue for isolating Bd from tadpoles is reasonable because infected specimens of some species can be selectively collected based on depigmentation of mouthparts.


Introduction
A chytrid fungus, Batrachochytrium dendrobatidis (Longcore et al., 1999) (hereafter Bd) infects keratinized tissue of amphibians, and can be lethal to post-metamorphic individuals (Berger et al., 1998;Lamirande and Nichols, 2002).This pathogen is associated with worldwide population declines of amphibian species (Berger et al., 1998;Daszak et al., 1999;Stuart et al., 2004;Lips et al., 2006).Although Bd skin infections can kill metamorphs and adults, most larvae (= tadpoles) survive when infected and may serve as reservoirs of infective zoospores (Lips, 1999;Lips et al., 2003Lips et al., , 2004;;Daszak et al., 2004).Important research topics for this pathogen are geographic origin and circumstances under which it arose, relative pathogenicity of different strains (Berger et al., 2005), and genetic differences among strains; research of these topics requires pure cultures of Bd (Voyles et al., 2009(Voyles et al., , 2010)).Most Bd isolates now in culture were isolated from haphazardly contributed animals, whereas isolates from specific geographic areas are needed.If Bd, which in adults is not detectable in the field, can be identified in tadpoles in the field, fewer animals will need to be euthanized and examined to obtain infected tissue for isolation (Longcore et al., 1999).Isolation of Bd from freshly euthanized larvae is also relatively easy, because mouthparts are seldom contaminated by other fungi or bacteria (J.E.Longcore, personal observation).

Material and Methods
While in North America (Orono, Maine) LFT participated in sampling a population of bullfrog tadpoles (Lithobates catesbeianus; Anura, Ranidae) with a history of Bd infections (Longcore et al., 2007).With this effort in mind, LFT returned to Brazil and with CAV sampled a population of torrent frogs (Hylodes cf.ornatus; Anura, Hylodidae) in southeastern Brazil to determine if tadpole mouthparts were depigmented and similarly infected with Bd.Our purpose was to evaluate tadpole mouthpart data regarding pigmentation and status of Bd infection for two anuran species with distant phylogenetic relationships and from different eco-regions.Moreover, we examined tadpoles in the field to confirm the ability to selectively collect infected tadpoles on the basis of depigmentation status of mouthparts.
In Brazil we collected 243 Hylodes cf.ornatus tadpoles from March to May 2010 in the Reserva Biológica Serra do Japi, county of Jundiaí, state of São Paulo (23° 14' 57.8" S; 46° 56' 59.4" W; about 1100 m a.s.l.).We separated these individuals into 5, 5-mm body length classes (i.e., <5.1, 5.1-10, 10.1-15, 15.1-20, and 20.1-25 mm; Table 1).We examined the mouthparts of these specimens in the laboratory with a stereo-microscope (400×) for signs of Bd infections (i.e., lack of keratin in the tooth rows and jaw sheaths).We considered structures as depigmented if mouthparts were brown or pale and lacked the dark melanin of keratin on teeth rows or jaw sheaths and appeared light sepia or nearly clear (Rachowicz and Vredenburg, 2004;Knapp and Morgan, 2006).To confirm that depigmentation was associated with Bd infection, we killed all sub-sampled individuals and observed mouthparts under a microscope (400×).

Results
Out of the 8 bullfrog tadpoles in the North American (Maine) sample, jaw sheaths for specimen numbers 7, 19 and 21 were clearly depigmented and specimen 14 was questionable with only slight depigmentation at the base of a jaw sheath; no depigmentation was observed for specimens 28, 35, 42, and 49.Three out of 4 specimens with depigmented jaw sheaths, but not specimen 14, were infected with Bd; none of the 4 specimens with intact jaw sheaths were infected as determined by microscopic examination of the jaw sheaths at 400 and 1000×.We isolated pure cultures from tadpoles 10 and 19.We separated individuals into 5, 5 mm body length classes (i.e., 14-20, 21-27, 28-34, 35-41, and 42-47 mm; Table 1) and regressed the percent depigmented on length class of tadpoles.Presence of Bd was closely associated with depigmentation (r 2 = 0.769, df = 4, P = 0.0507) (Figure 1).For the South American (Brazil) sample of torrent frogs 113 of 243 (46.5%) tadpoles had depigmented mouthparts (Table 1) and the larger tadpole classes had higher percentages of depigmented mouthparts.A linear regression (alpha value of P < 0.05) confirmed this relationship (r 2 = 0.903; df = 4; P = 0.013) (Figure 1).Under a microscope we examined 92 specimens with depigmented mouthparts; 87 (94.5%) were positive for Bd infection.From three (3.4%) of these 87 tadpoles, we isolated Bd with standard methods (Longcore and Berger, 2000).

Discussion
Although some researchers have reported that mouthpart dekeratinization is not a good predictor of infection by Bd (Padgett-Flohr and Goble, 2007), our data from L. catesbeianus and H. cf.ornatus support findings of Fellers et al. (2001), who reported that 67% of Rana muscosa tadpoles with missing or abnormally dekeratinized mouthparts were infected with Bd; Knapp and Morgan (2006), who reported that depigmentation of R. muscosa tadpole upper jaw sheaths was strongly with Bd infections (89%, 50 of 56) and Drake et al. (2007), who reported that 70.3% of tadpoles from 5 frog and 3 toad species in southeastern United States had deformed jaw sheaths with dekeratinization the most common (50.3%) defect.Mouthpart deformations were explained mostly by taxonomic group (family) and presence of Bd.
Because keratinized mouthparts of amphibians atrophy during metamorphosis after Gosner stage 40 (Gosner, 1960) and jaw sheaths fall off after tooth rows are gone (Thibaudeau and Altig, 1988) age class of tadpoles is important in finding Bd.Rachowicz (2002) observed that variability of pigmentation in keratinized cells in tooth rows and jaw sheaths is also related to seasonal changes in temperature.Because loss of labial tooth rows occurred in L. catesbeianus tadpoles at sites polluted by coal ash (Rowe et al., 1996) and jaw sheaths are the last keratinized structures to be shed (Thibaudeau and Altig, 1988), depigmentation of jaw sheaths may be more reliable in indicating Bd infections than depigmentation or loss of other mouthparts.We believe that, at least for some species, larvae can be evaluated with a hand lens and only those with dekeratinized erosions of the upper jaw sheaths, i.e., the ones most likely to be infected with Bd, can then be brought into the laboratory for isolation procedures.We have been successful in finding and isolating Bd from L. catesbeianus and H. cf.ornatus and 10 other Brazilian amphibian species (L.F.Toledo, personal observation) after screening and selecting larvae by this method.
This method may not work for some amphibian species and will be more informative at certain Gosner stages; we found that only the more advanced age classes had dekeratinized structures and were infected with Bd.Amphibians that have 2-year larval stages such as the bullfrog (Lithobates catesbeianus) (Albright, 1999) and Mountain Yellow-Legged Frog (Rana muscosa) (Rachowicz, 2002) are exposed to Bd zoospores for substantially longer periods than species that undergo metamorphosis during their first year.The larger and older an individual is, the longer the period that it would be exposed to infective zoospores in an aquatic environment.These findings are congruent with those from Smith et al. (2007), where infected tadpoles were larger than uninfected ones.However, our data is complementary as we show a gradual increase in the infection prevalence with the increase in body size, similarly as recently reported for Telmatobius jelskii in Peru (Catenazzi et al., 2013).Lower rates of infection of Bd in younger tadpoles also may be related to the smaller area of keratinized cells.Thus, exposure of larger tadpoles (either single or multiple exposure times) is greater than that of younger individuals.This method of selecting larvae for isolation of Bd will miss early infections because recently infected individuals may not have visibly depigmented mouthparts.
Our results expand knowledge for a South American frog species and corroborate observations for a wild population of bullfrogs (L.catesbeianus) in the northern United States of America; both showed a similar depigmentation/ infection pattern (Figure 1) of smaller tadpoles being less frequently infected than larger tadpoles and Bd infection corresponding with depigmentation of keratinized mouthparts.This similarity of response to Bd infections between phylogenetically distant species (Ranidae vs. Hylodidae) and between species of different eco-regions (Neotropics vs. Nearctics) suggests a consistent and perhaps a universal pattern.The high association of upper jaw sheath dekeratinization with Bd infection can be used to select tadpoles to examine for Bd infection, which can reduce the number of specimens killed to obtain cultures.This is an important issue as amphibians are the most threatened vertebrates in the world (Barnosky et al., 2011).

Table 1 .
Number and percentage of tadpoles with depigmented mouthparts from a population of Hylodes cf.ornatus in Jundiaí state of Sao Paulo, Brazil collected March-May 2010 and from a population of Lithobates catesbeianus tadpoles collected during October 2009 from Crocker Pond, Bethel, Maine, USA.