Water collection by the body in a viperid snake, Bothropsmoojenz

Denis V. Andrade1, Augusto S. Abe2

Departamento de Zoologia, Universidade Estadual Paulista, c. p. 178, Rio aaro 13506-900, São Paulo, Brasilte-mail: denis@rc.unesp.br2e-mail: asabe@rc.unesp.br

Abstract. We describe a previously unreported behavior for water collection in juveniles of a neotropica1 viperid

snake, Bothrops moojeni. When sprayed over, this snake displays a stereotyped coiling, bringing its body loops in

elose contact with each other, so that water is retained between the loops and over the body surface. This water is

continuously ingested during and after its collection. The functional significance of the water collecting behavior

is suggested to be related with the acquisition of water from short rainfalls, and with the special elimatic and°geologic conditions of B. moojeni habitaL Rates of evaporative water loss did not differ between juvenile and

adult snakes, but since juveniles have a greater surface-to-volume ratio, they were significantly more sensitive todesiccation than the adults.

Introduction

One of the biggest chalIenges faced by animaIs living on Iand is the potential riskof dehydration (Dejours, 1988; Little, 1990). According1y, in terrestrial animais, theacquisition and conservation of water are pivotal to the maintenance of an adequate osmoticequilibrium (Schmidt-NieIsen, 1979; Kirschner, 1991). Reptiles have met these basicrequirements of terrestriality very successfulIy, as evidenced by their impressive radiationin almost alI terrestrial habitats, inc1udingdeserts (Pianka, 1986; Zug, 1993; Pough et a1.,1998). Occupation of water stressing habitats requires mechanisms for decreasing waterloss and increasing water gain (Schmidt-NieIsen, 1979; Kirschner, 1991). Mechanismsused to prevent excessive water loss, however, have received much more attention thanthose invoIvedwith the improvement of water acquisition (see Mautz, 1982; Minich, 1982;Lil1ywhiteand Sanmartino, 1993).

The water used by reptiIes may come from drinking, from water contained in foodstuffs,from absorption through the integument, and from oxidation (CIoudsIey-Thompson, 1971;

@ Koninklijke Brill NV. Leiden. 2000 Amphibia-Reptilia 21: 485492

486 Denis V. Andrade, Augusto S. Abe

Minnich, 1982).The importance of such routes varies depending on the species consideredand environmental contingencies, but in most species drinking appears to be an especiallyimportant avenue for water influx since the scaly keratinized reptilian skin largely preventswater flux (Minich, 1982; Zug, 1993). Drinking, however, is not always possible becausewater availability may be temporaly and/or spatially restricted. Under such conditions,°drinking may involve specialized morphology and behavior aimed to improve watercollection and ingestion (e.g., Bentley and BIummer, 1962; Lasiewski and Bartholomew,1969; Medica et aI., 1980; Ashton and Johnson, 1998). Herein, we report on an unusualwater collecting behavior displayed by juvenile Bothrops moojeni Hoge, 1966 whensprayed with water. Wefurther correlate the occurrence of such behavior with the hydrationstate of the snakes and with a possible induction of water search. Finally, we compare ratesof evaporative water loss (EWL) in juvenile and adult B. moojeni to determine whethervulnerability to desiccation varies ontogenetically in this species..'

. .

Material and methods

Animais. Most observations were made on juvenile snakes (SVL :::; 450 mm) bom in captivity from females

collected at the municipalities of Araçatuba (21°38'S; 50025'W) and Pereira Barreto (20038'S; 51°06'W), São

Paulo state, southeastem Brazil. AduIt snakes collected at the municipality of Rio aaro (22°23'S; 4~32'W)also in São PáUlo state, southeastem Brazil, were used in the EWL experiments. The animais were maintained

individually (subaduIt and aduIt snakes) or in groups Guveniles) in wood cages (27 x 41 x 35 em) covered withcardboard and provided with a sliding gIass doar and holes for ventilation. Temperature of maintenance, as wellas during experiments, was kept constant at 30:1::2°C. Snakes were fed mice every other week and had free access

to water during the time before the experiments, except when otherwise noted. By the time of the experiments,the snakes appeared to be used to the experimenter and displayed no signs of disturbance at human presence. Forali experiments we fasted the snakes for two weeks before experimentation and animais involved with skin moltwere not used.

Behavioralobservations. For behavioral observations, the snakes were placed in individual cages (described

above) and left without water for a week before experimentation. We then gently removed the front gIass cover of

the cage and, using a manual plant mister, sprinkled droplets of water ali over the snake's body. At each triaI, wesprayed water over the snakes uninterruptedly for 10-20 seconds. AlI behavioral responses of the snakes, duringand following water spray, were recorded. Some sequences were documented photographically and others werevideo-recorded. These video-recorded sequences were later analyzed frarne-by-frame to capture postura! details.

To test the influence of hydration state on water collecting behavior, we compared its frequency of itsoccurrence in individual juvenile B. moojeni with free access to water until experimentation as opposed toindividuais that were prevented from drinking for two weeks before being sprayed with water. The protocolutilized in this set of experiments was the same as described above.

We finally investigated whether water collecting behavior could provoke a water-induced chemosensory search

analogous to the strike-induced chemosensory search reported for some snake species (see Ford and Burghardt,1993). For this purpose, we applied the protocol described above, but immediately after the snm had beensprayed with water we introduced a plastic opaque water bowl (9 x 7 x 4 em) within the cage, as far as possiblefrom where the snake was. Then, we monitored the snake's behavior for two hours. We recorded the time lag

to find the water bowl and drinking behavior (if it occurred), tongue-flick frequency, and the overall movementof the snake within the cage. AlI these variables were later compared with a control group. This control groupreceived the same treatment as the experimental group, but was not sprinkled with water (the plant mister wasused empty). For baseline purpose, ali individual snakes had their tongue-flick frequency recorded during the 20

minutes preceding experimentation. The one minute tongue-flick rates were averaged to give a grand mean foreach individual snake.

Water coUection by the body in snakes 487

Evaporative water loss. Evaporative water loss rates were measured indirectly through change in body mass (see

Shoemaker et ai., 1992) in adult and juvenile snakes subjected to a desiccating environment for five consecutive

hours. The snakes were weighed (to the nearest 0.001 g) and confined singly in hermetically cIosed plasticchambers (2-5 liters) inside a temperature controlled cabinet (at 25 :I: 1°C). To promote quiescence, the animais

were left undisturbed in this condition for two hours. Thereafter, we began to ventilate the chamber with dry air

using an air pump (Ametek RI; Sub-Sampler 1.0, Sable Systems) connected to the chamber by plastic tubes.

The flow rate of the pump was adjusted to renew 44% of the chamber air volume every minute. Water vapor ofthe incurrent airflow was removed by passage through a tube fiUed with silica gel. At the end of the experiment

the snakes were re-weighed, anesthetized with C02 (see Wang et ai., 1993), measured and sexed. If the animal

did not void during the time of the experiment, we assumed that ali change in body mass were due to EWL.

We express EWL as a percentage ofbody mass lost in an hour, and in mg.cm 2.hr 1 that represents the rate of

EWL per unit area of skin. Snake surface area was estimated from body mass foUowing the exponential equationprovided by Gans et ai. (1968).

Data analysis. All means are presented with their respective standard deviations. Differences between the

probability of a given snake to perform the water coUecting behavior and its hydration state were evaluated by the

Fisher Exact Test. This same test was aIso employed to identify a possible difference between the probability of a

given snake finding water and the occurrence, or no!, of the water collecting behavior. EWL rates were compared

between juvenile and adult snakes by Student's f-test. The difference between tongue-flick frequency in snakes

that coUected water with their bodies and those that did no!, were aIso evaluated employing Student's f-test. 10

those cases in which data variation compromised the premises of normality and homoscedasticity of the f-teS!,

we employed the Mann-Whitney Sum test. All statistical analyses foUowed Sokal and Rolf's (1986) procedures.

Differences were considered significant at P ~ 0.05.

Results

Water collecting behavior

During water collection, the snake resting on ground begins to coil immediately after theinitiation of water sprinkling. The coiling begins headfirst and progresses in an antero-posterior direction, with the head always tumed to the center of the coil and kept in tightcontact with the body (fig. 1). Body loops are brought together in close contact with eachother, so water is retained between them. Moreover, water droplets may also adhere to thesnake's body surface. Such water is continuously ingested during water collection (fig. 2a)with the head moving in an orientation contrary to scales overlap. The water collectingbehavior lasts from 10 to 30 seconds, but following it, the snake continues to ingest thewater collected, for as long as 2-5 niinutes. During this phase, the snake begins to loosen thebody coils and pass its lips slowly over their entire body (see figo2b) - short movementsof the mouth can be perceived at this stage. IndividuaIs of B. moojeni involved in the watercollecting behavior never displayed aggressive/defensive behaviors, the head was neverelevated over the body facing the experimenter, and the snakes did not try to strike, evenwhen disturbed by a finger touch.

Following water collection, B. moojeni may initiate an exploratory behavior thatultimately can lead to water discovery.Snakes that collected water using the body presenteda significantly higher tongue-flick frequency (8.13:f: 7.91; n = 9) than those that did not(0.15 :f:0.29; n = 8) (P = 0.002). Tongue-flick rate was also higher than baseline values

488 Denis V. Andrade, Augusto S. Abe

.. .

Figure 1. Schematic view, from upper left to right bellow, depicting the postoral sequence displayed by juvenites

Bothros moojeni during the collection oí water through body coiting. Note the orientation oí the head, tumed to

the center oí the coit, and moved in a direction contrary to the sca1es overlap. Arrows indicate the directions oí

body movements.

(0.053 :!: 0.23; n = 19) in snakes dispIaying the water collecting behavior (P < 0.001),but this was not the case for the controI group (P = 0.23).

Snakes that collected water using the body had a success rate in finding water (36.8% ofcases, n = 19) not different from individuaIs that had not engaged in this behavior (33.3%of cases, n = 9) (P > 0.9). SimilarIy, no significant difference was found between thetime needed to find water between snakes that dispIayed (51.3:!: 33.4 min; n = 7) or didnot dispIay (6.3:!:: 11 min; n = 3) the water collecting behavior (P = 0.25).

Dehydrated snakes showed a greater propensity to dispIay the water collecting behavior(80.8% of cases, n = 26) than well hydrated individuaIs (16.6% of cases, n = 24)(P < 0.0001). IndividuaIs Iarger than 600 mm SVL were never observed performing thecharacteristic coiling associated with water collection. However, following spraying theypromptIy ingested any water dropIets that adhered to their body surface.

Water colIection by the body in snakes 489

Figure 2. (Upper) A dorsal view of a juvenile male Bothrops moojeni perfarming the water collecting behavior.

Note the orientation of the head contrary to that of sca1es overlap. (Below) A lateral view af the same individual,

soon after water collection, loosening his body coils to drink the water that remained adhered to body surface.

490 Denis V. Andrade, Augusto S. Abe

Evaporative water loss

EWL rates were measured in three juvenile snakes (body mass = 10.6::1:3.7 g) and infour adults (body mass = 495.8::1: 242.1 g). Water loss for adults, as a percentage ofbody mass (0.049 ::I:0.013%), was significantly lower (P = 0.012) than that of juveniles(0.19::1:0.75%). On the other hand, no difference was found for EWL rate per unit area(P = 0.93) between adult (0.18::1:0.015 mgH20.cm 2.hr 1) and juvenile (0.18::1:0.055mgH20.cm 2.hr 1)snakes.

Discussion

The use of specialized morphology and/or behavior for water colIection and ingestioninc1udesexamples with lizards (Bentley and Blummer, 1962; Lasiewski and Bartholomew,1969; Louw and Holm, 1972; Gans et a1., 1982; Schwenk and Greene, 1987), snakes(Lou\v, 1972; Greene, 1986;.Ashton and Johnson, 1998), and tortoises (Auffenberg, 1963;Medica et a1., 1980). In snakes, alI these examples refers to the ability of drinking fromthe skin displayed by some viperids (Louw, 1972; Greene, 1986; Ashton and Johnson,1998). Nevertheless, to our knowledge, the stereotyped coiling used to colIect waterherein described for B. moojeni has never been reported before, and this may reflect theunlikelihood of recognizing a behavioral act not familiar to humans, and that is brief andsporadic rather than its absence in other snake species.

Bothrops moojeni occurs in Central/Southwestern Brazil, occupying habitats coveredwith semi-arid (or seasonally dry) tropical savannahs (CampbelI and Lamar, 1989).RainfalI in this region is not equalIy distributed over the year, and a welI defined dry seasonlasts about 4-5 months (Ferri, 1980; Radambrasil, 1981). During this season, rainfall occurssporadicalIy, for example in only 15 days out of 153 (Goodland and Ferri, 1979), and airhumidity often falIs below 50% (Radambrasil, 1981). The soil of much of B. moojeni'sgeographic range has a superficial sandy layer, welI drained and permeable, through whichrainwater rapidly percolates and where water pools rarely form (Goodland and Ferri, 1979;Ferri, 1980). Under these conditions, it seems plausible that the functional significance ofthe water colIecting behavior be related to the colIection of water from sporadic and briefrainfalIs. Moreover, at the micro-habitat levei, B. moojeni is regular1y found associatedwith humid zones, 10calIynamed "veredas" (I. Sazima, personal communication; see alsoStrussman and Sazima, 1993). Therefore, micro-habitat selection and water colIectingbehavior may be acting in tandem, permitting B. moojeni to thrive in an otherwise waterrestricted macro-environment.

Dehydrated snakes were more prone to display the water colIecting behavior than welIhydrated individuais, indicating that this behavioral act may help the maintenance of thehydration state. Following the water colIecting behavior there was an increase in tongue-flick frequency, but this increase did not result in an improved ability to find water.Therefore, it seems that vomerolfaction plays a limited role, if any, in the recognition

Water collection by the body in snakes 491

of water sources. This accords with observations of Meyer (1966) and Bradshaw andShoemaker (1967) who found that lizards appear to rely on water movement and/or thereflection/refraction of light by water as the main cues to find water sources.

Bothrops moojeni skin, as compared to other reptile species (see Mautz, 1982) presentsno significant improvement of the integumental barrier to decrease EWL. Ontogenetically,we found no difference in EWL rate, by unit area (mgH20.cm 2.hr 1),between juvenileand adult B. moojeni, despite the greater susceptibility of smaller snakes to desiccation,evident from their highertotal EWL (% ofbody mass). This observation indicates that: (1)juvenile snakes are more vulnerable to the risks of desiccation, because of their greatersurface-to-volume ratio, and (2) that this risk is not overcome by an improvement inskin isolation. Thus, the water collecting behavior and the strict association with humidzones observed in the juveniles of B. moojeni (see Sexton, 1956-57; Sazima, 1992) maycounterbalance their greater susceptibility to water 1088.Furthermore, water falling over thejuvenile's skin surface seems to coalesce more readily than on adult snakes. The juvenile

. skiri covered with small, keeled, and finely imbricated scales probably affords a higherdegree of hydrophoby than the skin of the adults, facilitating the formation of water beadsand enabling the juvenile snakes to drink more promptly.

Acknowledgements. We are indebted to Ariovaldo P. Cruz-Neto and Célio F.B. Haddad for their help with

the statistical procedures and for the loan of some articles. Ivan Sazima discussed many parts of the work

contributing with insightful comments, and providing unpublished information. We thanks Jaime Somera for

his expert drawing of the figures. An earlier version of this manuscript was largely improved by the comments of

Donald C. Jackson, Steven F. Perry, and Ivan Sazima.

References

Ashton, K.G., Johnson, J. (1998): Crotalus viridis concolor (Midget Faded Rattlesnake). Drinking from skin.Herpetol. Review 29: 170.

Auffenberg, W. (1963): A note on the drinking habits of some land tortoises. Anim. Behav. 11: 72-73.

Bentley, P.J., Blumer, W.F.C. (1962): Uptake ofwater by the lizard Moloch horridus. Nature 194: 699-700.

Bradshaw, S.D., Shoemaker, Y.H. (1967): Aspects of water and electrolyte changes in a field population ofAmphibolurus lizards. Comp. Biochem. Physiol. 20: 855-865.

Campbell, J.A., Lamar, W.W. (1989): The Venomous Reptiles of Latin America. New York, Comell UniversityPress.

Ooudsley- Thompson, J.L. (1971): The Temperature and Water Relations of Reptiles. Durham, Merrow Publ. Co.

Dejours, P. (1988): Respiration in Water and Air: Adaptations - Regulation - Evolution. Amsterdam, Elsevier.

Ferri, M.G. (1980): Vegetação Brasileira. São Paulo, Ed. Itatiaia/EDUSP.

Ford, N.B., Burghardt, G.M. (1993): Perceptual Mechanisms and the Behavioral Ecology of Snakes. In: Snakes:

Ecology and Behavior, p. 117-164. Seigel, R.A., Collins, J.T., Eds, New York, McGraw Hill.

Gans, C., Krakauer, T., Paganelli, C.Y. (1968): Water 1085 in snakes: Interspecific and intraspecific variability.

Comp. Biochem. Physiol. 27: 747-761.

Gans, C., Medin, R., Blumer, W.F.c. (1982): The water-collecting mechanism of Moloch horridus re-examined.

Amphibia-Reptilia 3: 57-64.

Goodland, R., Ferri, M.G. (1979): Ecologia do Cerrado. São Paulo, Ed. Itatiaia/EDUSP.

492 Denis V. Andrade, Augusto S. Abe

Oreene, H.W. (1986): Natural History and Evolutionary Biology. 10: Predator-Prey Relationships: Perspectives

and Approaches fiom the Study of Lower Vertebrates, p. 99-108. Feder, M.E., Lauder, O.V., Eds, Chicago,The University of Chicago Press.

Kirschner, L.B. (1991): Water and lons. 10: Environmental and Metabolic Animal Physiology, p. 13-107. Prosser,C.L., Ed., New York, Wiley-Liss, Inc.

Lasiewski, R.C., Bartholomew, O.A (1969): Condensation as a mechanism for water gain in nocturnal desertpoikilotherms. Copeia 1969: 405-407.

Lillywhite, H.B., Sanmartino, V. (1993): Permeability and water relations of hygroscopic skin of the file snake,Acrochordus granulatus. Copeia 1993: 99-103.

LittIe, C. (1990): The Terrestrial Invasion - An Ecophysiologica1 Approach to the Origins of Land Animais.Cambridge, Cambridge University Press.

Louw, O.N. (1972): The role of advective fog in the water economy of certain Namib Desert animais. 10:

Comparative Physiology of Desert Animais, p. 297-314. Maloiy, O.M.O., Ed., London, Zool. Soc. London.

Louw, O.N., Holm, E. (1972): Physiological, morphologica1, and behavioral adaptations of the ultrapsam-

mophilus, Namib Desert lizardAporosaura anchietae (Bocage). Madoqua, Ser. 111: 67-81.

Mautz, W.J. (1982): Patterns of Evaporative Water Loss. 10: Biology of the Reptilia, vol. 12: Physiology C,Physiological Ecology, p. 443-481. Oans, C., Pough, EH., Eds, New York, Academic Press.

Medica, P.A, Bury R.B., Luckenbach R. (1980): Drinking and the construction ofwater catchments by the desert

toftoi~e, Gopherus agassizi, in the MojaveDesert. Herpetologica 36: 301-304.

Meyer, D.E. (1966): Drinking habits in the earless lizard, Holbrookia maculata, and in two species of homedlizards (Phrynosoma). Copeia 1966: 126-128.

Minnich, J.E. (1982): The Use ofWater. 10: Biology ofthe Reptilia, Vol. 12: Physiology C, Physiologica1 Ecology,p. 325-395. Oans, c., Pough, EH., Eds, New York, Academic Press.

Pianka, E.R. (1986): Ecology and Natural History of Desert Lizards. Princeton, Princeton Univ. Press.

Pough, EH., Andrews, R.M., CadIe, J.E., Crump, M.L., Savitzky, AH., Wells, K.D. (1998): Herpetology. UpperSaddIe River, Prentice HaIl.

Radambrasil (1981): Folha SD. 22 Ooiás; Oeologia, Oeomorfologia, Pedologia, Vegetação e Uso Potencial daTerra. Rio de Janeiro, Ministério das Minas e Energia.

Sazima, I. (1992): Natural history of the jararaca pitviper, Bothrops jararaca, in Southeastern Brazil. 10: Biologyof the Pitvipers, p. 199-216. Campbell, J.A, Brodie, E.D., Jr., Eds, Texas, Selva.

Schmidt-Nielsen, K. (1979): Desert Animais: Physiologica1 Problems of Heat and Water. New York, Dover Publ.,Inc.

Schwenk, K., Greene, H.W. (1987): Water collection and drinking in Phrynocephalus helioscopus: a possiblecondensation mechanism. J. Herpetol. 21: 134-139.

Sexton, O.J. (1956-57): The distribution of Bothrops atrox in relation to food supply. BoI. Mus. Cienc. NaL 2:47-54.

Shoemaker, V.H., Hilmann, S.S., Hillyard, S.D., Jackson, D.C., Withers, P.c., Wygoda, M.L. (1992): Exchange

of water, ions, and respiratory gases in terrestrial amphibians. 10: Environmental Physiology of Amphibians,p. 125-150. Feder, M.E., Burggren, W.W., Eds, Chicago, The University of Chicago Press.

Soka1, R.R., Rohlf, EJ. (1986): Biometry. San Francisco. Freeman Co.

Strussman, c., Sazima, I. (1993): The snake assemblage of the Pantanal at Poconé, western Brazil: faunalcomposition and ecologica1 summary. Stud Neotrop. Fauna Environ. 28: 157-168.

Wang, T., Fernandes, W., Abe, AS. (1993): Blood homeostasis upon C02 anesthesia in the rattIesnake (Crotalusdurissus). The Snake 25: 21-26.

Zug, O.R. (1993): Herpetology: An Introductory Biology of Amphibians and Reptiles. New York, AcademicPress.

,.

Received: January 18,2000. Accepted: April27, 2000.