282 LVol.r 75

THE FUNCTION OF THE SALT GLAND IN THE BROWN PELICAN

BY KNUT SCHMIDT-NIELSEN AND RAGNAR I•'ANGE It has long been a matter of speculationhow oceanicbirds cover their needs for water. Some marine remain at sea for weeks or months,hundreds of miles from land and any possiblesources of fresh water. Sea water is known to be toxic to man and most other , the reasonbeing its high salt content. Someauthors have statedthat marine birds do drink sea water (Murphy, 1936: 337), while others have maintainedthat they can subsistwholly on water obtainedfrom the food (Smith, 1953:163). In order to profit from the ingestionof sea water it is necessaryfor an to excrete salts in a concentrationat least as high as that in the water ingested. The eliminationof the salt would otherwise require an additional amount of water which would be taken from the bodytissues. Therefore, if the organismcannot excrete a highlycon- centratedsalt solution,the drinking of sea water only will lead to a progressivedehydration or a harmful accumulationof salt. The kidneyis ableto excretesalts in a concentrationonly about one-halfof that found in . Hence, if the bird should excrete the salts from a given amount of sea water, it would be neces- sary to producetwice as much urine as the amount of water ingested. Thus, the kidneyis not ableto keepa marinebird in a favorablewater balance if it drinks sea water. However, it was recentlydiscovered that in marine birds a major part of the ingestedsalt is excretedextrarenally. The nasalglands, or salt glands,are able to producea highly concentratedsalt solution, makingit possibleto toleratedrinking of sea water (Schmidt-Nielsen ½tal., 1957, 1958). Cormorantsas well as a number of other marine birds have been found to secretesalt in high concentrationsfrom the salt gland, and there is little doubt that this is the generalavenue for salt excretion in all marine birds. This function is well correlated with the size of the salt gland, which is large in marine birds, as contrasted with terrestrialbirds, which have a very smallnasal gland. Our findings on the functionof the salt glandin the Brown Pelicanare here reported.

MATERIALS AND METYIODS The experimentalanimals were commonBrown Pelicans(Pelecanus occidentaliscarolinensis). They were caught on the Gulf Coast of Florida immediatelybefore being brought to the laboratoryfor the physiologicalstudies. Altogether eight birds were used, four with •J•l•] S½}•M•)x-N•zLSZ•^N•)F^NaE, Salt Gland inBrown Pelican 283 mature plumage (average weight 3.36 kg.), and four with juvenal plumage(average weight 2.59 kg.). Most of the birds were unwilling to eat in captivity,and force feedingresulted in vomiting. The experi- mental work was therefore done on fasting . The purposeof the investigationwas to establishwhether the salt gland plays the samerole in salt excretionin the pelicanas it doesin other marinebirds. Stimulationof the gland was accomplishedin two different ways, either by imposinga heavy salt load by infusion of hypertonicsodium chloride solution, or by injection of a solutionof mecholyl,a syntheticdrug with an effectsimilar to that of parasympa- thetic nerve stimulation. Although these stimuli do not correspond preciselyto the normal situationthe bird meetsin nature, the salt load is an exaggeratedrepresentation of the physiologicaleffects of ingestion of sea water, and the mecholylinjection simulatesthe action of the secretorynerves that normallycontrol the gland. All injectionswere made intravenouslyin the foot of unanesthetized birds. We have found in other speciesof marine birds that anesthesia blocksthe normal responseof the gland to a salt load. Samplesof the secretionfrom the salt gland were collectedby meansof thin poly- ethylenetubings introduced one or two mm into the narrow external nares. Occasionalsamples were taken of urine and of lacrymal fluid for comparativepurposes. The sodiumand potassiumcontent of the sampleswas determined by meansof a flamephotometer, and chlorideby a modificationof the Volhardtitration method (Rehberg, 1926).

RESPONSETO A SALT LOAD A salt load was imposedby injectionof variousamounts of a 10• sodironchloride solution in four pelicans (2 adult and 2 juveniles). Within 1 to 5 minutesafter the injection,drops of a clear, water-like liquid appearedat the externalnares. The liquid ran downalong the grooveson the upper side of the beak until it reachedthe tip of the beak,where it drippedoff. The headshaking which was characteristic in the behavior of cormorantsduring secretionfrom the salt gland (Schmidt-Nidsen,½taf., 1958) was not observed in pelicans.The secre- tion continued for some one to two hours. In a typicalsalt load experimentwith a pelican(2.54 kg. body weight) 28 ml of a 10• NaC1 solutionwere injectedintravenously. Within one minute of the beginningof injectionthe nasal secretion hadstarted. During an initlaJperiod of 7 minutesthe rate of secretion increasedrapidly. When it had reacheda high level it continuedat a rather constantrate for 110 minutesand then declinedabruptly and 284 SCHMIDT-NIELSENANDFANGE, Salt Gland in BrownPelican tVol[ Auk75 ceasedaltogether during the next 15 minutes. Except for the terminal observationperiod the flow remainedbetween 0.26 and 0.31 ml/minute (seetable 1). TABLE 1

VOLUME AND COMPOSITION OF NASAL SECRETION IN A PELICAN SUBJECTEl)TO SALT LOAI) Time, Vol. o] secr. ml per Conc.mEq/liter min. Min. ml. min. Na Cl K

O-7 7 -- -- 65O -- 11 7-17 10 2.885 0.29 637 656 10 17-27 10 2.686 0.27 700 695 12 27-37 10 2.997 0.30 764 850 15 37-57 20 6.155 0.31 736 740 16 57-77 20 5.879 0.29 721 730 14 77-97 20 5.209 0.26 704 708 14 97-117 20 5.119 0.26 696 704 13 117-132 15 2.815 0.19 676 694 14

Total: 33.745 Aver.: 0.27 698 722 13

The total amount of liquid secretedfrom the salt gland in this experimentwas 33.75 ml. In the same period four sampleswith a total volume of about 22 ml of cloacalcontents were passed(most of this was only slightlycontaminated with fecal material). The amount of liquid producedby the salt gland therefore exceededthe amount producedby the kidney. However, the important differencein the roles of these two organs appearswhen the salt concentrationsare compared. While the average salt concentrationin the nasal secretionwas over 700 mEq/liter (ab. 4 g NaC1per 100 ml), the averageurine concentra- tionswere about250 mEq/liter (ab. 1.5 g NaCI per 100 ml) (see table 2). With the smallervolume of urine and its much lower salt concen-

TABLE 2

VOLUME AND COMPOSITIONoF URINI• SAMPLES DISCIIAR(;E1) DURIN(; TIlE EXPERI]';IENTDESCRIBED IN TABLE 1

Urine vol. Conc.mEq/1

ml Na Cl K

6.7 222 251 18 7.1 207 205 7 5.9 261 316 38 1.9 156 269 52

21.6 212 260 29 July1958/-] SCHMIDT-NIELSENANDFANGE, Salt Gland in BrownPelican 285 trations, it appearsthat the amount of salt eliminatedby the kidney is about one fifth of the total, while about four-fifths of the salt was excretedfrom the salt gland in the observationperiod of slightlymore than two hours (see table 3). In this connectionit is desirableto emphasizethat the salt load imposedin this experiment was in the form of a salt solution about three times as concentrated as sea water. Such loads would not occur in nature. The total volumeof nasalsecretion (33.75 ml) exceededthe injected volume of salt solution (28 ml). Its concentrationof salt (ab. 700 mN), althoughnot as high as in the injectedsolution (1710 mN) was well abovethat of seawater (ab. 550 mN). In other words, the salt glandhas the capacityto excretethe saltscontained in ingested sea water. TABLE 3

THE AMOUNTS OF SALT EXCRETEDIN A PELICAN DURING TWO HOURS AFTER INJECTIONOF A SALT LOADOF 28 ML OF 10% NACL SOLUTION

N aC l Nasal Renal Total injected excretion excretion excretion Volume, ml 28 34 22 56 Total NaC1, mEq 47.9 23.9 5.3 29.2 Total NaC1, gram 2.8 1.4 0.3 1.7

RESPONSETO OTHER STIMULATION Previous work on cormorants has shown that the secretion from the salt gland is stimulatednot only by a load with sodiumchloride, but also by a generalosmotic load in the absenceof increasedsalt concen- trations(Schmidt-Nielsen et al., 1958). This was shownby the injection of hypertonicsolutions of sucrose,which stimulateda secretionsimilar to that observed after a salt load. Studieson gulls,which will be describedin anotherpublication, have shown that the innerration of the salt gland is of parasympathetic nature,and that secretionfrom the glandcan be stimulatedby mecholyl (methacholinechloride), a drug that mimicsparasympathetic stimula- tion and causessecretion from a numberof glands. It was thereforeof interestto comparethe stimulationby mecholylin the pelicanwith its effect on gulls. Mecholyl was injdcted intravenouslyin amountsof from 0.1 mg to 0.17 mg per kg body weight. Secretionfrom the salt glandsas well as secretionof tear fluid startedimmediately and con- tinuedfor periodsup to ten minutes.The rate of secretionfrom the salt glands was rather low, the highestrate observedunder mecholyl 286 SCHMIDT-NIELSENANDFANCE, Salt Gland in BrownPelican LVol.I' Auk stimulationwas 0.17 ml/minute for a five minuteperiod. I-Ienee,some samplesobtained were too small for completeanalysis, but sodium determinationscould be madeon five samples.These ranged from $90 to 750 m]•q Na/liter (aver. 652 m]•q/1). In other words, the con- centrationwas similar to that obtainedby a salt load (see table 1). Potassiumwas from 10-18 m]•q/1, and chloride from 632 to 708 mEq/1. The mecholylstimulation also providedan opportunityto obtain samplesof lacrymalfluid. In the three samplesobtained the sodium concentrationswere 52, 83 and 93 mEq/liter, respectively.This is lower than the sodiumconcentration of the plasma,which is about 150 to 160 mEq/liter, and it clearly showsthat the glandsthat produce the lacrymal fluid have no role similar to that of the salt gland in the excretion of salt.

Ti•]• MAJORGI•ANDS IN •I-t]• ORnI•AL In the Brown Pelicanthe salt gland,or nasalgland, is locatedin the upper anterior portion of the orbital cavity of the skull closeto the interorbitalsepturn (see figure 1). A shallowdepression in the under- side of the bony roof of the orbit (praefrontale) marks the place of thegland. The positionof thegland is the sameas reportedby T'echnau (1936, p. 560) in Pelecanusonocrotalus. The glandhas an oblongate pear-shapedform with a lengthof 2.6--3cm and a width of 0.6-0.8 cm. The surfaceof the glandis slightlylobulated. The attenuatinganterior end continuesforward horizontallyas a duct that is about 1 cm long

SALT GLAND

GLAND

'HARDERIANGLAND ..-,--"'

FxovR•;1. Orbital region of Brown Pelican, showing location of major glands. l•J•] SCltMmT-NIELSEN^NVF^N(;E, Salt Gland inBrown Pelican 287 and opensinto a narrow cavity a few mm behind the external nares. A glandtaken from a bird whichhas just beensecreting has a reddish color,probably due to the rich supply. Accordingto one single observationa non-secretinggland is paler and of a somewhatsmaller sizethan a secretinggland. The microscopicalstructure is very similar to that of the salt.gland of gulls (manuscriptin preparation).The secretoryelements consist of parallel closelypacked glandular tubules radiatingfrom centralducts. The glandtissue is surroundedby a thin connective tissue membrane. Several blood vessels and thin nerves passinto the gland but the anatomyof thesewas not studied. The weightsof the major glands of the orbit were determinedin six pelicans.The resultsof theseweighings are tabulatedin table 4.

TABLE 4 WEIGHTS IN MG OF ;THEM•AJ'OR C•LANDS IN ;THEORBITAL REGION OF THE BROWN PELICAN Salt gland HarderJangland Animal Bodywt. mg/kg mg/kg No. kg L R Total BW L R Total BW

A (ad.) 3.74 520 474 994 266 1776 1805 3581 958 C (ad.) 3.17 483 480 963 304 2065 2053 4118 1290 D (ad.) 3.41 -- 536 -- est. 314 2491 2487 4978 1462 E (juv.) 2.86 382 384 766 268 1894 1935 3829 1339 F (juv.) 2.36 370 353 723 306 2082 2052 4134 1753 G (juv.) 2.54 540 530 1070 421 2194 2126 4310 1697

The moststriking finding is that althoughthe salt gland(nasal gland) is quitelarge, the HarderJangland is somefour to five timesas large again. This glandforms a whiteor yellowishrounded mass in front of the eye bulb. Its histologicalstructure is vesicularand different from the salt gland. The lacrymalgland is situatedbehind the eye. It is a small gland which in three casesweighed from 8 to 12 mg. The normalfunction of the large Harderiangland is not knownwith certainty. However, it is likely that the "tears" that appearedunder mecholyl stimulation (see above) came from the Harderian gland. The amountof "lacrymal"fluid that was obtainedin about a minute ran as high as 342 mg, and it is unthinkablethat this amountof fluid was secretedby the lacrymalgland which weighsonly some 10 mg. These"tears" could, on the otherhand, easily be producedby a gland as largeas the Harderiangland. As describedearlier, the fluid is hypo- tonic, and it is visciousand slow flowing. It could well serve for pro- tectionof the eye of a diving bird, where the high viscositywould be a help in keepingit from beingwashed away. 288 SCHMIDT-NIELSEN^•) F^•CE, Salt Gland in BrownPelican Vol.Auk75

In contrast,the secretionfrom the salt (nasal) gland is always high in salt concentrationand water-like in consistency.With the rates of secretionobserved, which rangedup to 0.38 ml/minute (Pelican C) it can be estimatedthat the salt glands can secretea volume of liquid about four tenthsof their weight per minute. This is indeeda very high rate of secretion,particularly in view of the considerableosmotic work involved in the secretion.

DISCUSSION While sea water is known to be toxic to most mammals and terrestrial birds, it has been debated whether marine birds drink sea water. For an animalto toleratedrinking of seawater it is necessaryto excretethe saltsin a concentrationat leastas high as in the water taken in. While the bird kidneycannot produce a urine as concentratedas sea water, the excretoryfunction of the salt gland (nasal gland) is very efficient, producinga fluid with salt concentrationshigher than that in ingested sea water, therebyleaving a net gain of water. It is well known that the nasal gland is particularlywell-developed in marine birds. The significanceof this has been discussed,and the commonlyaccepted conclusion has beenthat the large nasalgland pro- ducesa secretionwhich will rinseaway the harmfulor irritating effect of seawater that penetratesinto the nasalcavity (Marpies, 1932). Of 83 speciesof birds examinedby Technau(1936) the 24 speciesthat had particularlylarge glands were all marine. It is interestingto note that Technau'scareful investigationsalso revealed a correlationwith the habitatof the bird within a singlegenus. For example,among gulls the size of the gland increasesas we move from the EuropeanBlack- headedGull (Larus ridibundus),through the Mew Gull (Larus canus) and the Herring Gull (Larus argentatus)to the Great Black-backed Gull (Larus marinus). In Europe the Black-headedGull is mostlya fresh water species;it breedsat fresh water and spendsmuch of the yearon the big Europeanrivers. The Great Black-backedGull, on the other hand, has the most pronouncedmarine habitat of all the species mentioned. The same trend to a correlation with the extremeness of the marine habitat is evident in other birds listed in Technau's well com- piledtables.. It has, furthermore, been shown that the size of the salt gland to someextent dependson the adaptationof the bird to its habitat. For example,Schildmacher (1932), working in the Berlin Zoo, found that when ducksof the samespecies were broughtup in salt water and in fresh water, thosebrought up in salt water had considerablylarger nasalglands. •] $CnMt•r-NtE•.SENANOFANCE,Salt Gland inBrown Pelican 289

While these correlationsbetween the size of the salt gland and the marineenvironment are well established,it is necessaryto modify the interpretationthat the functionof the gland is to protectthe nasal mem- branesagainst sea water. Instead, it is evident that the gland has the unique role of being the main for excretion of salt in marine birds. ACKNOWLEDGMENTS This study was supportedby Grant No. H-2228 from the National Institute of Health.

SUMMARY The salt gland (nasal gland) of the Brown Pelicancan excretea highlyconcentrated solution of sodiumchloride. The excretorycapacity of the salt gland permits the bird to tolerate ingestionof sea water, and to profit from it becausethe salt is excreted in a concentration higher than in sea water. Quantitatively,the role of the salt gland in the eliminationof sodiumchloride is greaterthan that of the kidney.

LXTF•ATu• C•T• MAmm•:s,B.J. 1932. The structureand developmentof the nasal glands of birds. Proc. Zool. Soc. London,829-844. MuRrn¾, R. C. 1936. Oceanicbirds of South America, vol. 1. 640 pp. Amer. Mus. Nat. Hist. New York. PoRTX•m,P. 1910. Pression osmotiquedes liquides des oiseaux et mammiferes marins. J. Physiol. Pathol. Gen., 12: 202-208. R•:HsF•c, P. B. 1926. The determinationof chlorine in blood and tissuesby micro- titration. Blochem.J., 20: 483-485. ScHx•,•)Macnr•,H. 1932. Ueber den Einfiussdes Salzwassersanf die Entwicklung der Nasendrfisen.J. Ornithol., 80: 293-299. SCHMIIYr-NIELSEN,K., C. B. JSRGENSENand H. OSAKI. 1957. Secretionof hyper~ tonic solutionsin marinebirds. Fed. Proc., 16: 113-114. SCaMm'r-NmI,S•;N, K., and W. J. L. SI,a•)•;N. 1958. Nasal Salt Secretionin the Humboldt Penguin. Nature, 181: 1217-1218. SCHM•rrr-N•;I,S•;N,K., C. B. J6sC•;NS•;Nand H. Os^•L 1958. Extrarenal salt excretion in birds. Am. J. Physiol., 193: 101-107. S•x•/, H. W. 1953. From Fish to Philosopher.Little, Brown and Co., Boston. 264 pp. Tr,CHN^u, G. 1936. Die Nasendrfiseder V6gel. J. Ornithol.,84: 511-617. Departmentof Zoology,Duke University,Durham, North Carolina, and Archbold BiologicalStation, Lake Placid, Florida.