o X o ^ vV^) 2 Z t ^ /■ t WUi^y ?.r S " 'j^ ^ \.iVA5vii/ 2 5* ’ ^ Xijrosji^/ > 3NIAN^lNSTITUTION‘^N01iniIlSNI_NVINOSHilWs‘^S31 HVdan_LlBRARl ES‘^SM1THS0N1AN_INSTITI — c/> —»■ V ^ (/) (/> O ~ O X2;vos5i^ ~ O ^^ilWS""s3lavaan"'UBRARIES SMITHSONIAN^'iNSTITUTlON NOIiniIXSNj“'NVINOSHXIlNS S3IBV ~ z r- .. z onian'institution NoiiniiiSNi wviNOSHiu^s SHiavyan libraries Smithsonian instit 4 X . O ffig' X Hiiws"''s3 lavaan^LIBRARl ES^^SMITHSONIAN^ institution N0lifUIlSNI_NVIN0SHilWs‘^S3 1 U) — tn ^ 7*^ . (/y N R Iff ;0NiAN""lNSTITUT10N"'N0lini!iSNi“'NVIN0SHillMS"S3 I aVB a H LI BRAR I ES~SMITHSONIAN_!NSTn >H1I1AIS^S3 I ava a n”LI B R AR l ES^SMITHSONIAN~INSTITUTION^NOIiniIiSNI NVIN0SHiHAiS^^S3 I a’ < ^ ~ 8 I I CCJI i I I >' s 50NIAN"'lNSTITUTl0N'^N0liniliSNl^NVIN0SHims‘"s3 I a Va a IT^LI B RAR I Es'^SMITHSONIAN JNSTIl — C/) ^ \ (/> — u> E w SMITHSONIAN INSTITUTION NOlinillSNI NVlNOSHillAlS S3iavaan LIBRARIES z < 5 {/) in 2 NVlNOSHillAlS S3 1 a va an libraries smithsonian institution ^ I ‘^NOIiniIlSNI_ ,l B RAR I ES^SMITHS0NIAN“"lNSTlTUTI0N"^N01iniIlSNl“'fNVIN0SHlllA!S^S3 I ava 813 libraries u/ in CO NOlinillSNI NVlNOSHillAlS S3iavaail libraries SMITHSONIAN X CO o i xo^D^ I CO Z libraries SMITHSONIAN INSTITUTION to INSTITUTION NOlinillSNI V- ^ ^ 2 to Z £0 NOlinillSNI NVlNOSHillAlS to X CO o z > r„. r. ¥ I?.!'- Wttjii: I: ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 56 • 1971 • NUMBERS 1-7 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont T reasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Edward R. Ricciuti, Editor Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor EDITORIAL COMMITTEE Robert G. Goelet, Chairman; William G. Conway, Donald R. Griffin, Hugh B. House, F. Wayne King, Peter R. Marler, Ross F. Nigrelli, James A. Oliver, Edward R. Ricciuti, George D. Ruggieri, S.J. William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology; Hugh B. House, Curator, Mammalogy; James G. Doherty, Assistant Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Robert A. Brown, Assistant Curator, Animal Departments; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director; Stephen H. Spotte, Curator; H. Douglas Kemper, Assistant Curator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director and Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung, Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology; Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate; George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists; Robert M. Beck, Research Fellow ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants; John Budinger, Pathology; Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology; Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology c& Toxicology: Harold S. Goldman, Radiology; Roy Bellhorn, Paul Henkind, Alfred Freidman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine; Theodore Kazimiroff, Dentistry © 1972 New York Zoological Society. All rights reserved. Contents Issue 1. May 5, 1971 PAGE 1. The Effect of Holothurin on Leucocyte Migration. By B. J. Easley and Ross F. Nigrelli 1 News and Notes New York Medical College to Sponsor Cooperative Program in Compara- tive Pathology 13 Issue 2. August 20, 1971 2. Species Identification of Commercial Crocodilian Skins. By F. Wayne King and Peter Brazaitis. Figures 1-41 15 News and Notes Crocodylus intermedins Graves. A Review of the Recent Literature. By Peter Brazaitis. Figures 1-3 71 Issue 3. January 18, 1972 3. Inheritance of Melanophore Patterns and Sex Determination in the Monte- zuma Swordtail, Xiphophorus montezumae cortezi Rosen. By Klaus D. Kallman. Figures 1-10 77 4. Eastern Pacific Expeditions of the New York Zoological Society. Stoma- topod Crustacea. By Raymond B. Manning. Figures 1-3 95 5. On the Resemblance of the Young of the Fishes Platax pinnatus and Plec- torhynchus chaetodonoides to Flatworms and Nudibranchs. By John E. Randall and Alan R. Emery. Figures 1-4 115 Issue 4. August 7, 1972 PAGE 6. Chromosome Numbers of Heliconiine Butterflies from Trinidad, West Indies (Lepidoptera, Nymphalidae). By Esko Suomalainen, Laurence M. Cook, and John R. G. Turner. Figures 1-4 121 7. The Genetics of Some Polymorphic Forms of the Butterflies Heliconius melopomene (Linnaeus) and H. erato (Linnaeus). II. The Hybridization of Subspecies of H. melpomene from Surinam and Trinidad. By John R. G. Turner. Plate-figures 1-37; Text-figures 1-4 125 News and Notes Underwater Sounds of Southern Right Whales. By Roger Payne and Katharine Payne. Plate I; Text-figures 1-2 159 INDEX TO VOLUME 56 166 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 56 • ISSUE 1 • SPRING, 1971 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President John Pierrepont T reasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Edward R. Ricciuti, Editor & Curator, Publications & Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor EDITORIAL COMMITTEE Robert G. Goelet, Chairman; William G. Conway, Donald R. Griffin, Hugh B. House, F. Wayne King, Peter R. Marler, Ross F. Nigrelli, James A. Oliver, Edward R. Ricciuti, George D. Ruggieri, S.J. William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology; Hugh B. House, Curator, Mammalogy; James G. Doherty, Assistant Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Herpetologist; Robert A. Brown, Assistant Curator, Animal Departments; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus AQUARIUM James A. Oliver, Director; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology; Jay Hyman, Consultant Veterinarian OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director & Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempier, Jr., Assistant to the Director & Bio-Organic Chemist; Eli D. Goldsmith, Scientific Consultant; William Antopol, Research Associate, Comparative Pathology; C. M. Breder, Jr., Research Associate, Ichthyology, Jack T. Cecil, Virologist; Harry A. Charipper, Research Associate, Histology; Paul J. Cheung, Microbiologist; Erwin J. Ernst, Research Associate, Estaurine & Coastal Ecology; Kenneth Gold, Marine Ecologist; Jay Hyman, Research Associate, Comparative Pathology; Myron Jacobs, Neuroanatomist; Klaus Kallman, Fish Geneticist; John J. A. McLaughlin, Research Associate, Planktonology; Martin F. Schreibman, Research Associate, Fish Endocrinology INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR (Operated jointly by the Society and the Rockefeller University) Peter R. Marler, Director & Senior Research Zoologist; Paul Mundinger, Assistant Director & Research Associate; Donald R. Griffin, Senior Research Zoologist; Jocelyn Crane, Senior Research Zoologist; Roger S. Payne, Research Zoologist; Fernando Nottebohm, Research Zoologist; George Schaller, Research Zoologist; Thomas T. Struhsaker, Research Zoologist; Alan Lill, Research Associate; O. Marcus Buchanan, Resident Director, William Beebe Tropical Research Station ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Ben Sheffy, Consultant, Nutrition; Roy Bellhorn, Consultant & Research Associate, Comparative Opthomalogy; John Budinger, Joseph Conetta, Edward Garner, Henry Kobrin, Ralph Strebel, Consultants & Research Associates, Comparative Pathology. 1 The Effect of Holothurin on Leucocyte Migration B. J. Lasley* and Ross F. Nigrelle Crude Holothurin, purified Holothurin A, and desulfated Holothurin A, water soluble steroid saponins from the Cuvierian organs of the Bahamian sea cucumber Actino- pyga agassizi Selenka, cause a stimulation of leucocyte migration from buffy layer in capillary tubes in relatively low concentrations and an inhibition of the move- ments at higher concentrations. The inhibitory properties are compared with similar effects by Ouabain, a non-hemolytic saponin, and by liberated hemoglobin. The Holothurins had no effect on the migration pattern of the leucocytes in the presence of anaerobic and aerobic inhibitors. The effect of Holothurin on the amoeboid movements of the leucocytes is dis- cussed in relation to surface alterations and cell membrane permeability. Introduction The main purpose of this study was to obtain additional information on the bio- logical properties of Holothurin and its fractions. Many toxic materials obtained from plant and animal tissues cause similar host re- sponses when released from combination with cellular components. The effects of crude Holo- thurin, Holothurin A, and desulfated Holothurin were studied in systems of mammalian leuco- cytes present in whole blood or in saline suspen- sions of white blood cells using the migration technique of Ketchel and Favour ( 1955). Com- parative experiments were made with the sapo- nin Ouabain, while hemolytic properties were investigated in relation to the inhibition of leuco- cyte activity. By the use of aerobic and anae- robic inhibitors, an estimate was made of bio- chemical pathways associated with a stimulation of leucocyte migration. Holothurin is a water soluble, steroid saponin with surface-active properties found in the Cuvierian organ of the Bahamian and West Indian sea cucumber Actinopyga agassizi Se- lenka. The active principle may be obtained from granules in branching filaments of Cuvierian tubules. Crude Holothurin represents dried tubules and analysis shows the presence of gly- cosides, pigment, cholesterol, insoluble proteins, salts, free amino acids, and polypeptides (Ni- grelli and Jakowska, 1960). Holothurin, ob- tained from the water extract of these granules, consists of steroid aglycones, bound individually I Psychobiology Research Unit of the Payne Whitney Clinic, New York Hospital, Cornell Medical Center, New York, New York 10021. 2 Osborn Laboratories of Marine Sciences, New York Aquarium, Brooklyn, New York 11224. to four molecules of monosaccharides (Chanley, et ai, 1960). Holothurin A represents 40% of the crude Holothurin and is a cholesterol-pre- cipitated fraction (Nigrelli and Jakowska, 1960) with the empirical formula C50-52H81-85O25-20^Na (Chanley, et ai, 1959). Infrared spectrum analysis indicates a five-or- six-membered ring lactone and one double bond. On acid hydrolysis, Holothurin A yields water- soluble aglycones, sulfuric acid and water soluble reducing sugars (Chanley, et ai, 1960). Recent work has shown the presence of a half-esterified sulfate residue (— OSOa^Na+l, probably at- tached at some point in the sugar chain. De- sulfation of Holothurin A occurs when treated with 0.2M methanolic HCl at 37 C. This results in a neutral product devoid of the sulfate group, while retaining unaltered glycoside-genin bonds (Friess, et ai, 1967). Investigations into properties of Holothurin have shown it to be a powerful neurotoxic agent (Freiss, et al., 1959) and to possess antitumor activity capable of suppressing growth of Sarcoma-180 (Nigrelli, 1952; Nigrelli and Zahl, 1952). Similar effects were noted on Krebs-2 ascites tumors in Swiss mice (Sullivan, et al., 1955). Hemopoietic effects have been demon- strated in Rana pipiens (Jakowska, et ah, 1958), while hemolytic properties were noted using rabbit red blood cells (Nigrelli and Jakowska, 1960) and human erythrocytes (Thron, 1964). In the current study, the effect of Holothurin and its fractions in a system of leucocytes dis- playing amoeboid movement is described. The effect of these compounds on this physiological property was interpreted in relation to surface alterations and cell membrane permeability. 1 2 New York Zoological Society: Zoologica, Spring, 1971 Materials and Methods A. Holothiirin solutions Stock solutions of crude Holothurin, purified Holothurin A, and desulfated Holothurin A were prepared in physiological saline solution. Final concentrations are listed in Tables 1 through 1 1. (For Tables, see page 8 ff.) B. Leucocyte migration A modification of the technique developed by Ketchel and Favour (1955) was used through- out this study. Leucocyte ability to migrate by amoeboid motion was measured quantitatively as white blood cells moved away from a leuco- cyte buffy layer. The technique utilized a micro-hematocrit formed inside capillary tubes each of which was approximately 7 cm long with an internal diam- eter of 0.8 ± 0.1 mm. New tubes were used for each experiment. By means of capillarity, each tube was filled with heparinized blood (0.1 mg heparin/ml blood) in the presence or absence of each test compound, to two-thirds its length. One end of each tube then was heat sealed. Temperature effects associated with this method of sealing did not alter migration patterns. The tube was centrifuged at 3000 rpm for 2 minutes, produc- ing a three component system: a lower layer of packed erythrocytes, an intermediate buffy layer of leucocytes, and an upper layer of plasma (semi-clot). Upon vertical incubation of each tube at 37 °C for 5 hours, leucocytes migrated into the plasma layer. Extent of migration from the buffy coat was then measured by means of an ocular micrometer. Trypan Blue (Tennant, 1964) staining (1% solution) was always performed on leucocytes from the buffy layer after each five hour incu- bation to estimate the percentage of viable cell populations and data analyzed statistically using Student’s test (Snedecor, 1956). 1 ) Whole blood containing Holothurin Hospitalized psychiatric patients, receiving no medication, served as blood donors. Two ml of heparin (2 mg/ml) in isotonic saline was added to 38 ml venous blood. This final concentration of heparin (0.1 mg/ml blood) delayed clotting. Various amounts of crude Holothurin, Holo- thurin A, or desulfated Holothurin A in saline or saline alone, were added to the heparinized blood, incubated for two hours at 25°C, and then used to fill each capillary tube. 2) Leucocyte suspensions containing Holothurin Heparinized blood (0.1 mg heparin/ml blood) placed in long stoppered glass tubes, 6x250 mm, were incubated at a 45-degree angle for 45 mins. at 37 °C. The plasma-leucocyte layer was then withdrawn, pooled, and gently mixed with a pipette. Varying amounts of crude Holothurin in saline were added to the plasma suspension to attain higher concentrations than possible in whole blood, without hemolysis from erythro- cytes. Incubation of the suspension porceeded for 2 hrs. at 37 °C in rubber stoppered tubes (15 X 100 mm), after which the suspension was centrifuged at 1500 rpm for 5 mins, and washed once with 5-10 ml heparinized saline. After additional centrifugation at 1500 rpm for 5 mins, the saline solution was discarded and leu- cocytes resuspended in homologous plasma, using amounts of Holothurin-free heparinized plasma comparable to the original ratio in whole blood. In order to use the Ketchel and Favour (1955) migration procedure, it was found nec- essary to have the leucocyte buffy layer overlay packed erythrocytes in each capillary tube. To obtain a leucocyte-plasma-erythrocyte ratio com- parable to that which initially existed in the whole blood, heparinized blood was centrifuged in hematocrit tubes at 3000 rpm for 15 mins. From divisions on the tubes and a known total volume, it was possible to estimate the amount of each component initially present in a blood sample. On this basis, reconstituted leucocyte-plasma- erythrocyte suspensions were placed in capillary tubes and taken through the procedure as de- scribed for whole blood. Experience has shown that leucocytes will not migrate when a buffy layer is located at the basal end of each capillary tube. Toxic factors are probably released from the glass tip during the heat sealing process. 3) The effect of hemolysis a) To obtain plasma containing a high con- centration of hemoglobin, several ml of heparin- ized blood (0.1 mg heparin/ml blood) were frozen and thawed three times using dry ice in absolute alcohol. Centrifugation at 3000 rpm for 10 minutes allowed removal of cell debris. The resulting solution consisted mainly of hemo- globin in plasma. b) Hemolyzed blood prepared with crude Holothurin in a concentration of 20 fxg/mX blood, was allowed to remain at 25 °C for 1.5 and four hours. The hemolytic action of Holo- thurin caused a liberation of hemoglobin. After centrifugation at 3000 rpm for 10 minutes, a cell-free solution of hemoglobin remained. To equate the intensity of liberated hemo- globin in these two plasma samples, optical density readings were made using a Klett- Summerson colorimeter (#54 filter). Heparin- ized blood in contact with Holothurin for 1.5 Lasley and Nigrelli: Effect of Holothiirin on Leucocyte Migration 3 hours at 25°C and hemoglobin added to normal plasma to give a corresponding O.D., consti- tuted solutions and readings (A). (See Table 6.) The same procedure was used for (B) solutions, except that Holothurin was allowed to remain in heparinized blood four hours at 25 °C. Cor- respondingly more hemoglobin was then re- quired in plasma to give higher O.D. readings in (B). After establishing the exact amount of hemo- globin (liberated by freezing and thawing or Holothurin treatment) necessary to give com- parable concentrations in whole blood, capillary tubes were filled and incubated at 37°C for five hours. Leucocyte migration was then measured by means of an ocular micrometer. 4) Crude Holothurin versus Ouabain Experiments were designed to show the effect of a non-hemolytic saponin (Ouabain) in rela- tion to a hemolytic saponin (Holothurin) on leucocyte migration. To obtain the same molar concentration, cal- culations based on molecular weights of Ouabain and Holothurin were made to estimate the amount of each compound necessary to give a final concentration of 1.4 x 10~°M in heparin- ized blood. Stock solutions of each compound were prepared in saline and appropriate dilu- tions added to heparinized blood. Capillary tubes were then filled, sealed, centrifuged, and incu- bated for five hours at 37°C. 5) The influence of enzymatic inhibitors Freshly prepared saline solutions of the fol- lowing inhibitors were added to heparinized blood and allowed to remain at 25°C for two hours: Sodium fiuoride, iodoacetic acid (sodium salt), 2,4-dinitrophenol, and potassium cyanide. Concentrations are given in Tables 8, 9, 10, 11. Crude Holothurin, at a final concentration of 1 /xg/ml blood was added in some cases. All quantities of inhibitors and Holothurin in blood were of volumes less than one part inhibitor or Holothurin to four parts of heparinized blood. Saline controls always contained the same vol- ume ratio. After filling, sealing, and centrifuging (3000 rpm for two minutes) each capillary tube was incubated in a vertical position for five hours at 37°C. Results In all cases with crude Holothurin, Holo- thurin A or desulfated Holothurin A, low con- centrations showed a significant increase in the rate of leucocyte amoeboid motion. The stimu- lation by crude Holothurin (Table 1) occurred mainly in the range of 0.1 to 4 /zg/ml blood. Inhibition by crude Holothurin (Table 2) re- sulted in a decrease in leucocyte migration. principally at levels between 16 and 22 /jig crude Holothurin/ml blood. Trypan Blue staining showed the presence of viable cells in all experi- ments. Effects were not due to massive altera- tions in the test system; however, considerable hemolysis was noted in tubes treated with higher concentrations of Holothurin. No sig- nificant results were observed in concentrations ranging from 1 to 14 /xg/ml. Holothurin A was found to be effective in lesser amounts, producing a stimulation at a con- centration of 0.02 /xg/ml and a decrease in the rate of amoeboid motion at a concentration of 0.1/xg/ml. Movement of leucocytes was com- pletely inhibited at a concentration of 20 /xg/ml, with the subsequent liberation of inhibitory concentrations of hemoglobin from erythrocytes. When desulfated Holothurin A is compared to crude Holothurin and Holothurin A it will be noted that the biological activity of this com- pound has been greatly changed. Thus, a stimu- lation in migration will occur at 1/xg/ml, while concentrations greater than 100 /xg/ml were necessary to inhibit leucocyte motion. Observations indicate that the hemolytic properties of crude Holothurin in higher con- centrations may have some effect on leucocytes in whole blood (Table 5). A crude Holothurin concentration of 40 /xg/ml leucocyte suspension (23 /xg/ml blood) does not significantly alter the rate of white blood cell migration, however concentrations from 100-1000 /xg/ml leucocyte suspension (greater than 58 /xg/ml blood) will significantly decrease the rate of motion. A significant alteration in cell viability occurred in the presence of higher concentrations of Holo- thurin when levels greater than 100 /xg/ml leucocyte suspension were reached. To substantiate the fact that crude Holothurin per se had affected leucocyte mobility, experi- ments were designed to compare results obtained when the non-hemolytic saponin Ouabain was used in the same concentration as the hemolytic saponin (crude Holothurin). Results showed no significant differences between the two saponins at a concentration of 1 .4 x 10~®M (16 /xg/ml blood). (Table 8.) The data indicate approxi- mately the same decrease in migration when compared with the control. Trypan Blue stain- ing demonstrated the viable nature of these cells within an acceptable range. No gross mortality had caused alterations in the rate of leucocyte migration. Holothurin is known to possess strong hemo- lytic properties with concentrations as linear functions of the red blood cell count and reac- tion time (Thron, 1964). The effect of homolo- gous hemoglobin, in the absence of crude Holo- 4 New York Zoological Society: Zoologica, Spring, 1971 thurin, was found to affect leucocyte migration only in concentrations (O.D.) comparable to that produced by crude Holothurin (O.D. 250) in contact with red blood cells for four hours at 25 °C. The hemoglobin concentration (O.D. 103) comparable to crude Holothurin hemoly- sis after contact for 1.5 hours did not affect leucocyte migration. It appears that the amount of hemoglobin liberated by crude Holothurin after four hours wilt significantly decrease the rate of leucocyte motion (Table 6). The effect of selected inhibitors on white cell migration was studied with the hope of obtain- ing information on metabolic pathways in the presence and absence of crude Holothurin. Both anaerobic inhibitors (sodium fluoride and iodo- acetate) depressed migration in concentrations less than 1 x lO^^M, white the aerobic inhibi- tors (2,4-dinitrophenol and potassium cyanide) required greater amounts (1 x lO^^M or more) to produce a decrease. In the case of dinitro- phenol, solubility problems prevented an accu- rate estimation of the amount required to pro- duce an inhibition. A study of inhibitors, in the presence of crude Holothurin, was undertaken in an attempt to explain the stimulation of migration resulting from use of low concentrations (8.8 x 10~'^M or 1 jxg/m\ blood) of this compound. In all cases with both aerobic and anaerobic inhibitors, a decrease in migration occurred in the presence of crude Holothurin. When compared to the rate of leucocyte mobility in the absence of in- hibitors, it is possible to state that crude Holo- thurin stimulation is the result of both aerobic and anaerobic metabolism. Leucocyte non- viability was not a factor since Trypan Blue staining indicated a cell survival greater than 95%. All concentrations used allowed each ex- periment to be carried out under optimal conditions. Discussion The biological and chemical activities of Holothurin put this compound in the class of steroid saponins (Nigrelli and Jakowska, 1960) with surface-active properties (Seeman, 1967; Thron, 1964) and ability to alter cell membrane structure. High surface activity of echinoderm toxins, structural arrangement, and chemical properties contribute to their ability to pene- trate membranes and demonstrate an affinity for cellular components (Ruggieri, 1965). These properties are significant in the study of Holo- thurin action on leucocytes in relation to sys- tems incorporating surface phenomena, such as amoeboid movement. By definition (White, et al., 1964), all sapo- nins lower surface tensions and contribute to cytolytic effects when used in sufficient quantity. The work of Ponder and McLeod (1936) has shown that saponin hemolytic substances will affect white cells in a manner similar to their effect on red cells. Current experimental data using suspensions of leucocytes confirms these earlier reports. Crude Holothurin in concentra- tions greater than 100 //.g/ml will cause a cer- tain percentage of white cell disintegration, de- creased cell counts, and altered cell viability. Since Holothurin is chemically a member of the saponin family, it is expected that this com- pound will show activity towards cell mem- branes. Studies by Friess, et al., (1960) postu- late that the Holothurin effect may be a non- specific surface action on cell membranes or that it could be a specific attack at one or more sites in the membrane. The demonstration that Holo- thurin can produce cytotoxic effects on leuco- cytes, as shown by Trypan Blue uptake and altered morphology in higher concentrations, would suggest that Holothurin in vitro, and pos- sibility in vivo, can affect integrity and permea- bility of this particular cell type. The migration system had advantages over others since it util- ized a huffy layer of leucocytes from which isolated cells could move away from the main population. It has been suggested (Parpart and Ballentine, 1952) that cell membrane structure is a mosaic of cylinders containing phospholipid and cho- lesterol surrounded by a protein meshwork. Observations by Dourmashkin, et al., (1962) have shown that, in the presence of suitable agents, extensive re-arrangement of these mem- brane lipids may occur. The original membrane is rendered more permeable by incorporation of saponin, thereby forming a more permeable structure with spaces 90 Angstroms across that appear to contain water. Saponin hemolytic activity is thought to depend on these spaces in the lipid component of cell membranes. Altera- tions in lipid structures of this kind are impor- tant in controlling the permeability of cells under physiological conditions. It is therefore suggested, that changes in leucocyte permeabil- ity in the presence of Holothurin could be re- sponsible for several effects observed during the current study with this compound. Holothurin is known to be a surface active agent and it is suspected that a change in celt permeability and eventual destruction of Holothurin-treated leucocytes (in concentrations greater than 100 fxg/m\) are initiated by reactions on the cell surface. In addition, Holothurin may alter cell metabolism by virtue of its action at cell sur- faces to allow freer passage of this compound to an intracellular environment. Lasley and Nigrelli: Effect of Holothiirin on Leucocyte Migration 5 Treatment of human leucocytes with crude Holothurin, Holothurin A, or desulfated Holo- thurin A resulted in a stimulation of migration as well as inhibition of phagocytic motility. Concentrations were chosen which could affect these physiological properties but not cause cell death in most cases. Results may be explained on the basis of structure, surface properties, or by consideration of the role of chemical inter- action between a surface-active agent (Holo- thurin) and the cells involved. Chemical ap- proaches to amoeboid movement have been lacking; however, it is suspected that Holothurin may enhance migration through a stimulation of leucocyte glycolysis. It is well established that certain substances, as bacterial endotoxins, hor- mones, and polysaccharides (Woods et al., 1961 ), produce a glycolytic stimulation and that this stimulation can be associated with an in- crease in functional capacities of the cells. On the basis of chemical structure, it will be noted that the concentration of desulfated Holo- thurin A required to produce a decrease in mi- gration was considerably higher (greater than 100 /i,g/ml) than that necessary to inhibit leu- cocyte mobility in the presence of crude Holo- thurin or Holothurin A. Certain structural con- figurations of Holothurin modify leucocyte motility in systems involving physiological re- sponse. The action of a saponin devoid of the acidic sulfuric acid group may be interpreted in view of its ability to combine with basic groups of membrane proteins. It has been noted by Friess, et al., (1967) that the anionic nature of Holothurin facilitates permeability through membranes, allowing greater speed of action and therefore a more complete reaction in a given time. Desulfated Holothurin A demon- strates an impeded degree of membrane per- meation due to a separation of lipid and poly- saccharide tissue phases. Toxic manifestations accordingly would be inhibited during the initial phases of interaction within cells. Friess, et al., ( 1967), also note that a resulfation of desulfated Holothurin may be necessary for this compound to become biologically active. An alternative approach to explain the stimu- lation of leucocyte migration resides in an area related to the lytic properties of Holothurin. Holothurin has a strong hemolytic potency since it has a high affinity for erythrocytes. In suspen- sions of heparinized blood, where both erythro- cytes and leucocytes were present, low concen- trations of Holothurin were probably taken up by red cells, leaving little or no free lysin to act on the white cells. More Holothurin caused greater absorption by red cells at a constant number; however more free lysin also remained in solution to act on the leucocytes. Under these conditions a stimulation of leucocyte migration is merely the rate occurring where extremely small amounts of Holothurin have remained free in the system (less than 0.02-4 yixg/ml). Eventually a concentration of free Holothurin is reached that will be too great to cause a stimulation but not sufficient to produce an in- hibition. This wilt occur in a suspension of blood initially receiving 6-14 fig of crude Holothurin/ ml blood, and is referred to as the intermediate range. In crude Holothurin concentrations ini- tially greater than 16 fig/ml blood, much free saponin will be available to inhibit leucocyte activity. As the concentration of this compound is further increased, results (Table 5) indicate much cell destruction concomitant with a sig- nificant decrease in leucocyte mobility and the presence of numerous non-viable cells. On mi- croscopic examination, distortion of cell mor- phology, altered staining properties and frag- mentation were evident. Macroscopically at a concentration of 580 fig crude Holothurin/ml blood, viscous turbid solutions of leucocyte debris resulted after approximately one hour in the presence of this compound. It is obvious that cell integrity was being altered, membranes were being attacked, and leucocyte structure was being destroyed. The effect of liberated hemoglobin also was studied in the migration system. Recent studies by Rideal & Taylor (1958) support the view that hemolysis, caused by saponins, involves ad- sorption of hemolytic agents on the cell wall. This adsorption alters bound cholesterol, with eventual cell wall destruction and release of hemoglobin. Rate of liberation is dependent on the concentration of saponin present in the sys- tem, and a period of time is required for all hemoglobin to be released from the erythrocytes. Table 6 shows the effect of hemoglobin on leuco- cyte migration. Low concentrations of hemo- globin do not alter migration patterns whereas concentrations attained after four hours, using a crude Holothurin concentration of 20 |U,g/ml, seem to display an inhibitory effect. Additional studies with leucocyte suspensions (Table 5) indicate that free hemoglobin has been a factor to depress migration in crude Holothurin con- centrations of approximately 20 fig/m\ blood (Table 2). To further substantiate this conclu- sion, the non-hemolytic saponin Ouabain was used in a system of heparinized blood (Table 7) at a lesser concentration (1.4x 10~®M). Migra- tion results at this level of saponin are due to an effect of the compound and not to the pres- ence of hemoglobin, assuming its action on cell membranes and biochemical processes is similar to that of crude Holothurin. No significant dif- ferences were noted between crude Holothurin 6 New York Zoological Society: Zoologica, Spring, 1971 and Ouabain when used in the same concentra- tion (Table 7). Both saponins were able to de- press significantly leucocyte migration when compared to saline controls. Several inhibitors were studied in the migra- tion system to further elucidate factors involved in leucocyte metabolism in the presence of Holothurin. A decrease in migration may be explained as the result of inhibition of either glycolytic or oxidative respiratory metabolism of leucocytes. In present studies two anaerobic inhibitors, iodoacetate (8 x 10~^M) and sodium fluoride (1 x 10“®M), were found to be most effective towards inhibiting leucocyte mobility. The aerobic inhibitors, 2,4-dinitrophenol and potassium cyanide, were not inhibitory until concentrations greater than 1 x lO^^M were reached. The fact that leucocyte metabolism is mainly glycolytic and mobility is inhibited by glycolytic inhibitors would imply that the energy provided by glycolysis is a factor in amoeboid movement. The use of greater amounts of aerobic inhibitors, noted in current studies, would further suggest that aerobic systems affected by cyanide and dinitrophenol also contribute to the production of energy. Crude Holothurin in a concentration of 1 /xg/ml used concomitantly with these inhibitors showed the same migration patterns as inhibitors in the absence of Holothurin. However, a stimu- lation resulting from a low concentration of crude Holothurin, was evident in all experi- ments where this saponin was used in the ab- sence of inhibiting compounds. The stimulation of leucocyte migration in the presence of crude Holothurin at low concentrations (1 /xg/ml) was due to both aerobic and anaerobic metabo- lism. Summary 1 ) A stimulation of leucocyte migration oc- curred in the concentration range 0.1-6 ju.g crude Holothurin/ml blood, at 0.02 jxg Holothurin A/ml blood and 1.0 fjLg desulfated Holothurin A/ml blood. 2) An inhibition of leucocyte migration was produced in concentrations of 16-58 ^ig crude Holothurin/ml blood, at 0.1 /xg Holothurin A/ ml blood and 300 fjig desulfated Holothurin A/ ml blood. 3) Ouabain, a non-hemolytic saponin, inhib- ited leucocyte migration to the same extent as crude Holothurin when used in a concentration of 1.4 X 10~®M. ( 16 fxg/m\). 4) Liberated hemoglobin was found to be inhibitory in high concentrations, while lesser amounts did not alter white cell mobility. 5) Crude Holothurin (1.0 /xg/ml blood) used concomitantly with anaerobic inhibitors NaF and iodoacetic acid and with aerobic in- hibitors 2,4-dinitrophenol and KCN, showed no alteration in migration patterns different from that obtained in the absence of this saponin. Both aerobic and anaerobic metabolism are im- portant for the stimulation of leucocyte migra- tion in the presence of low concentrations of crude Holothurin. Literature Cited Chanley, J. D., R. Ledeen, J. Wax, R. F. Nigrelli AND H. SOBOTKA 1959. Holothurin. I. The isolation, properties and sugar components of Holothurin A. J. Amer. Chem. Soc. 81: 5180-5183. Chanley, J. D., J. Perlstein, R. F. Nigrelli and H. SOBOTKA 1960. Further studies on the structure of Holo- thurin. Ann. N. Y. Acad. Sci. 90: 902-905. Dourmashkin, R. R., R. M. Dougherty and R. J. C. Harris 1962. Electron microscopic observations on Rous Sarcoma virus and cell membranes. Nature 194: 1116-1119. Friess, S. L., F. G. Standaert, E. R. Whitcom, R. F. Nigrelli, J. D. Chanley and H. Sobotka 1959. Some pharmacologic properties of Holo- thurin, an active neurotoxin from the sea cucumber. J. Pharmacol. Exptl. Therap. 126: 323-329. Friess, S. L., F. G. Standaert, E. R. Whitcomb, R. E. Nigrelli, J. D. Chanley and H. Sobotka 1960. Some pharmacologic properties of Holo- thurin A, a glycosidic mixture from the sea cucumber. Ann. N. Y. Acad. Sci. 90: 893-901. Friess, S. L., R. C. Durant, J. D. Chanley and F. J. Fash 1967. Role of the sulfate charge center in irre- versible interactions of Holothurin A with chemoreceptors. Biochem. Pharmacol. 16: 1617-1625. Jarowska, S., R. F. Nigrelli, P. M. Murray and A. M. Veltri 1958. Hemopoietic effects of Holothurin, a ste- roid saponin from the sea cucumber, Actinopyga agassizi, in Rana pipiens. Anat. Rec. 132: 459. Ketchel, M. M. and C. B. Favour 1955. The acceleration and inhibition of migra- tion of human leucocytes in vitro by plasma protein fractions. J. Exptl. Med. 101: 647-663. Lasley and Nigrelli: Effect of Holothnrin on Leucocyte Migration 1 Nigrelli, R. F. 1952. The effect of Holothurin on fish and mice with Sarcoma-180. Zoologica 37: 89-90. Nigrelli, R. F. and P. A. Zahl 1952. Some biological characteristics of Holo- thurin. Proc. Soc. Exptl. Biol. Med. 81: 379-380. Nigrelli, R. F. and S. Jakowska 1960. Effects of Holothurin, a steroid saponin from the Bahamian sea cucumber (Actino- pyga agassizi), on various biological sys- tems. Ann. N. Y. Acad. Sci. 90: 884-892. Parpart, a. K. and R. Ballentine 1952. Molecular anatomy of the red cell plasma membrane. In: Modern Trends in Physiol- ogy and Biochemistry, ed. Barron, E. S. G., Academic Press, N. Y. pp. 135-148. Ponder, E. and J. McLeod 1936. The effect of hemolytic substances on white cell respiration. J. Gen. Physiol. 20: 267-281. Rideal, E. and E. H. Taylor 1958. On hemolysis and hemolytic acceleration. Proc. Roy. Soc. London Ser. B 148: 450- 464. Ruggieri, G. D. 1965. Echinoderm toxins-II. Animalizing action in sea urchin development. Toxicon 3: 157-162. Seeman, P. 1967. Transient holes in the erythrocyte mem- brane during hypotonic hemolysis and stable holes in the membrane after lysis by saponin and lysolecithin. J. Cell Biol. 32: 55-70. Snedecor, G. W. 1956. Statistical methods applied to experiments in agriculture and biology. 5th ed. Iowa State College Press, Ames, Iowa. Sullivan, T. D., K. T. Ladue and R. E. Nigrelli 1955. The effects of Holothurin, a steroid sapo- nin of animal origin, on Krebs-2 ascites tumors in Swiss mice. Zoologica 40: 49-52. Tennant, J. R. 1964. Evaluation of the Trypan Blue technique for determination of cell viability. Trans- plantation 2: 685-694. Thron, C. D. 1964. Hemolysis by Holothurin A, Digitonin and Quillaria: Estimates of the required cellular lysin uptake and free lysin con- centrations. J. Pharm. Exptl. Ther. 145: 194-202. White, A., P. Handler and E. L. Smith 1964. Principles of biochemistry, 3rd ed., McGraw-Hill Book Co., N. Y. Woods, M. W., M. Tandy, J. L. Whitby and D. Burk 1961. Symposium on bacterial endotoxins. HI. Metabolic effects of endotoxins on mam- malian cells. Bact. Rev. 25: 447-456. 8 New York Zoological Society: Zoologica, Spring, 1971 Table 1. Crude Holothurin Stimulation of Leucocyte Migration Using Whole Blood Concentration of Holothurin in iJ-g/ml Holothurin + leucocytes (migration in mm) Saline* + leucocytes ( migration in mm) Per cent viable Holothurin- treated leucocytes N P 0.1 1.94 1.63 100 30 < .001 2.55 2.38 — 30 < .001 2.49 2.11 — 30 <.01 1.88 1.79 - 30 <•1 0.5 2.68 2.11 — 30 < .001 2.02 1.63 100 30 < .001 2.22 2.10 — 30 < .001 2.61 2.38 — 30 < .001 1.96 1.79 - 30 <.01 1.0 1.97 1.63 99-100 30 < .001 2.01 1.79 _ 30 < .001 1.50 1.34 — 30 < .001 2.78 2.11 — 30 < .001 1.98 1.79 — 30 < .001 1.89 1.84 - 30 < .3 2.0 2.09 1.63 — 30 < .001 1.47 1.34 — 30 < .001 1.81 1.79 — 30 <■8 1.82 1.79 - 30 < .6 3.0 2.18 1.63 - 30 < .001 4.0 1.90 1.79 99-100 30 < .02 2.16 2.02 - 30 <1 6.0 1.84 1.79 - 30 < .5 * Control cells always showed 95-100% viability. Table 2. Crude Holothurin Inhibition of Leucocyte Migration Using Whole Blood Holothurin Saline* Per cent + + viable Concentration leucocytes leucocytes Holothurin- of Holothurin (migration (migration treated in pg/ml in mm) in mm) leucocytes N P 0.1 1.72 1.84 - 30 < .01 0.5 1.79 1.84 - 30 < .2 1.0 2.34 2.38 — 30 <•3 1.77 1.79 99-100 30 < -8 2.0 2.09 2.38 - 30 < .001 4.0 1.73 1.76 — 30 <•7 1.74 1.79 99-100 30 < .6 6.0 2.08 2.08 100 30 < 1 1.20 1.30 - 30 <■2 8.0 1.19 1.30 97 30 <•4 1.77 1.79 - 30 < .9 10.0 1.18 1.30 97-100 30 <■3 12.0 1.19 1.30 90 30 <•2 14.0 1.18 1.30 95 30 <•2 16.0 1.52 2.12 90 30 < .001 1.06 1.30 90-95 30 < .02 18.0 0.66 1.30 90 30 < .001 20.0 0.70 1.30 90-95 30 < .001 22.0 0.51 1.30 90 30 < .001 * Control cells always showed 95-100% viability. Lasley and Nigrelli: Effect of Holothurin on Leucocyte Migration 9 Table 3. The Effect of Holothurin A on Leucocyte Migration Using Whole Blood Concentration of Holo- thurin A in fxg/ ml Holothurin A + leucocytes (migration in mm) Saline* + leucocytes (migration in mm) Per cent viable leucocytes N P 20** none 1.62 98-100 28 — 0.1 1.53 1.80 99-100 20 < .001 0.025 1.46 1.51 99-100 23 <.05 * Control cells showed 100% viability. ** Much hemolysis present. Table 4. The Effect of Desulfated Holothurin A on Leucocyte Migration Using Whole Blood Concentration of desulfated Holothurin A in ixg/ml Desulfated Holothurin A + leucocytes (migration in mm) Saline* + leucocytes (migration in mm) Per cent viable leucocytes N P 1 1.24 1.70 100 33 < .01 20 1.69 1.62 100 28 <.8 100 1.38 1.43 100 34 < -4 300 0.68** 1.42 70 7 < .001 * Control cells showed 100% viability. ** Slight amount of hemolysis. Table 5. Crude Holothurin Inhibition of Leucocyte Migration Using White Blood Cell Suspensions Holothurin concentration in WBC sus- pension, pg/ml leucocytes Comparable Holothurin concentration in whole blood, fjLg/ml whole blood Holothurin- treated leucocytes (migration in mm) Saline-treated leucocytes* (migration in mm) Per cent viable Holothurin- treated leucocytes N P 40 23 1.54 1.67 99-100 15 < .2 100 58 0.90 1.67 85 20 < .001 300 174 0.66 1.67 30-50 14 < .001 1000 580 0.49** 1.67 5*** - < .001 * Control cells showed 99-100% viability. ** Few cells. *** Most cells disintegrated. 10 New York Zoological Society: Zoologica, Spring, 1971 Table 6. The Effect of Hemoglobin on Leucocyte Migration Compound Lapsed time of contact Klett O.D. A verage migration (in mm) Per cent viable N P Crude Holothurin (20 ;ttg/ml) 1 .5 hrs. 107 1.50 100 29 Hemoglobin (A ) - 103 2.44 100 20 <•8 Saline — — 2.42 100 34 Crude Holothurin (20/tg/ml) 4.0 hrs. 250 1.37 99 25 Hemoglobin (B ) - 270 1.97 100 20 < .001 Saline — — 2.42 100 34 Table 7. Leucocyte Migration in the Presence of Crude Holothurin and Ouabain Compound Concentration Average migration (in mm) Per cent viable N P Crude Holothurin 1.4 X 10-=M 1.52 90 20 <•5 Ouabain 1.4 X 10-=M 1.44 92-95 31 Saline - 2.12 95-100 28 Table 8. Leucocyte Migration in the Presence of Crude Holothurin and Sodium Fluoride Compound Concentration A verage migration (in mm) Per cent viable N 1 ) Holothurin ( 1 Mg/ml ) 8.8 X lO-’M 1.86 99-100 25 2) Sodium fluoride 8.0 X 10-“M 1.65 98-99 31 3) Holothurin (8.8 x 10“'M) + NaF 8.0 X 10-“M 1.68 100 29 4) Holothurin (8.8 x 10‘'M) + NaF 1.0 X 10-^M 1.61 99 30 5) Holothurin (8.8 x 10“'M) + NaF 1.0 X lO-'M 1.03 98-99 28 6) Saline - 1.63 98 28 7) Sodium fluoride 8.8 X lO-’M 1.72 99 25 8) Sodium fluoride 1.0 X lO-'M 1.10 99-100 28 9) Saline — 1.73 99-100 28 1) versus 6); P < .001 2) versus 6) ; P < .6 3) versus 6) ; P < .3 4) versus 6) ; P < .7 5) versus 6) ; P < .001 7) versus 9); P < .7 8) versus 9); P < .001 Lasley and Nigrelli: Effect of Holothnrin on Leucocyte Migration 11 Table 9. Leucocyte Migration in the Presence of Crude Holothurin and Iodoacetic Acid Compound Concentration A verage migration (in mm) Per cent viable N 1 ) Holothurin ( 1 ,ag/ml ) 8.8 X lO-’M 1.82 98-100 32 2) Iodoacetic acid 8.0x 10-=M 0.98 100 31 3) Holothurin (8.8 x 10"’M) + lAA 8.0 X 10-°M 0.91 97-100 31 4) Saline - 1.66 99-100 31 5 ) Iodoacetic acid 8.8 X lO-’M 1.65 100 20 6) Iodoacetic acid 1.0 X lO-'M 0.26 98-100 20 7) Saline — 1.53 99-100 18 1) versus 4); P < .001 2) versus 3) ; P < .3 2) versus 4); P < .001 3) versus 4) ; P < .001 5) versus 7); P < .2 6) versus 7); P < .001 Table 10. Leucocyte Migration in the Presence of Crude Holothurin and Dinitrophenol Compound Concentration A verage migration (in mm) Per cent viable N 1 ) Holothurin 8.8 X lO-’M 1.90 99-100 35 2) Dinitrophenol 8.0 X lO-'^M 1.65 100 35 3) Holothurin (8.8 x 10"’M) + DNP 8.0 X 10-=M 1.78 100 34 4) Saline - 1.73 100 36 5 ) Dinitrophenol 1.0 X 10-^M 1.64 100 26 6) Saline - 1.60 100 22 7 ) Dinitrophenol 1.0 X 10""M 1.50 95 20 8) Dinitrophenol 8.8 X lO-’M 1.55 100 15 9) Saline — 1.50 99-100 25 1) versus 4); P < .001 1) versus 3); P < .02 2) versus 4); P < .02 3) versus 4); P < .4 5) versus 6); P < .6 8) versus 9); P < .7 12 New York Zoological Society: Zoologica, Spring, 1971 Table 11. Leucocyte Migration in the Presence of Crude Holothurin and Potassium Cyanide Compound Concentration A verage migration (in mm) Per cent viable N 1 ) Holothurin 8.8 X lO-’M 1.79 98-100 31 2 ) Potassium cyanide 8.0 X 10-=M 1.54 99-100 35 3) Holothurin (8.8 x IQ-^M) + KCN 8.0 X 10““M 1.50 98 33 4) Saline - 1.50 99-100 33 5) Potassium cyanide 1.0 X lO-'M 1.60 99 27 6) Potassium cyanide 1.0 X lO-'M 1.66 98-99 29 7) Potassium cyanide 1.0 X 10-=M 0.76 99-100 27 8) Saline - 1.60 100 25 9) Potassium cyanide 8.8 X lO-’M 2.14 99-100 30 10) Saline — 2.14 100 27 1 ) versus 4); P < .001 2) versus 4); P < .5 6) versus 8) ; P < .7 7) versus 8); P < .001 News and Notes 13 NEWS AND NOTES New York Medical College to Sponsor Cooperative Program in Comparative Pathology A group of scientists at New York Medical College believes that a major contribution to mankind can be made from veterinary medicine within the next decade, and has instituted a cooperative program with local zoos in which a flow of information on spontaneously occurring animal diseases can be applied to the human health sciences. The Department of Pathology, through its chairman, David Spiro, M.D., Ph.D., has announced the formation of a Comparative Pathology Program in which animal disease models which have human counterparts will be studied in depth. Man shares many diseases with his non- human brothers and the program’s Directors, Ralph E. Strebel, Ph.D., associate professor of pathology, Edward Garner, D.V.M., assistant professor of pathology, and Emil Dolensek, D.V.M., veterinarian of the Bronx Zoo and assistant professor of comparative pathology at New York Medical College, believe that studies of these animals could shed light on many as- pects of human pathology. In order to provide future practitioners and research pathologists with a broad-based ap- proach to the pathogenesis of disease while also meeting the need for improved animal health management, the investigators plan to utilize material from a wide variety of wild, domestic, and zoo animals. The program functions as part of a coopera- tive interprofessional arrangement with the New York Zoological Society, through its director, William G. Conway, and Dr. Dolensek. Other cooperating institutions include the Staten Island Zoo, which is operated by the Staten Island Zoo- logical Society, and the Prospect Park, Flushing, and Central Park zoos, which are under the jurisdiction of New York City’s Parks, Recre- ation, and Cultural Affairs Administration, They are represented respectively by William Summer- ville, General Curator; Ronald Ellis, Supervisor; Mrs. Gilette Infante, Supervisor; and John Fitz- gerald, Supervisor. Calling material from these zoo populations a “vast unexplored reservoir of valuable disease models,” Dr. Strebel and Dr. Garner believe that material thus obtained will become a unique resource for the study of such conditions as arteriosclerosis, cancer, heart disease, diabetes, hepatitis, and many other diseases shared by man with animals. To expedite the search for animal diseases which might prove fruitful in comparative studies, the program will also call upon the services of two pathologists experi- enced in human disease. Dr. Henry I. Kobrin of New York Medical College, and Dr. John Budinger, consultant to the New York Zoologi- cal Society. Profiles of Normal Values Sought The medical treatment of zoo animals is hampered presently because of a dearth of in- formation on what constitutes the norm for blood values in all orders of zoo animals. There- fore, efforts will be made to develop compre- hensive normal blood profiles for zoo animal populations. This should greatly enhance future attempts to compare normal animals of differ- ent orders and species as well as to establish normal base lines necessary for the appropriate treatment of their diseases. Registry of Animal Diseases The information which will be derived from diagnostic and necropsy procedures will be utilized to develop a comprehensive registry of animal diseases. When completed, this registry will provide valuable reference material to in- vestigators, medical students, and graduate stu- dents studying comparative pathology. Graduate Degree Program Offered New York Medical College will offer qualified individuals a graduate degree program in com- 14 New York Zoological Society: Zoologica, Spring, 1971 parative pathology. The unique aspect of this program will be the depth of exposure to an exceptionally broad spectrum of animal as well as human pathology. In addition to compre- hensive course work and enrollment in the hu- man pathology program offered at New York Medical College, students will spend several months at the various zoos on a rotation basis. The broad scope of this orientation will afford the student an exposure to the pathogenesis of disease on a comparative basis in a wide variety of animal life. The combination of excellent training in both animal and human pathology is a rarity. A person, so exposed, would be in a superior position to make effective use of animal pathology and animal models of disease for pur- poses of study and research in regard to the pathogenesis of human disease. NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate, and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 1. The Effect of Holothurin on Leucocyte Migration. By B. J. Lasley and Ross F. Nigrelli 1 News and Notes New York Medical College to Sponsor Cooperative Program in Compara- tive Pathology 13 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx Park, Bronx, N. Y. 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, N. Y. Published May 5, 1971 ©1971 New York Zoological Society. All rights reserved. rio.S ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 56 . ISSUE 2 • SUMMER, 1971 I j i PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Edward R. Ricciuti, Editor & Curator, Publications & Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor EDITORIAL COMMITTEE Robert G. Goelet, Chairman; William G. Conway, Donald R. Griffin, Hugh B. House, F. Wayne King, Peter R. Marler, Ross F. Nigrelli, James A. Oliver, Edward R. Ricciuti, George D. Ruggieri, S.J. William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology; Hugh B. House, Curator, Mammalogy; James G. Doherty, Assistant Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Robert A. Brown, Assistant Curator, Animal Departments; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus AQUARIUM James A. Oliver, Director; H. Douglas Kemper, Assistant Curator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology; Jay Hyman, Consultant Veterinarian OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director & Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempier, Jr., Assistant to the Director & Bio-Organic Chemist; Eli D. Goldsmith, Scientific Consultant; William Antopol, Research Associate, Comparative Pathology; C. M. Breder, Jr., Research Associate, Ichthyology, Jack T. Cecil, Virologist; Harry A. Charipper, Research Associate, Histology; Paul J. Cheung, Microbiologist; Erwin J. Ernst, Research Associate, Estaurine & Coastal Ecology; Kenneth Gold, Marine Ecologist; Jay Hyman, Research Associate, Comparative Pathology; Myron Jacobs, Neuroanatomist; Klaus Kallman, Fish Geneticist; John J. A. McLaughlin, Research Associate, Planktonology ; Martin F. Schreibman, Research Associate, Fish Endocrinology INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR (Operated jointly by the Society and the Rockefeller University) Peter R. Marler, Director & Senior Research Zoologist; Paul Mundinger, Assistant Director & Research Associate; Donald R. Griffin, Senior Research Zoologist; Jocelyn Crane, Senior Research Zoologist; Roger S. Payne, Research Zoologist; Fernando Nottebohm, Research Zoologist; George Schaller, Research Zoologist; Thomas T. Struhsaker, Research Zoologist; Alan Lill, Research Associate ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Ben Sheffy, Consultant, Nutrition; Roy Bellhorn, Consultant & Research Associate, Comparative Ophthalmology ; John Budinger, Joseph Conetta, Edward Garner, Henry Kobrin, Ralph Strebel, Consultants & Research Associates, Comparative Pathology 2 Species Identification of Commercial Crocodilian Skins F. Wayne King’ and Peter Brazaitis- ( Figures 1-41) Gross similarities in the morphology of crocodilian skins has made specific identification of individual commercial hides extremely difficult. Qualitative and quantitative differences between the hides of various species are defined and form the basis of a key to commer- cial crocodilian hides. The distribution, common, and commercial names, distinguishing characteristics of the hides, and the status of the wild populations of each of the 27 species and subspecies is given. Introduction To THE LAYMAN most crocodiUans look similar — all have relatively long toothy snouts, scaly backs, flattened oar-like tails, large webbed hindfeet, and most have crossbanded color patterns. As a consequence, most Americans mentally lump all crocodilians under the collective heading “alligator.” Pro- found differences exist between the species of crocodilians, but the untrained eye notices the gross similarities rather than the less obvious differences. When it comes to identifying a spe- cies of crocodilian from a commercial hide, however, even a trained herpetologist faces serious difficulty. All commercial skins are grossly alike. All crocodilian leather is retailed throughout the United States as “alligator,” while in Europe, Africa, and Asia the same hides are sold as “crocodile.” In this paper, we attempt to provide means to identify commercial crocodilian hides. Since the paper will be read by layman, trained herpe- tologist, government inspector, and commercial dealer alike, we have endeavored to use termi- nology comprehensible to all. Where there is a chance of confusion, we have provided photo- graphs and line drawings for clarification. Materials and Methods Comparisons were made between the skins of live specimens in zoos and private collec- tions, preserved specimens and dried skins in museum collections, and tanned and finished ’ Curator of Herpetology, New York Zoological Park, Bronx, New York 10460. “ Assistant Animal Manager, Department of Herpetol- ogy, New York Zoological Park, Bronx, New York 10460. commercial skins supplied by the Reptile Prod- ucts Association of the United States. A total of over 350 specimens were examined. Museum specimens of every species and subspecies were studied. Living specimens of every form, ex- cept Caiman crocodihis apaporiensis and Croco- dyhis siamensis, were seen. Commercial hides of most species were examined. The notable exceptions were Gavialis gangeticiis. Alligator sinensis, Paleosuclnis palpehrosus, Paleosachus trigonatus, Crocodylns paliistris, and Crocody- lus rliombifer. The raw data are on deposit at the New York Zoological Park. The characters and terms used in this study are defined below. COMMERCIAL HIDES. Hides used in the crocodilian hide trade for the manufacture of leather goods are termed commercial hides, whether they are raw skins or are in the process of being tanned and finished. HORNBACK HIDES. Rough dorsal (back) skins obtained by skinning the animals begin- ning from an incision made along the midventral (belly) line. Large bony dorsal scales, usually with raised keels, occupy the center of the hide. Smooth squarish scales from the ventral surface are located along the lateral edges of the hide. Hornback hides usually are skinned from rela- tively small specimens since the heavily ossified dorsal scales of adults make their hides stiff and limits its use for leather. Skin from the tail and proximal portion of the legs is attached to the hide (figure 1 ) . BELLY HIDES. Smooth ventral (belly) skins obtained by skinning the animal beginning at an incision just below the large bony dorsal scales high on one side and continuing down the side, under the body, and up the other side to the 15 16 New York Zoological Society: Zoologica, Summer, 1971 edge of the large dorsal scales on that side of the back. The scales of belly skins are squarish or rectangular in shape over most of the center of the skin. The round or oval scales from the side of the body are located along the lateral edges of the hide. The vent is represented by a hole along the midline of the skin. Skin from the tail and proximal portion of the legs is at- tached (figure 1). BUTTON-BELLY AND SOLT-BELLY HIDES. Belly skins that possess osteoderms, or buttons (the commercial name), are called button hides. Belly skins that lack osteoderms are called soft-belly hides. Button-belly hides will not flex through the middle of a scale be- cause of the osteoderm button. Soft-belly hides not only flex between scales but also bend to a lesser degree through the middle of scales (fig- ure 2). As a result the most sought after hides are soft-bellies. Button-bellies with large osteo- derms are most frequently used for making the flat, non-flexing sides of purses or attache cases. The stiffness of these skins limits their use where flexibility is demanded, as in shoes, belts, and watch-bands. SIDES. A narrow strip of soft skin taken from a point under the lower jaw and extending back over the front leg, along the side of the trunk, under the rear leg and ending near the vent (figure 15). Such strips are skinned from large caimans (Caiman, Melcmosacluis, Paleosiicluis) which have such heavy osteoderm buttons in the belly skin, as to make this skin almost worth- less as leather. The hide-hunters avoid both the bony dorsal scales and the bony belly scales by removing only the soft side skin. The scales on sides consist of large, round to diamond-shaped scales separated by soft skin or small scales. The largest scales usually possess osteoderm buttons and may have a low keel. Sides could, of course, be cut from soft- bellied alligators and crocodiles, as well as button-bellied caimans. The reason they are not is that soft-belly hides are worth more with the sides attached than they are with the strips re- moved. The best tanned soft-belly hides sell for as much as $12.00 per square foot. Tanned sides sell for only $5.00 to $15.00 apiece. THROATS, SIDES, AND GIRDLES. Throats are V-shaped pieces skinned from under the chin and the sides of the neck. Girdles are taken from the thighs and belly immediately anterior to the vent. Sides, which accompany throats and girdles, are stripped from the sides of the trunk between the legs (figure 20). These cuts of skin are from exceptionally large cai- mans (Melanosachus, and possibly Caiman) which have heavy bony belly osteoderms. The large size of the scales on these pieces attests to the size of the animals they come from. ELIPPERS. The small, irregular-shaped pieces of skin from the legs. The scales are usually uniformly small, smooth, and squarish. RAW HIDES OR SALTED HIDES. Un- tanned commercial hides. They may be dry or moist from the salt. They usually are rolled up for shipment, and retain the color pattern of the live specimen — most frequently dark spots or dark crossbands on the side of the trunk and tail. CRUSTS. Hides which have been tanned, but not dyed or polished. Crusts are usually ash gray or tan, and have a dull, unpolished finish (figures 9, 10, and 15). The next step in the finishing process is to bleach or dye the hide its final color. If it is a button-belly, the osteo- derms are shaved from the inside of the hide to eliminate as much of them as possible. It is not possible to remove every osteoderm button in its entirety, but it is possible to remove enough of them to make a stiff hide much more flexible (figures 14 and 19). Unskilled tanners may shave the hide too closely and leave thin, weak sections between the rows of scales. POLISHED AND EINISHED HIDES. After dyeing, skins customarily are given a high-gloss finish by burnishing the scales under the pres- sure of a polishing wheel. The glossy finish is characteristic of most crocodilian products. After the hide has been finished, it is ready for cutting into the pieces that are to be made into the manufactured product. FLAT FINISH AND BOMBE FINISH. Fin- ished skins with flat, level scales are flat finished skins. Those in which the individual scales are slightly curved, with the center of each one arching up from the crease where it meets adja- cent scales, are given the French name bomhe. SALVAGE FINISH. Not all finished hides have a highly polished surface. Some are fin- ished with a process that retains much of the texture of the original crust. The result is a textured, non-glossy, oiled-leather appearance called sauvage (figure 9). VENTRAL SCALES AND LATERAL SCALES. Scales from the underside of the throat, body, and tail are ventral scales (figure 1 ) . They are large and squarish in all croco- dilians. The individual square scales are in con- tact with adjacent scales and are arranged in rows across the belly of the animal. Scales from the side of the body are lateral scales (figure 1). King and Brazaitis: Species Identification of Commercial Crocodilian Skins 17 They frequently are arrayed in two distinct size classes, the larger composed of oval and dia- mond-shaped scales. These individual scales usually are not in contact with adjacent scales. Ventral and lateral scales possess a number of characteristics which may be used to identify a particular species or group of species. Most commercial skins used in the United States are either belly skins or sides, and since even horn- back hides normally retain ventral scales along their edges, only the ventral and lateral scales are considered in this paper. BUTTONS. The bony osteoderms present in the scales of many crocodilians. The presence of buttons in the ventral scales of belly skins gives them their commercial name, button-belly hides. Buttons can be seen if the hide is turned over and the backside (inside) examined. Even though they are decalcified during tanning, the osteoderm buttons are a dilferent color. In crusts and light-colored finished skins, buttons are usually slightly darker than the rest of the skin (figures 13 and 14). In dark finished skins, they may be lighter than the surrounding tissue, al- though they are usually darker (figures 19, 34, 35, 36, and 37). In many cases buttons also can be detected from the front side (outside) of a finished skin as ill-defined light-colored blotches in the center of the ventral scales (figures 18 and 34). They may also be evident as raised or slightly sunken areas on the surface of the ventral scales. They can best be seen by holding the polished surface of the scale in a position where light reflected off its surface reaches your eye. In this position any surface irregularity becomes apparent (fig- ures 35 and 41 ) . SURFACE PITTING. During the tanning process, the bony osteoderms do not shrink as much as the non-calcified tissue. As a result, button skins may exhibit some of the pock- marked or wrinkled texture of the underlying osteoderms. This condition is called surface pit- ting. Surface pitting is best seen where the osteo- derms are most heavy, as in the dorsal scales of hornback hides, and the ventral scales of the skins of caiman (Caiman, Melanosucluis, Paleo- snchus), slender-snouted crocodile (Crocody- lus cataphractus) , dwarf crocodile (Osteolae- mus), or button-belly Nile crocodile (Croco- dyliis niloticiis) (figures 10, 11, 12, 18, 36, and 37). The skins of some species are more strongly pitted than others; adult specimens tend to be more pitted than juveniles; and not all indi- viduals exhibit the same degree of pitting throughout all their scales. SINGLE BUTTONS AND DOUBLE BUT- TONS. Those crocodile and alligator specimens which have osteoderms are characterized by having single buttons, one osteoderm button per scale (figures 8, 34, 36, and 37), while all caimans have double buttons, two osteoderms per scale (figures 13, 14, and 19). In caimans, the posterior of the two buttons occupies almost the entire area of the scale. The anterior button occupies the anterior one-quarter of the scale and curves upward and inward (mesially) to protectively overlap the posterior margin of the next anterior scale and the intervening soft skin. The overlapping of osteoderms is an evolution- ary adaptation that affords more protection to the weak hinge point between the scales, although it causes the loss of some freedom of movement. The dwarf caimans { Paleositchus) , with pro- portionally the largest double buttons of any species of crocodilian, are well on the way to evolving the stiff analog of a turtle plastron. The two parts of the caiman double button are more evident before the buttons are shaved during finishing (figure 14). During shaving, the ante- rior, inward-curving button is nearly completely removed. In most cases, however, part of the anterior button remains even after the shaving is complete (figure 19). FOLLICLE GLAND. Belly scales of gavials and all crocodiles have a pit-like structure, the follicle gland, near the posterior margin (fig- ures 22 and 24). The glands are clearly visible in live specimens, raw skins, and crusts. In fin- ished skins dyed dark, the follicle glands may be lighter in color (figure 25). The glands may be partly obscured during the polishing process. If this happens, the gland usually can be seen as a short deep wrinkle running from the gland to the posterior edge of the scale (figure 35). In finished skins, follicle glands are detected most easily on the throat and in the area just anterior to the vent as well as on the underside of the tail (figures 28 and 30). SPIDER-WEB UMBILICUS. Alligator hides have neither surface pitting nor follicle glands. The ventral scales of finished alligator skins are glossy smooth without ornamentation. It is also possible to distinguish the belly skin of the American alligator (Alligator mississippiensis) from all other species on the basis of the shape of the umbilicus scar. Crocodilians with buttons in the ventral scales tend to lose all evidence of the umbilicus once it is healed and the osteo- derms are formed. Other species may retain the posterior one-third of the scar as a zigzag ar- rangement of small scales scattered along the midline just anterior to the vent. In the Ameri- can alligator, and no other species, this same 18 New York Zoological Society: Zoologica, Summer, 1971 posterior portion of the scar remains as an area of soft skin lacking scales, and because of the profusion of creases and lines it has a distinctly spider-web appearance (figure 7). VENTRAL COLLAR. Most crocodilians have a prominent row of enlarged scales, called a ventral collar, across the throat just anterior to the front legs (figure 1). A few species lack an enlarged row of scales, so the collar is not conspicuous. One species has a double collar, two enlarged rows of scales. TRANSVERSE VENTRAL SCALE ROWS. Ventral scales are arranged in transverse rows in all crocodilians, but the number of rows found between the neck and vent differs between spe- cies. The transverse rows of ventral scales are counted from the first row posterior to the ven- tral collar to, but not including, the row of scales encircling the vent (figure 1). The row of scales around the vent may be missing from a commercial hide because the hide-hunter was careless when skinning the animal. In that event the position of the missing rows must be esti- mated. Only rows which cross the midline -are counted. Incomplete or missing rows will add confusing variation to the count, so to eliminate doubt, the count first should be made only to the right of the midline and then repeated on the left side, and the two counts compared. LARGE-SCALE AND SMALL-SCALE HIDES. Soft-belly crocodilians which have 26 to 35 transverse ventral scale rows are called small-scale hides by the hide trade. Soft-belly species with 20 to 25 transverse ventral scale rows are called large-scale hides (see the key that follows and figure 26). TAIL WHORLS. The transverse rows of scales under the tail are the ventral portions of the whorls of scales that completely encircle the tail. The ventral portion of these whorls, like the rows of ventral scales on the body, are usu- ally complete and evenly arranged (figure 28). Morelet’s crocodile (Crocodylus moreletii), however, possesses irregular or incomplete whorls 66 percent of the time (figure 30). No other species shows as high an incidence of irregularity in this character. Identification of Crocodilian Hides The following keys can be used to identify the species, or species groups, of commercial belly skins, hornbacks, and sides. The keys are of limited use in identifying throats and girdles, and are useless for flippers. They may be use- less in identifying skins already manufactured into finished products. Keys are identification tools which employ a series of alternative choices. To use the keys, first decide whether or not the hide you wish to identify is a side, belly, or hornback. Once this determination has been made, proceed to the appropriate key. Each set of alternative choices, or couplets, is numbered. Starting with couplet 1, decide which of the two choices, “a” or “b,” best describes the hide to be identified. The number that follows the correct choice indi- cates the next couplet. By moving from couplet to couplet following the numbers shown after each correct choice, you will arrive at a final choice which indicates the species, or species group, of crocodilian from which the hide was taken. Once the identification has been made, you should turn to the text that follows the keys for information on the distribution of the species, the commercial names under which it is sold, additional distinguishing characteristics, and status of the wild populations. Species identifications supplied by manufac- turers are not to be relied on until verified by means of the keys. In the past two years, the authors have seen live African slender-snouted crocodiles and South American caimans shipped into the United States from Bangkok, Thailand, as Siamese crocodiles; finished African dwarf crocodile hides enter from a tanner in Erance who labelled them gavial; and wallets made from South American caimans arrive from an Italian manufacturer who declared they were Nile crocodile. A KEY TO COMMERCIAL CAIMAN SIDES The key to sides is based on the assumption that the sides being identified are from caimans {Caiman, Melanosiichiis, or Paleosiichns) , and not from other species. At the present time, caimans are the only crocodilians being skinned in this manner. This may not be the case at some future date. In addition, small finished products such as belts may be pieced together from scrap left over from the manufacture of large belly hide products. These small pieces may come from any species, therefore, the key is of little use in identifying pieced items. 1. a) Rows of large oval scales alternating with rows of small scales (figure 21) . . . Melanosiichiis niger. b) Rows of large scales alternating with network of creases and small irregular scales (figure 16) ... 2 2. a) Large oval scales, usually smooth, and arranged in distinct rows . . . Caiman crocodiliis (four subspecies), Caiman kitirostris. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 19 b) Large oval scales usually keeled, and usually not arranged in distinct rows . . . Paleosuchiis palpehrosits, Paleo- siicluis trigonatiis. A KEY TO CROCODILIAN BELLY SKINS AND HORNBACK SKINS This key is for use with belly skins. Hornback hides can be identified if you limit your atten- tion to the belly scales found along the lateral edges of the hide. Surface pitting is not evident in untanned hides. 1. a) Ventral (belly) scales with follicle glands (figures 22 and 24) ... 2 b) Ventral (belly) scales without follicle glands ( figures 4, 1 1, and 18) ... 4 2. a) Osteoderm buttons present ( figures 34 through 37) ... Crocodyliis cata- phractiis, Crocodyliis niloticiis, Osteo- laeniits tctraspis (two subspecies), b) Osteoderm buttons not present (figure 27) ... 3 3. a) Transverse rows of ventral scales 20 to 25 . . . Crocodyliis iicutiis (south of Panama), Crocodyliis intermedins, Crocodyliis jolinsoni, Crocodyliis no- viiegiiineae (two subspecies), Tomi- stonia sclilegelli. b) Transverse rows of ventral scales 26 to 35 . . . Crocodyliis aciitiis (north of Panama), Crocodyliis moreletii, Cro- codyliis niloticiis, Crocodyliis paliis- tris (two suhspecies), Crocodyliis po- rosiis, Crocodyliis rhombifer, Croco- dyliis sianiensis, GaviaUs gangeticiis. 4. a) No osteoderm buttons present in mid- belly (figure 6), or single buttons present ( figure 8 ) ... 5 b) Double osteoderm buttons present in midbelly (figures 14 and 19) ... 6 5. a) Umbilicus scar has spider-web ap- pearance (figure 7); transverse rows of ventral scales 29 or more . . . Alli- gator mississippiensis. b) Umbilicus scar not evident or lacks spider-web appearance; transverse rows of ventral scales 28 or fewer . . . A lligator sinensis. 6. a) Large osteoderm buttons present me- dially only, not over pelvic girdle (figure 19); surface pitting slight; transverse rows of ventral scales 25 or more . . . Melanosiicliiis niger. b) Large osteoderm buttons in all large ventral scales, throat to pelvis (figure 13); surface pitting slight to pro- nounced; transverse rows of ventral scales 1 8 to 30 ... 7 7. a) Surface pitting pronounced; trans- verse rows of ventral scales 20 to 30 ... 8 b) Surface pitting slight or absent; trans- verse rows of ventral scales 18 to 22 . . . Paleosuchiis palpehrosiis, Paleo- sitcliiis trigonatiis. 8. a) Transverse rows of ventral scales 26 to 30; double ventral collar . . . Caiman latirostris. b) Transverse rows of ventral scales 20 to 27; single ventral collar . . . Caiman crocodiliis (four subspecies). In the text that follows, the species and sub- species are listed alphabetically by scientific name within each family. The systematic ar- rangement follows Wermuth and Mertens ( 1961 ). Family ALLIGATORIDAE AMERICAN ALLIGATOR Alligator mississippiensis (Daudin) DISTRIBUTION. Southeastern United States — the states of North Carolina, South Carolina, Georgia, Florida, Alabama, Mississippi, Loui- siana, Arkansas, and Texas. This species does not occur outside the United States (Schmidt, 1953; U.S. Department of Interior, 1968; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the Florida and the Mississippi alligator, or gator. Hides are marketed as American, Florida, or Louisiana alligator or soft-belly. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. No osteoderm buttons (large specimens from Florida have single osteoderm buttons on the throat). Transverse ventral scale rows 29 or more. Umbilicus scar prominent and with spider-web appearance. Maximum length of live specimen is 18 feet. STATUS OF WILD POPULATIONS. En- dangered (Honegger, 1968; Pan American Union, 1967; U.S. Department of Interior, 1 968 ) . Now protected by state law in every state in which it occurs; by federal prohibition on in- terstate traffic in illegal hides; and by local and state prohibitions on sales of live specimens, hides, and hide products. 20 New York Zoological Society: Zoologica, Summer, 1971 CHINESE ALLIGATOR Alligator sinensis Fauvel DISTRIBUTION. The lower Yangtze River drainage of China (Pope, 1935; Wermuth, 1953; Wermuth and Mertens, 1961 ; Werner, 1933). OTHER COMMON NAMES. In China it is called To, Ton Lung, Yow Lung. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skims: No follicle glands. Usually no osteoderm buttons, but occasionally single buttons may he present in the midbelly and collar areas. Transverse ventral scale rows 28 or fewer. Umbilicus scar not evident. Maximum length of live specimen is 6V2 feet. STATUS OE WILD POPULATIONS. Prob- ably endangered (Honegger, 1968). This species is known from a strip of territory only a few hundred miles long. A. E. Oeming of the Alberta Game Farm, Canada, recently returned from a trip to China and reported (in litt.) that the species is totally protected by law and the law is rigidly enforced. Dr. Cheng, of the Institute of Zoology, Academia Sinica, Peking, is study- ing the species. SOUTH AMERICAN CAIMAN Caiman crocodilus crocodilus (Linnaeus) DISTRIBUTION. Northern South America — Colombia east of the Andes, Peru, Ecuador, Venezuela, Guyana, Surinam, French Guiana, Trinidad, and, with the exception of a few southern tributaries, the Amazon drainage of Brazil [the exceptions are listed under Caiman crocodilus yacare'] (Carvalho, 1955; Medem, 1968; Schmidt, 1928b; Wermuth, 1953; Wer- muth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In the United States it is frequently called spectacled caiman (and equally frequently given the old synony- mous scientific name Caiman sclerops). In Cen- tral and South America it is called “alligator,” baba, babilla, cachirre, caiman, caiman bianco, caiman del Paraguay, cascarudo, cocodrillo, jacare, jacare de Lunetas, jacaretinga, lagarto, lagarto bianco, lagarto negro, ocoroche, tinga, and yacare bianco. Hides are frequently mar- keted under these names. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons present. Surface pitting evident. Transverse ven- tral scale rows 20 to 24. Single prominent ventral collar. Highest point of ventral scale slightly anterior of center (figure 39). Sides: Rows of large oval scales alternating with network of small irregular scales and creases. (The network is actually soft skin folds and creases without scales.) Maximum length of live specimen is 8V2 feet. STATUS OF WILD POPULATIONS. Most wild populations are declining and some have all but disappeared due to slaughter by hide hunters and capture by live animal collectors (Pan American Union, 1967). The subspecies is con- sidered endangered by some experts (Honegger, 1968). South American countries require that hides be tanned before export. Colombia pro- tects specimens less than 1.2 meters in length (Medem, 1970, in litt.) and Peru protects those less than one meter long (Crowe, 1965). RIO APAPORIS CAIMAN Caiman crocodilus apaporiensis Medem DISTRIBUTION. Colombia - known only from the Apaporis River and its tributaries be- tween the Falls of Jirijirimo and Puerto Yaviya (Medem, 1955, 1968; Wermuth and Mertens, 1961). OTHER COMMON NAMES. In Colombia it is called babilla, cachirre, cocodrillo, jacaretinga, lagarto negro and ocoroche. If marketed, hides would be sold under these names, or as tinga. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: No follicle glands. Eull double osteoderm buttons present. Surface pitting evident. Transverse ven- tral scale rows 20 to 24. Single prominent ventral collar. Highest point of ventral scale is slightly anterior of center (figure 39). Sides: Rows of large oval scales alternating with network of small irregular scales and creases. Maximum length of live specimen is 7 feet. STATUS OE WILD POPULATIONS. Criti- cally endangered. This subspecies has the most restricted range of any crocodilian. It is known only from an area 125 miles long in one river. Hide hunters can completely decimate this form in one or two years unless hunting is prohibited immediately. Colombia prohibits the export of untanned hides and protects specimens less than 1.2 meters in length (Medem, 1970, in litt.). BROWN CAIMAN Caiman crocodilus fuscus (Cope) DISTRIBUTION. Central America - south- ern Mexico to Colombia, west of the Andes (Medem, 1968; Schmidt, 1928b; Smith and Taylor, 1950; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In the United States it is also called spectacled caiman. Central King and Brazaitis: Species Identification of Commercial Crocodilian Skins 21 American caiman, dusky caiman, and Magda- lena caiman. In Central America it is called “alligator,” caiman, cocodrillo, and cuajipal. In Colombia it is known as bahilla, lagarto negro, and jacaretinga. Hides are marketed under these names, and Central American tinga. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons present. Surface pitting evident. Transverse ven- tral scale rows 20 to 24. Single prominent ventral collar. Highest point of ventral scale slightly anterior of center (figure 39). Sides: Rows of large oval scales alternating with network of small irregular scales and creases. Maximum length of live specimen is 7 feet. STATUS OF WILD POPULATIONS. Many wild populations are disappearing due to hide- hunting (Pan American Union, 1967). The sub- species is considered endangered by some ex- perts (Honegger, 1968). South American coun- tries prohibit the export of untanned hides. Colombia and Panama protect specimens less than 1.2 meters in length (Medem, 1970, in litt: D. Tovar, 1970, in lift.). Peru protects speci- mens less than 1.5 meters in length (Honegger, 1968). Mexico’s laws regulate hunting of this species. YACARE Caiman crocodiliis yacare (Daudin) DISTRIBUTION. Southern South America — specifically the Paraguay and Parana river drain- age systems of Paraguay, Uruguay, Argentina, and Brazil, and the southern tributaries of the Amazon in Bolivia [the Mamore, Itenez, and Beni drainages] and Brazil [the Guapore drain- age, and the Araguaia River above its confluence with the Tapirape] (Carvalho, 1955; Medem, 1968; Schmidt, 1928b; Wermuth, 1953; Wer- muth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the Paraguay caiman and red caiman in the United States. In South America it is called caiman del Paraguay, cascarudo, jacare, jacare de Limetas, jacaretinga, lagarto, tinga, yacare, and yacare de hocico angosto. Hides are mar- keted under these names. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons present. Surface pitting evident. Transverse ven- tral scale rows 20 to 25. Single prominent ventral collar. Highest point of ventral scale is slightly anterior of center (figure 39). Sides: Rows of large oval scales alternating with network of ir- regular shaped small scales and creases. Maxi- mum length of live specimen is 8 feet. STATUS OF WILD POPULATIONS. En- dangered (Pan American Union, 1967; U.S. Department of Interior, 1970). Most wild popu- lations declining in numbers (Jose Cei, 1970, in litt.). South American countries prohibit the export of untanned hides. Its import is pro- hibited under provisions of the Endangered Species Conservation Act (U.S. Department of Interior 1970) . BROAD-SNOUTED CAIMAN Caiman latirostris (Daudin) DISTRIBUTION. Southern South America — the drainages of the Paraguay, Parana, and Uruguay rivers in Argentina, Uruguay, Para- guay, and Brazil, and the rivers emptying into the southeast coast of Brazil south of Recife (Carvalho, 1955; Medem, 1968; Schmidt, 1928b; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In South America it is called jacare de Papo Amarelo, overo, itriiraii, and yacare de hocico ancho. Hides are marketed under these names. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons present. Surface pitting slight or absent. Trans- verse ventral scale rows 26 to 30. Double ventral collar. Highest point of ventral scale located at center of scale. Sides: Rows of large oval scales alternating with network of irregular shaped small scales and creases. Maximum length of live specimen is 9 feet. STATUS OF WILD POPULATIONS. En- dangered (Pan American Union, 1967). This species is nearly extinct from excessive hide hunting (Jose Cei, 1970, in litt.). South Ameri- can countries prohibit export of untanned hides. BLACK CAIMAN Melanosuchiis niger (Spix) DISTRIBUTION. Northern and central South America — Amazon basin drainages of Brazil, Colombia, Venezuela, Guyana, Peru, and Bolivia (Carvalho, 1955; Medem, 1963, 1968; Schmidt, 1928b; Wermuth, 1953; Wer- muth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In South America it is called asu, caiman, caiman negro, cocodrillo, jacare agii, jacare assit, jacare asu, jacare uassii, jacare una, and yacare assii. Hides are marketed under these names. 22 New York Zoological Society: Zoologica, Summer, 1971 DISTINGUISHING CHARACTERISTICS OF COMMERCIAT HIDES. Belly skins: No follicle glands. Full double osteoderm buttons, at least medially. Lateral scales may lack osteo- derms or possess small osteoderms in center of scales. Surface pitting slight. Transverse ventral scale rows 25 to 28. Sides: Parallel rows of large oval scales alternating with rows of small oval scales. Maximum length of live specimen is 16 feet. STATUS OF WILD POPULATIONS. En- dangered (Honegger, 1968; Pan American Union, 1967). Rapidly declining everywhere, and exterminated in many areas. South Ameri- can countries prohibit the export of untanned hides. Peru prohibits the killing of specimens less than 2 meters in length (Honegger, 1968). DWARF CAIMAN Paleosiicluis palpebrosus (Cuvier) DISTRIBUTION. Northern and central South America — Amazon and Orinoco river drainages of Colombia, Venezuela, Guyana, Brazil, Peru, Ecuador, and Bolivia (Carvalho, 1955; Medem, 1967, 1968; Schmidt, 1928b; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called musky caiman and Cuvier’s smooth-fronted caiman. In South America it is called cachirre, jacare corod, and yacare corod. Hides are mar- keted under these names. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons on all large ventral scales. Surface pitting slight or absent. Transverse ventral scale rows 18 to 22. Single prominent ventral collar. Sides: Large scales scattered, not in well-defined rows, and separated by wide areas of soft skin. Maximum length of live specimen is 5’/2 feet. STATUS OF WILD POPULATIONS. De- clining in numbers (Pan American Union, 1967). Paleosuchns is possibly the least perse- cuted of the crocodilians at the present time. Its small size and heavy osteoderm buttons make the skins less desirable than skins from the larger caimans and crocodiles of South America. South American countries prohibit the export of untanned hides. SMOOTH-FRONTED CAIMAN Paleosuchns trigonatus (Schneider) DISTRIBUTION. Northern and central South America — the Amazon and Orinoco river drainages of Colombia, Venezuela, Guyana, Brazil, Ecuador, Peru, and Bolivia (Carvalho, 1955; Medem, 1967, 1968; Schmidt, 1928b; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called Schneider’s smooth-fronted caiman. In South America it is called cachirre, jacare corod, ja- care ciirud, and yacare corod. Hides are mar- keted under these names. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: No follicle glands. Full double osteoderm buttons on all large ventral scales. Surface pitting slight or absent. Transverse ventral scale rows 18 to 22. Single prominent ventral collar. Sides: Scattered large keeled oval scales, not in well- defined rows, and separated by wide areas of soft skin. Maximum length of live specimen is 7 feet. STATUS OF WILD POPULATIONS. De- clining in numbers (Pan American Union, 1967). Paleosuchns is possibly the least perse- cuted of the crocodilians. Its small size and heavy ossification of the osteoderms makes the skins less desirable than skins from the larger caimans and crocodiles of South America. South American countries prohibit the export of un- tanned hides. Family CROCODYLIDAE AMERICAN CROCODILE Crocodylns acntns Cuvier DISTRIBUTION. Florida, West Indies, Cen- tral and northern South America — southern Florida, Cuba, Hispaniola (Haiti and Domini- can Republic), Jamaica, Mexico south to Co- lombia and Venezuela, exclusive of the Orinoco river drainage system (Cochran, 1941; Medehi, 1968; Smith and Taylor, 1950; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In Central America and Cuba it is called caimdn, and in South America it is known as caiman and cai- indn de agnja. Hides may be marketed under these names, or simply as Central or South American “alligator,” crocodile, soft-belly, small scale (north of Panama) or large scale (south of Panama) . DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 25 to 35. Tail whorls regular ventrally. Maximum length of live specimen is 23 feet. King and Brazaitis: Species Identification of Conunercial Crocodilian Skins 23 STATUS OF WITD POPULATIONS. De- clining everywhere due to excessive hide- hunting (Pan American Union, 1967). The species is considered endangered by some ex- perts (Honegger, 1968). Many populations in Central and South America have been totally exterminated. The species is protected by state law in Florida, and South American countries prohibit the export of untanned hides. Mexico regulates the hunting of the species, as does Nicaragua. Jamaica prohibits the export of crocodiles, their eggs, or skins (K.C. Hall, 1970, in litt.). The species is protected in Cuba and Colombia, although the law is not enforced in the latter (Honegger, 1968). AFRICAN SLENDER-SNOUTED CROCODILE Crocodyhis cataphractus Cuvier DISTRIBUTION. Western and central Africa — the Congo, Niger, and Volta river drainages, and the coastal rivers from Senegal south to northern Angola. Only once recorded from East Africa at Ujiji, Tanzania, on Lake Tanga- nyika (Schmidt, 1919; Wermuth, 1953; Wer- muth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is sometimes called the West African crocodile, African long- nosed crocodile, African gavial, or sub-water crocodile. Hides are sold under these names or as Nigerian, Congo, or Cabinde “alligator,” crocodile, or button hides. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: Fol- licle glands present. Round or elliptical single osteoderm buttons present. Surface pitting may or may not be present. Transverse ventral scale rows 25 to 29. Hides from Nigeria usually are missing the tip of the tail, due to local hunting practices. Skins from other parts of Africa usually have complete tails. Maximum length of live specimen is 13 feet. STATUS OF WILD POPULATIONS. Criti- cally endangered (A. C. Pooley, 1971, personal communication). This species is limited to large rivers, and is rarely abundant anywhere. Popu- lations are declining everywhere due to hide hunting and the spread of human population (Lowes, 1970). ORINOCO CROCODILE Crocodyhis intermedins Graves DISTRIBUTION. Northern South America — the Orinoco river drainage of Colombia (east of the Andes), Venezuela, and possibly Guyana (Medem, 1968; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is called caiman in South America. It is marketed under this name, or as Colombian, Venezuelan, or Venezuelan delta “alligator,” crocodile, large scale, or soft-belly. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 20 to 25. Tail whorls usually regular. Maximum length of live specimen is 23 feet. STATUS OF WILD POPULATIONS. En- dangered (Pan American Union, 1967; U.S. Department of Interior, 1970). Because of ex- cessive hide hunting the species is now rare in Venezuela, and apparently exterminated in Colombia (Honegger, 1968). South American countries prohibit the export of untanned hides. Colombia has legislation prohibiting the hunting of crocodiles, but it is not enforced ( Honegger, 1968). Import is prohibited under provision of the Endangered Species Conservation Act (U.S. Department of Interior, 1970). JOHNSON’S CROCODILE Crocodyhis johnsoni Krefft DISTRIBUTION. Northern Australia — from the Eitzroy River in northern Western Australia to Mackay in eastern Queensland (Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933; Worrell, 1963). OTHER COMMON NAMES. In Australia it is called the freshwater crocodile, Johnson's river crocodile, Johnstone’s crocodile, and fish crocodile. It may be marketed under these names, or as Australian or Singapore “alligator,” gator, crocodile, soft-belly, or large scale. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: Eol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 22 to 24. Tail whorls usually regular. Maximum length of live specimen is 9Vi feet. STATUS OF WILD POPULATIONS. Rare (Honegger, 1968). The species is completely protected by law in Western Australia and Northern Territories, but skins are still shipped from Queensland (Fauna Preservation Society, 1970b; Green, 1969; Honegger, 1968). MORELET’S CROCODILE Crocodyhis moreletii Dumeril, Bibron and Dumeril DISTRIBUTION. Northern Central America — Atlantic and Pacific coasts of Mexico, British Honduras, and Guatemala (Smith and Taylor, 24 New York Zoological Society: Zoologica, Summer, 1971 1950; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is some- times called Belize crocodile or Central Ameri- can crocodile. In Central America it is called "alligator,” caiman, and lagarto de El Peten. Hides are marketed under these names, or as Mexican “alligator,” crocodile, small scale, or soft-belly. DISTINGUISHING CHARACTERISTICS OE COMMERCIAT HIDES. Belly skins: Eol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 27 to 32. Tail whorls irregular (66 percent of the time). Maximum length is 8 feet. STATUS OE WILD POPULATIONS. En- dangered (Honegger, 1968; Pan American Union, 1967; U.S. Department of Interior, 1970). This species has all but been eliminated from British Honduras and parts of Guatemala (Charnock-Wilson, 1970). It is still locally abundant in parts of Mexico (Eauna Preserva- tion Society, 1 969b ) . Mexico has protective laws but they are unenforced (Honegger, 1968). Guatemala began enforcing its protective legis- lation in 1970. Importation is prohibited under provision of the Endangered Species Conserva- tion Act (U.S. Department of Interior, 1970). NILE CROCODILE Crocodyliis niloticiis Laurenti DISTRIBUTION. Africa (all of Africa ex- cept the northwest corner and central Sahara); east along the Mediterranean coast to Syria; Malagasy Republic (Madagascar); and Sey- chelles, Comoros, and Mauritius (Schmidt, 1919; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the Nilotic crocodile. Hides are marketed as African, Ethiopian, Kenya, Madagascan, or Nile “alligator,” “caiman,” crocodile, small scale, button-belly, or soft-belly. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. Usually no buttons, but occasionally single buttons may be present in the midbelly and collar area. Transverse ventral scale rows 26 to 32. Tail whorls usually regular. Hides from Nigeria have the tip of the tail missing due to local hunting practices. The tails are complete on hides from elsewhere. Maxi- mum length of live specimen is probably 18 feet. STATUS OF WILD POPULATIONS. En- dangered (Cott, 1961; Honegger, 1968; Pooley, 1969; U.S. Department of Interior, 1970). This species has been exterminated over large areas of Africa by hide hunters (Fauna Preservation Society, 1969c, 1969d, 1970a; Lowes, 1970; Pooley, 1970, in litt.). It can be found in num- bers only in small local populations. It is ex- tinct in the Seychelles and Mauritius. It is protected by law in most East African countries and in national parks and game preserves (Cott, 1969). Hunting of this species is to be regulated throughout all of Africa by the African Con- vention for the Conservation of Nature and Natural Resources (Burhenne, 1970; Honegger, 1968). South Africa has set up a research pro- gram in hopes of saving the species and restock- ing it in areas where it has been exterminated (Pooley, 1970). Importation is prohibited under provision of the Endangered Species Conserva- tion Act (U.S. Department of Interior, 1970). NEW GUINEA CROCODILE Crocodyliis novaeguineae novaeguineae Schmidt DISTRIBUTION. New Guinea (Schmidt, 1928a, 1932; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the New Guinea freshwater crocodile. Hides may be marketed as Australia, New Guinea, or Singapore “alligator,” crocodile, soft-belly, or large scale. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 24 to 25. Tail whorls usually regular. Maximum length of live specimen is 9Vi feet. STATUS OF WILD POPULATIONS. Rare (Honegger, 1968). Populations are declining rapidly due to hide hunting. Specimens over 20 inches in belly width are protected by laws in most of Papua and Northeast New Guinea (Bustard, 1970; Fauna Preservation Society, 1969a; Honegger, 1968). PHILIPPINE CROCODILE Crocodyliis novaeguineae niindorensis Schmidt DISTRIBUTION. Philippine Islands -Luzon, Mindoro, and Mindanao Islands (Schmidt, 1935; Wermuth, 1953; Wermuth and Mertens, 1961). OTHER COMMON NAMES. Also called the Mindoro crocodile and Philippine fresh- water crocodile. Hides may be marketed under the name Philippine or Singapore “alligator,” crocodile, soft-belly, or large scale. King and Brazailis: Species Identification of Commercial Crocodilian Skins 25 DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 24 to 26. Tail whorls usually regular. Maximum length of live specimen is 8 feet. STATUS OF WILD POPUL.ATIONS. Rare, possibly endangered. Hide hunting is eliminat- ing the species from parts of its former range. MUGGER CROCODILE Crocodylns paliistris palitstris Lesson DISTRIBUTION. India and Pakistan - from the Dasht River in West Pakistan through all the river systems of India to the Brahmaputra River drainage in the east ( De Rooij, 1915; Schmidt, 1935; Smith, 1931; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. Also called the marsh crocodile, broad-snouted crocodile, swamp crocodile, and Indian freshwater croco- dile. Hides may be marketed as Indian “alli- gator,” crocodile, soft-belly, or small scale. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 26 to 32. Ventral collar not distinct (no enlarged scales). Tail whorls usually regular. Maximum length of live specimen is 1 3 feet. STATUS OF WILD POPULATIONS. En- dangered. The species is protected in India by a ban on the export of crocodile hides, and in Pakistan by a ban on the export of all wild animal hides (Fauna Preservation .Society, 1967, 1970c; Mountfort, 1969). CEYLON MUGGER CROCODILE Crocodylns palnstris kiinhula Deraniyagala DISTRIBUTION. Ceylon (Deraniyagala, 1936, 1939, 1953; Wermuth, 1953; Wermuth and Mertens, 1961 ) . OTHER COMMON NAMES. It is also called the Ceylon swamp crocodile, Ceylon marsh crocodile, and lake crocodile. In Ceylon it is known as hale kimbula, ala kimbiila, and kiilathi innihele. It may be marketed as Ceylon “alligator,” crocodile, soft-belly, or small scale. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 26 to 32. Ventral collar present and distinct. Tail whorls regular. Maximum length of live specimen is 18 feet. .STATUS OF WILD POPULATIONS. De- clining in numbers. Hunting is regulated by the Ceylon government (Fauna Preservation So- ciety, 1970e). SALTWATER CROCODILE Crocodylns porosns Schneider DISTRIBUTION. India and Ceylon east to Australia and New Guinea — the coastal rivers, lagoons, and marshes from Cochin in extreme southwestern India east to Ceylon, Burma, Malaysia, Thailand, Cambodia, Vietnam, Indo- nesia, the Philippines, Palau Islands, northern Australia, New Guinea, Solomon Islands, New Hebrides, and Fiji (Deraniyagala, 1939, 1953; De Rooij, 1915; Schmidt, 1932; Taylor, 1970; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933; Worrell, 1963). OTHER COMMON NAMES. It is also called the estuarine crocodile, gator (in Austra- lia), and sea-going crocodile. In Ceylon it is known as pita gatteya, gatte kiinbnla, gorekeya, and senunnklian: in Indonesia, bnaja; in Malay- sia, bnaja, bnaya, baya, and rawing. Hides may he marketed under these names, or as Indian, Javan, Philippine, Singapore, Sumatran, or Thailand “alligator,” crocodile, soft-belly, or small scale. DISTINGUISHING CHARACTERLSTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral .scale rows 30 to 35. Tail whorls regular. Maximum length of live speci- men is probably 25 feet. STATUS OF WILD POPULATIONS. Most populations are declining rapidly due to hide hunting, and the species is non-existent in some parts of its former range where it was once abundant (Fauna Preservation .Society, 1970d; Honegger, 1968). It is partially protected in most of Papua and North East New Guinea, where specimens over 20 inches belly width may not be killed (Fauna Preservation Society, 1969a; Bustard, 1970). The species is com- pletely protected in Western Australia until 1980 (Fauna Preservation Society, 1970b; Honegger, 1968). Indonesia has imposed size limits. Ceylon, India, and Pakistan protect the species completely by banning the export of all crocodile skins or the skins of all wild animals (Fauna Preservation Society, 1967; Honegger, 1968; Mountfort, 1969). Singapore requires export licenses. Deraniyagala ( 1939, 1953) mis- takenly listed this species as occurring on the .Seychelles and Mauritius where Crocodylns niloticns was known to occur in the past. 26 New York Zoological Society: Zoologica, Summer, 1971 CUBAN CROCODILE Crocodyliis rhombifer Cuvier DISTRIBUTION. Cuba and the Isle of Pines (Barbour and Ramsden, 1919; Varona, 1966; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In Cuba it is called cocodrilo, cocodrilo perla, cocodrilo cri- ollo, cocodrilo legitimo, caiman, and occasion- ally zaquendo. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 32 to 33. Tail whorls regular. Maximum length of live speci- men is 16 feet. STATUS OF WILD POPULATIONS. En- dangered (Honegger, 1968; U.S. Department of Interior, 1970). The species once occurred on the Isle of Pines from which it has been exter- minated. Today it only occurs in remnants of the Zapata Swamp on the south coast of Cuba, but hide hunting and land drainage has made it very nearly extinct even there. The Cuban gov- ernment protects this species rigidly and has established a captive breeding facility in the Zapata Peninsula National Park in an attempt to save it from extinction (Honegger, 1968). Importation is prohibited under provision of the Endangered Species Conservation Act (U.S. Department of Interior, 1970). SIAMESE CROCODILE Crocodyliis siainensis Schneider DISTRIBUTION. Southeast Asia— Thailand, Cambodia, Vietnam, and Java (De Rooij, 1915; Smith, 1931; Taylor, 1970; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It may also be called the Siamese freshwater crocodile. In Indonesia it is called hiiaja. Hides may be sold as Java, Singapore, or Thailand “alligator,” crocodile, soft-belly or small scale. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 30 to 34. Tail whorls regular. Maximum length of live speci- men is 13 feet. STATUS OF WILD POPULATIONS. En- dangered. It has always been a rare animal in Indonesia, and became scarce in Thailand 30 years ago due to hide hunting. Today fewer than 200 remain in the wild in Thailand, but approxi- mately 9,000 specimens are protected in the Sumatprakan Crocodile Farm in Bangkok (U. Youngparpakorn, 1971, personal communica- tion). WEST AFRICAN DWARF CROCODILE Osteolaennis tetraspis tetraspis Cope DISTRIBUTION. West Africa — the Niger and Senegal river drainages and other rivers south of the Sahara and north of the Congo River drainage (Schmidt, 1919; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the broad-snouted crocodile. Hides may be mar- keted as African "caiman,” button-belly, bony crocodile, black crocodile, or rough-back croco- dile. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. Large single osteoderm but- tons present. Surface pitting usually evident. Transverse ventral scale rows 21 to 27. Maxi- mum length of live specimen is 6V2 feet. STATUS OF WILD POPULATIONS. En- dangered (A. C. Pooley, 1971, personal com- munication). Populations declining due to hide hunting, destruction of habitat, and live animal collecting (Lowes, 1970). This species has never been as abundant as the other African species. CONGO DWARF CROCODILE Osteolaennis tetraspis oshorni (Schmidt) DISTRIBUTION. Central Africa-the Congo River drainage (Schmidt, 1919; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. Also called the Central African dwarf crocodile, Osborn’s dwarf crocodile, and African broad-snouted crocodile. Hides may be marketed as African “caiman,” button-belly, bony crocodile, black crocodile, or rough-back crocodile. DISTINGUISHING CHARACTERISTICS OF COMMERCIAL HIDES. Belly skins: Fol- licle glands present. Large single osteoderm but- tons present. Surface pitting usually evident. Transverse ventral scale rows 21 to 27. Maxi- mum length of live specimen is 5 feet. STATUS OF WILD POPULATIONS. En- dangered (A. C. Pooley, 1971, personal com- munication). This species does not occur in large populations. Its numbers are declining due to hide hunting. FALSE GAVIAL Tomistoma schlegelii (Muller) DISTRIBUTION. Southeast Asia - Indone- sia (Kalimantan and Sumatra) and Malaysia King and Brazaitis: Species Identification of Commercial Crocodilian Skins 27 ( De Rooij, 1915; Taylor, 1970; Wenmith, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. It is also called the Malay gavial, Malayan gharial, and Malayan fish crocodile. In Indonesia it is called hediai sanipit and bitaja supit; in Malaya, hiiaya senjii- long; in Sarawak, haya kaniilong. Hides may be sold under these names, or as Singapore “alli- gator,” crocodile, soft-belly, or large scale. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: Eol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 22 to 24. Tail whorls usually regular. Maximum length of live specimen is 16 feet. STATUS OE WILD POPULATIONS. De- clining in numbers, soon to be endangered. Hide hunters have so decimated the populations of this animal in Malaysia that protective legisla- tion is being considered (Lucas Chin. 1970, per- sonal communication). Family GAVIALIDAE GAVIAL Gavialis gangeticiis (Gmelin) DISTRIBUTION. India, Pakistan, and Burma — specifically the Indus, Mahandi, Gan- ges, Brahmaputra, and Kaladan river drainage systems, and possibly parts of the Irawaddy system in northwestern Burma (Smith, 1931; Wermuth, 1953; Wermuth and Mertens, 1961; Werner, 1933). OTHER COMMON NAMES. In India it is called gharial. Hides may be sold as Indian soft- belly, small scale, “alligator,” “crocodile,” or gavial. DISTINGUISHING CHARACTERISTICS OE COMMERCIAL HIDES. Belly skins: Fol- licle glands present. No osteoderm buttons. Transverse ventral scale rows 30 to 31. Ventral collar not prominent. Maximum length of live specimen is 21 Vi feet. STATUS OF WILD POPULATIONS. En- dangered (U.S. Department of Interior, 1970). Protected in India by a ban on the export of all crocodilian hides, and in Pakistan by a ban on the export of all wild animal hides (Fauna Preservation Society, 1967, 1970c; Mountfort, 1969). Importation is prohibited under provi- sions of the Endangered Species Conservation Act (U.S. Department of Interior, 1970). Acknowledgments We wish to thank David Klapisch (Southern Trading Corporation, Newark, New Jersey), Hyman Marx (Field Museum of Natural His- tory, Chicago), Henry Molt (Philadelphia Rep- tile Exchange), Charles Myers (American Mu- seum of Natural History, New York), Ray Pawley (Chicago Zoological Park), Barry Wake- man (Cincinnati Zoological Society), George Zug (U.S. National Museum, Washington, D.C.) for the opportunity of examining specimens and skins in their care. We also wish to thank Edward Baker and Warren Difendall (United States Department of Interior, Department of Sport Fish and Wildlife, New York), Claire Hagen (Hagen and Company, New York), and Glenn Kindler (Connecticut Import Export Cor- poration, New York) for their assistance. Rene Honegger kindly lent us an advance copy of the International Union for the Conservation of Nature and Natural Resources Red Data Book on amphibians and reptiles so that we might include some of the information here. Jose Cei, Lucas Chin, Federico Medem, A. C. Pooley, James Powell, and others provided personal observations on wild populations of crocodilians. William Meng (New York Zoological Society) provided photos. Li i erature Cited Barbour, T., and C. T. Ramsden 1919. The herpetology of Cuba. Mem. Mas. Comp. Zook, 47(2): 73-2 1 3. Blirhenne, W. E. 1970. The African Convention for the Conser- vation of NatLire and NatLiral Resources. Biol. Conservation, 2(2): 105-114. Bustard, H. R. 1970. A futLire for crocodiles. Oryx, 10(4): 249- 255. Carvaliio, A. L. DE 1955. Os jacares do Brasil. Arquivos Mus. Nac., 42( 1 ): 127-139. Charnock- Wilson, J. 1970. Manatees and crocodiles. Oryx, 10(4): 236-238. Cochran, D. M. 1941. The herpetology of Hispaniola. U.S. Nat. Mus. Bull. 177: 1-398. Cott, H. B. 1961. Scientific results of an inquiry into the ecology and economic status of the Nile crocodile (CrocodHus niloticiis ) in Uganda and Northern Rhodesia. Trans. Zook Soc. London, 29(4 ) : 21 1-356. 28 New York Zoological Society: Zoologica, Summer, 1971 1969. Tourists and crocodiles in Uganda. Oryx, 10(3): 153-160. Crowe, P. K. 1965. What is happening to the wildlife of South America. Oryx, 8(1): 28-37. Deraniyagala, P. E. P. 1936. A new crocodile from Ceylon. Ceylon J. Sci. (B) 19: 279-286. 1939. The tetrapod reptiles of Ceylon. Vol. I, Testudinates and Crocodilians. Colombo Mils. Nat. Hist. Ser. xxxii 412 pp. 1953. A colored atlas of some vertebrates from Ceylon. Vol. 2, Tetrapod Reptilia. Ceylon Nat. Mus. Publ. vii + 101 pp. De Roou, N. 1915. The reptiles of the Indo- Australian archi- pelago. Vol. I, Lacertilia, Chelonia, Emy- dosauria. E. .1. Brill Ltd., Leiden, xiv + 384 pp. Fauna Preservation Society 1967. No crocodiles from India. Oryx, 9(3): 184. 1969a. Protection for Papua’s crocodiles. Oryx, 10(2): 81. 1969b. Morelet’s crocodile near extinction. Oryx, 10(2): 88. 1969c. Crocodiles decrease in Natal. Oryx, 10(3): 144. 1 969d. Crocodiles in Blue Nile gorge. Oryx, 10(3): 144. 1970a. Crocodiles going in Uganda. Oryx, 10(4): 208-209. 1970b. Crocodiles in Australia. Oryx, 10(4) : 213. 1970c. Two crocodiles in trouble. Oryx, 10(5): 287. 1 970d. Sarawak crocodiles. Oryx, 10(5): 295. 1970e. Poaching in Ceylon. Oryx, 10(5): 295. Green, K. 1969. Rare fauna. Walkabout, 35(4): 6. Honegger, R, 1968, Red Data Book. Volume 3 — Amphibia and Reptilia. International Union for the Conservation of Nature and Natural Re- sources, Survival Service Commission, Morges, Switzerland. Loose leaf n.p. Lowes, R. H. G. 1970. Destruction in Sierra Leone. Oryx, 10(5): 309-310. Medem, F. 1955. A new subspecies of Caiman sclerops from Colombia. Fieldiana: Zook, 37: 339-344. 1963. Osteologia craneal, distribucion geografica y ecologia de Melanositchiis iiiger (Spix) (Crocodylia, Alligatoridae). Revista Acad. Colombiana Cien. Exactas, Fis. y Nat., 12(45): 5-19. 1967. El genero Paleosiichits en Amazonia. Simposio Biota Amazonica, 3: 141-162. 1968. El desarrollo de la herpetologia en Co- lombia. Revista Acad. Colombiana Cien. Exactas, Fis. y Nat., 13(50): 149-199. Mountfort, G. 1969. Pakistan’s progress. Oryx, 10(1): 39-43. Pan American Union 1967. Listas de especies de fauna y flora en vias de extincion en los estados miembros. La Convencion para la Proteccion de la Flora, de las Fauna y de las Bellezas Esci- encas Naturales deJos Estados America- nos. Organization of American States, ii + 48 pp. POOLEY, A. C. 1969. Rearing crocodiles in Zululand. African Wildlife, 23(4): 314-320. 1970. Crocodile rearing in Zululand. Animals, 13(2): 76-79. Pope, C. 1935. The reptiles of China. Nat. Hist. Central Asia. Vol. 10. Amer. Mus. Nat. Hist, lii + 604 pp. Schmidt, K, P. 1919, Contributions to the herpetology of the Belgian Congo based on the collection of the American Congo Expedition, 1909- 1915. Bull. Amer. Mus. Nat. Hist., 39(2): 385-624. 1928a. A new crocodile from New Guinea. Field Mus. Nat. Hist. Zook Ser., Publ. 247, 12(14): 177-181. 1928b. Notes on South American caimans. Field Mus. Nat. Hist. Zook Ser., Publ. 252, 12( 17) : 205-231. 1932. Notes on New Guinea crocodiles. Field Mus. Nat. Hist. Zook Ser., Publ. 310, 18(8): 167-172. 1935. A new crocodile from the Philippine Islands. Field Mus. Nat. Hist. Zook Ser., 20(8): 67-70. 1953. A checklist of North American amphibi- ans and reptiles. 6th ed. Univ. Chicago Press, vii + 280 pp. Smtih, M. 1931. The fauna of British India. Reptilia and Amphibia. Vol. 1, Loricata, Testudines. Taylor and Francis, London, xxviii + 185 pp. King ami Brazaitis: Species Identification of Commercial Crocodilian Skins 29 Smith, H. M., and E. H. Taylor 1950. An annotated checklist and key to the reptiles of Mexico exclusive of the snakes. U.S. Nat. Mus. Bull. 199: 1-253. Taylor, E. H. 1970. The turtles and crocodiles of Thailand and adjacent waters. Univ. Kansas Sci. Bull., 49(3): 87-179. U.S. Department of Interior 1968. American alligator. Rare and endangered fish and wildlife of the United States. U.S. Dept. Interior, Bur. Sport Eish. Wildlife, Washington, D.C., Sheet RA-2. 1970. List of endangered foreign fish and wild- life. Title 50— Wildlife and Eisheries. Eed- eral Register, 35(233 ): 18319-19322. Varona, L. S. 1966. Notas sobre los crocodilidos de Cuba y descripcion de una nueva especie del Pleis- tocene. Poeyana Inst. Biol., Ser. A, No. 16: 1-34. Wermuth, H. 1953. Systematik der Rezenten Krokodile. Mit- teil. Zool. Mus. Berlin, 29(2): 375-514. Wermuth, H., and R. Mertens 1961. Schildkroten, Krokodile, Bruckenechsen. Veb Gustav Eischer Verlag, Jena, xxvi + 422 pp. Werner, E. 1933. Reptilia, Loricata. Das Tierreich. Gruyter and Co., Berlin, xiii + 40 pp. Worrell, E. 1963. Reptiles of Australia. Angus and Robert- son, Sydney, xv + 207 pp. 30 New York Zoological Society: Zoologica, Summer, 1971 Figure I. Diagrammatic dorsal (A) and ventral (B) views of a crocodilian. Hornback hides consist of most of the skin seen in A (skull and feet are absent). Belly hides consist of most of the skin seen in B (skull and feet are absent and the lateral [side] skin is attached). Transverse scale rows are counted by beginning and ending with the rows indicated by the arrows. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 31 Figure 2. Comparative flexibility of a button-belly (A) and a soft-belly (B) hide. The hard osteoderms in the scales permit the button-belly hide to flex only between the scales, while the soft-belly hide will also flex through the scales. 32 New York Zoological Society: Zoologica, Summer, 1971 Figure 3. Outside surface of a finished American alligator (Alligator mississippiensis) belly hide. Closer views of the ventral scales and spider-web umbilicus are provided in figures 5 and 7. King and Brazaitis: Species Identification of Commercial Crocodilian Skins Figure 4. Diagrammatic illustration of the American alligator ventral scales shown in figure 5 lack of both surface pitting and follicle glands. 34 New York Zoological Society: Zoologica, Summer, 1971 Figure 5. Ventral scales of a finished American alligator (Alligator mississippiensis) belly hide. Compare it with figure 4. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 35 Figure 6. Inside surface of a finished adult American alligator {Alligator mississippiensis) belly hide. Note the total absence of osteoderm buttons, which indicates the specimen probably came from Louisiana. Compare the inside of the ventral collar, just visible at the top of the photograph, with figure 8. 36 New York Zoological Society: Zoologica, Summer, 1971 Figure 7. The spider-web umbilicus typical of American alligator {Alligator mississippiensis) belly hides — A is a diagrammatic illustration of the photograph B. Also note the absence of both surface pitting and follicle glands on the ventral scales. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 37 Figure 8. Inside surface of a finished American alligator (Alligator mississippiensis) hide from Florida. The portion shown is from the throat area as evidenced by the ventral collar. The dark round blotches in the center of the scales are single osteoderm buttons. 38 New York Zoological Society: Zoologica, Summer, 1971 Figure 9. Hornback (A) and belly hides (B and C) of a South American caiman (Caiman crocodiliis) . Note the presence of the vent in both belly hides. A and C are crusts. B is a hide with sanvage finish. King iiml Brazaitis: Species Identification of Connnerci(d Crocodilian Skins 39 Figure 10. Outside surface of a crust belly hide of a South American caiman (Caiman crocodilas ) . Note the surface pitting which is indicative of underlying osteoderm buttons. 40 New York Zoological Society: Zoologica, Summer, 1971 Figure 11. Ventral scales of a South American caiman (Caiman crocodiliis) crust. Note the surface pitting. Lateral scales are just visible on the right side of the photograph. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 41 Figure 12. Ventral scales of a finished South American caiman {Caiman crocodiliis) belly hide. Photo- graph B is a close view of the scales seen in A. Because of the technique used to dye this hide, the surface pits are white against a dark background. Note that the pitting is not as pronounced near the vent (lower half of A ) as near midbelly ( upper half of A ) . 42 New York Zoological Society: Zoologica, Summer, 1971 Figure 13. Inside surface of a South American caiman (Caiman crocodihts) crust. Note the presence of double osteoderm buttons in the ventral scales. Closer views are provided in figure 14. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 43 Figure 14. Double osteoderm buttons on the inside surface of a South American caiman {Caiman crocodilns) belly crust. A is a diagrammatic illustration of the photograph B. Each ventral scale contains two osteoderms, double buttons. The larger posterior button is shaded in A, while the smaller inward- curving anterior button is unshaded. Most of the anterior button is removed when the hide is shaved. Compare this figure with the shaved hide in figure 19. 44 New York Zoological Society: Zoologica, Summer, 1971 Figure 15. Sides of South American caiman (Caiman crocodilus) . The center hide is a crust. The other two are finished hides. The arrow indicates the anterior (cephalic) end of the hide. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 45 Figure 16. Scales of finished South American caiman (Caiman crocodilns) sides. A is a diagrammatic illustration of photograph B. Note that the rows of large oval scales alternate with strips of soft skin with a network of creases. Compare this with figure 21. 46 New York Zoological Society: Zoologica, Summer, 1971 Figure 17. Outside surface of a finished black caiman (Mehmosiiclins niger) belly hide. A closer view of the ventral scales is provided in figure 18. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 47 Figure 18. Ventral scales of a finished black caiman (Melanosnchiis niger) belly hide. A is a diagram- matic illustration of the photograph B. Note the wrinkles and fine surface pitting, as well as the lighter color in the centers of the scales. Both conditions are indicative of underlying osteoderm buttons. 48 New York Zoological Society: Zoologica, Summer, 1971 Figure 19. Inside surface of a finished black caiman (Melanosuchus niger) belly hide. Note the dark double osteoderm buttons in each scale. Photograph B is a close view of the buttons seen in A. This hide has been shaved so most of the anterior button has been removed. Compare it with figures 13 and 14. 49 King and Brazaitis: Species Identification of Commercial Crocodilian Skins Figure 20. Outside surface of finished black caiman { Melanosiicliiis niger) throat (A), girdle (B), and side (C). The scales of the side are shown in figure 21. 50 New York Zoological Society: Zoologica, Summer, 1971 Figure 21. Scales of finished black caiman {Melanosiicliiis iiiger) side. A is a diagrammatic illustration of photograph B. Note that the large oval scales alternate with rows of small scales. Compare this with figure 16. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 51 Figure 22. Diagrammatic illustration of the Nile crocodile belly hide shown in figure 23. Note the presence of follicle glands (only those visible in figure 23 are illustrated). 52 New York Zoological Society: Zoologica, Summer, 1971 Figure 23. Outside surface of a finished Nile crocodile (Crocodyliis uiloticus) belly hide. Compare it with figure 22. 53 King and Brazaitis: Species Identification of Coniniercial Crocodilian Skins Figure 24. Diagrammatic illustration of the ventral scales of the Morelet’s crocodile hide shown in figure 25. Note the prominent follicle glands. 54 New York Zoological Society: Zoologica, Summer, 1971 Figure 25. Ventral scales of a finished Morelet’s crocodile [Crocodylus moreletii) belly hide. Compare it with figure 24. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 55 Figure 26. Comparison of scale size on (A) large scale false gavial (Tomistoma schlegelii) and (B) small scale saltwater crocodile (Crocodylus porosits) belly hides. 56 New York Zoological Society: Zoologica, Summer, 1971 Figure 27. Inside surface of a finished saltwater crocodile (Crocodytus porosits) belly hide. Note the total absence of osteoderm buttons. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 57 Figure 28. Diagrammatic illustration of the Nile crocodile tail whorls shown in figure 29. Note both the presence of follicle glands (only the ones visible in figure 29 are illustrated ) and the regular arrangement of the whorls. Compare it with figure 30. 58 New York Zoological Society: Zoologica, Summer, 1971 Figure 29. Tail whorls of a finished Nile crocodile (Crocodylus niloticus) belly hide. Compare it with figure 28. King ami Brazaitis: Species Identification of Commercial Crocodilian Skins 59 Figure 30. Diagrammatic illustration of the Morelet’s crocodile tail whorls shown in figure 3 1, Note the presence of both follicle glands and irregular and incomplete (shaded) whorls. Compare it with figure 28. 60 New York Zoological Society: Zooiogica, Summer, 1971 Figure 31. Tail whorls of a finished Morelet’s crocodile [Crococlylus moreietii) belly hide. Compare it with figure 30. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 61 Figure 32. Outside surface of a finished Orinoco crocodile (Crocodylns intermedins) hide. The arrows indicate the location of parasitic “worm trails.” Close views of these trails are shown in figure 33. 62 New York Zoological Society: Zooiogica. Summer, 1971 Figure 33. Undulating “worm trails" on the ventral scales of an Orinoco crocodile (Crocodyliis inter- mciliiis) belly hide. Similar trails have been seen on Johnson’s crocodiles (C. joimsoni), Morelet’s crocodiles (C. moreletii) , Nile crocodiles (C. uiloliciis). and saltwater crocodiles iC. porosns). They probably occur on other species as well. King amt Brazaitis: Species Identification of Conunercial Crocodilian Skins 63 Figure 34. Outside (A) and inside (B) surfaces of finished ventral scales of either the African slender- snouted crocodile (Crocodylits catapliractns) , Nile crocodile (Crocodylus niloticiis) , or dwarf crocodile (Osteolaemns tetraspis). Note the lighter color in the center of the scales (A), which are indicative of the underlying dark single osteoderm buttons (B). 64 New York Zoological Society: Zoologica, Summer, 1971 Figure 35. Outside surface of a finished belly skin ( A ) and ventral scales (B ) of either an African slender- snouted crocodile {Crococlylus cataphractus), Nile crocodile (Crocodyliis niloticus) or dwarf crocodile lOsleolaemiis tetruspis). Note the shallow indentations, surface pits, indicative of underlying single buttons. Also note that follicle glands are reduced to deep wrinkles by polishing process. King and Brazaitis: Species Identification of Commercial Crocodilian Skins 65 Figure 36. Outside (A) and inside (B) surfaces of finished African slender-snouted crocodile (Crocodyliis cataphractus), Nile crocodile (Crocodyliis niloticiis), or dwarf crocodile (Osteolaemiis tetraspis) belly hides. Note the surface pitting (A) and dark single osteoderm buttons (B). Close views of the ventral scales are shown in figure 37. 66 New York Zoological Society: Zoologica, Slimmer, 1971 Figure 37. Ventral scales (A) and single osteoderm buttons (B) of finished African slender-snouted crocodile (Crocotlyliis cataphractiis), Nile crocodile (Crocodyiiis iiiloticiis), or dwarf crocodile (Osteolaemiis tciruspis} hides. Because of the technique used to dye this hide, the surface pitting is white against the dark scales. King unci Brazaitis: Species Identification of Conunercial Crocodilian Skins 67 Figure 38. Lady’s purse made from narrow South American caiman (Caiman crocodUns) sides. Seams where the sides are glued together are difficult to locate. Arrows indicate seams. 68 New York Zoological Society: Zoologica, Summer, 1971 Figure 39. Lady’s purse made from a South American caiman {Caiman crococliliis) belly. Note the wrinkles and surface pitting. Arrows indicate the high points of the scales. In this species the high point is just anterior to the center of the scale (the anterior end of this hide is touching the table, the posterior end is up ) . King nnd Brazuitis: Species Identification of Commercial Crocodilian Skins 69 Figure 40. Lady’s purse made from American alligator (Alligator mississippiensis) belly. Note the absence of both surface pitting and follicle glands. Also note the spider-web umbilicus indicated by the arrows. 70 New York Zoological Society: Zoologica, Summer, 1971 Figure 41. Man’s belt made from either African slender-snouted crocodile (Crocodylus cataphractiis), Nile crocodile (Crocodylus niloticus). or dwarf crocodile (Osteolaemus tetraspis) belly hide. Photograph B is a close view of some scales from A. Note follicle glands and also slight hump (arrows) indicating under- lying single osteoderm buttons. News and Notes: Crocodylus intermedins 71 NEWS AND NOTES Crocodylus intermedins Graves, A Review of the Recent Literature (Figures 1-3) Studies evaluating the definitive morphological characters of living crocodilians have disclosed some confusion in the recent literature on the Orinoco crocodile Crocodylus intermedins Graves ( Mook, 1921, Bull. Amer. Mus. Nat. Hist., 44(13): 165-173; Wermuth, 1953, Son- derdruck aus: Mitteilungen aus dem Zoologis- chen Museum in Berlin, 29(2) :493-495; Wer- muth and Mertens, 1961 Schildkroten, Kroko- dile, Bruckenechsen, Veb Gustav Fischer, Jena: 359 and 361). A complete biological profile of the species is given by Medem (1958, Caldasia, 8(37 ) : 175-215 ) and need not be repeated here. I thank Dr. F. Wayne King and the New York Zoological Society (= NYZS); Federico Me- dem; the American Museum of Natural History, New York ( = AMNH), and its Herpetological Information Search System; and the Field Mu- seum of Natural History, Chicago ( = FMNH) for their assistance in making specimens and difficult-to-obtain literature available for study and for reviewing the manuscript. Discussion Before Medem presented his collection to the Field Museum of Natural History in 1958, Crocodylus intermedins was poorly represented in zoological and museum collections. Those specimens which were available were supported by little or no collecting data. Consequently, one skull (AMNH 8790), bearing the data “Vene- zuela, South America, via the New York Zoolog- ical Society,” was described in detail and figured by Mook (1921), and subsequently figured by Wermuth (1953), and Wermuth and Mertens (1961). Unfortunately, the skull was not avail- able for re-examination until recently. Compari- son of AMNH 8790 to a female Crocodylus intermedins collected by Medem on the Rio Ariari, Territory of Meta, Colombia (FMNH 75658); and individuals of Crocodylus cata- phractus from K. P. Schmidt’s Congo Expedition (AMNH 10075), and from Liberia, West Africa (NYZS 610716 and 61 0504) discloses AMNH 8790 to be an example of Crocodylus cataphractus, the West African slender-snouted crocodile, erroneously identified as Crocodylus intermediusd’ - Medem (1958:184) pointed out that Mook described and figured AMNH 8790 with nasal bones not entering the external narial opening while those Crocodylus intermedins he had ex- amined from Colombia showed the nasals to enter the external narial opening. However, he did not realize Mook had incorrectly identified the specimen as C. intermedins. In addition, AMNH 8790 differs from Croco- dylus intermedins (FMNH 75658) and agrees with Crocodylus cataphractus (AMNH 10075), in the following aspects: The pre-maxillary/maxillary suture extends caudad to slightly beyond the level of the first maxillary teeth in AMNH 8790, while in FMNH 75658 the suture nearly reaches the level of the third maxillary teeth. The mandibular symphysis of AMNH 8790 and AMNH 10075 extends to the level of the eighth mandibular teeth, while in FMNH 75658 the symphysis barely reaches the level of the seventh mandibular teeth. The palatine/maxillary suture in AMNH 8790 is triangular, anteriorly pointed at its junc- tion with the median palatine suture, and occu- pies a space approximately equal to that of three adjacent maxillary teeth. FMNH 75658 has an elongated parallel-sided palatine/maxillary su- ture, square at its anterior face which is at right angles to the median palatine suture. Its length coincides to the space occupied by four maxil- lary teeth. The ninth maxillary teeth are largest in AMNH 8790 while the tenth are the largest in FMNH 75658. 1 While comparing plate figures, it was noted that the skull figure for Tomistoma schlegelii (S. Muller) shown in Wermuth and Mertens, 1961, page 376, was dupli- cated in error on page 360 as the skull figure for Croco- dylus cataphractus Cuvier. - De Rochebrune, 1883 (Faune de la Senegambie, J. Durand, Imprimeur de la Societe Linneenne, Bordeau, p. 47), includes Temsacus intermedins Gray (= Croco- dylus intermedins Graves) in the fauna of Senegambi (= Senegal and Gambia) although the species is un- known in Africa. The specimen he figures most closely resembles C. intermedins. 72 New York Zoological Society: Zoologica, Summer, 1971 The conformation of AMNH 8790 is sug- gestive of C. cataphractiis in the relatively high, square profile of the cranial table, the concave dorsal aspect of the snout, and the proportion- ately narrow frontal between the orbits. FMNH 75658 differs in having a relatively low cranial table, a slightly elevated or “swollen” snout im- mediately anterior to the orbits, and a frontal region which is wide in proportion to the over- all length of the skull. It should be noted that AMNH 8790 is the skull of a deformed specimen, probably result- ing from confined captive conditions over a prolonged period of time during shipment. Many of the maxillary and mandibular teeth are broken or twisted in their sockets. The mandible itself is broken, perhaps during preparation or damaged in life. Portions of the anterior mandi- ble and the pre-maxillaries are also damaged or worn away, a condition often seen in captive specimens poorly crated for shipment, in cramped quarters. Only two “Orinoco crocodiles” appear in the New York Zoological Society’s annual reports between the years 1900 and 1922. These coin- cide to the receipt of AMNH 8790 and another preserved juvenile specimen (AMNH 2206) bearing “Colombia, South America,” data, also “via the New York Zoological Society.” The latter preserved specimen is also an example of Crocodylus cataphractiis. One of these is re- ported to have been secured by the zoological park from a donor recently returned from a tour aboard a merchant vessel. One of these specimens was photographed in life while at the zoological park. The plates, mis- identified as Crocodylus intermedins, were re- produced in subsequent literature (Ditmars, 1913, Bull. Zool. Soc., 16(58): 1005; DeSola, 1933, Bull. Zool. Soc., 36(1): 14, Wermuth, 1953, 29(2) :493). These photographs are pre- served in the NYZS photographic archives. The identification of living crocodilians with- out the availability of accurate collecting data has been a problem for scientific staffs of zoolog- ical parks and museums, particularly during earlier years when the classic works of Boulen- ger, Cuvier, and Gray represented the only com- prehensive literature on crocodilians. These pub- lications, which stress osteological materials rather than living specimens, obviously were of little help in the identification of a rare species perhaps never seen before and seldom encoun- tered since. Peter Brazaitis, Department of Herpetol- ogy, New York Zoological Park, Bronx, New York 10460. News and Notes: Crocodylus intermedins 73 Figure 1. Crocodylus cataphractus Cuvier (AMNH 8790), misidentified and described in Mook (1921) as Crocodylus intermedins Graves. Figure adapted from Mook. 74 A^ch' York Zoological Society: Zoologica, Summer, 1971 Figure 2. Crocodylus cataphractus Cuvier (AMNH 10075), described in Mook (1921). Figure adapted from Mook. News and Notes: Crocodyliis intermedins 75 Figure 3. Crocodyliis intermedins Graves (FMNH 75658), a juvenile female from Rio Ariari, Territory of Meta, Colombia, collected by Federico Medem, Illustration by Lloyd Sandford, NYZS. ! ’s ■ I i: V- ■ V I %■ NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable), and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which much all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 2. Species Identification of Commercial Crocodilian Skins. By F. Wayne King AND Peter Brazaitis. Figures 1-41 15 News and Notes Crocodylus intermedins Graves, A Review of the Recent Literature. By Peter Brazaitis. Figures 1-3 Published August 20, 1971 ® 1971 New York Zoological Society. All rights reserved. S‘iO.,1) 7 3 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 56 • ISSUE 3 • FALL, 1971 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Edward R. Ricciuti, Editor & Curator, Publications & Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor EDITORIAL COMMITTEE Robert G. Goelet, Chairman; William G. Conway, Donald R. Griffin, Hugh B. House, F. Wayne King, Peter R. Marler, Ross F. Nigrelli, James A. Oliver, Edward R. Ricciuti, George D. Ruggieri, S.J. William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology ;Vi\xg\\ B. House, Curator, Mammalogy: James G. Doherty, Assistant Curator, Mammalogy: F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Robert A. Brown, Assistant Curator, Animal Departments; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Ctirator Emerittis AQUARIUM James A. Oliver, Director; H. Douglas Kemper, Assistant Ctirator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology; Jay Hyman, Consultant Veterinarian OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director & Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempier, Jf., Assistant to the Director & Bio-Organic Chemist; Eli D. Goldsmith, Scientific Consultant; William Antopol, Research Associate, Comparative Pathology; C. M. Breder, Jr., Research Associate, Ichthyology, Jack T. Cecil, Virologist; Harry A. Charipper, Research Associate, Histology; Paul J. Cheung, Microbiologist; Erwin J. Ernst, Research Associate, Estaurine & Coastal Ecology; Kenneth Gold, Marine Ecologist', Jay Hyman, Research Associate, Comparative Pathology; Myron Jacobs, Neuroanatomist; Klaus Kallman, Fish Geneticist; John J. A. McLaughlin, Research Associate, Planktonology; Martin F. Schreibman, Research Associate, Fish Endocrinology INSTITUTE FOR RESEARCH IN ANIMAL BEHAVIOR (Operated jointly by the Society and the Rockefeller University) Peter R. Marler, Director & Senior Research Zoologist; Paul Mundinger, Assistant Director & Research Associate; Donald R. Griffin, Senior Research Zoologist; Jocelyn Crane, Senior Research Zoologist; Roger S. Payne, Research Zoologist; Fernando Nottebohm, Research Zoologist; George Schaller, Research Zoologist; Thomas T. Struhsaker, Research Zoologist; Alan Till, Research A ssociate ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Ben Sheffy, Consultant, Nutrition; Roy Bellhorn, Consultant & Research Associate, Comparative Ophthalmology; John Budinger, Joseph Conetta, Edward Garner, Henry Kobrin, Ralph Strebel, Consultants & Research Associates, Comparative Pathology 3 Inheritance of Melanophore Patterns and Sex Determination in the Montezuma Swordtail, Xiphophorus montezumae cortezi Rosen Klaus D. Kallman^ (Figures 1-10) The inheritance of four melanophore patterns was studied in the teleost Xiphophorus montezumae cortezi endemic to parts of the Rio Panuco system, Mexico. Three of them, At (atromaculatus). Cam (carbomaculatus) , and Sc (spotted caudal) are composed of macromelanophores and the fourth one, Cb (caudal blot), of micromelanophores. The patterns are controlled by four loci that are not linked and not associated with sex. No abnormal sex ratios were obtained. At, Cam, and Cb are dominant, but Sc exhibits incomplete penetrance in the homozygous and heterozygous conditions. The penetrance of Sc in an inbred laboratory stock is about 88 percent; in hybrids between this stock and wild fish, the penetrance of Sc is only 30 percent. The frequency of the Sc factor in the population of the Rio Axtla has been estimated to be about 59 percent. Within the inbred stock, the expression of Sc may vary from a small elongate streak in the caudal fin to large melanomas that eventually destroy it. The melanoma may spread into the caudal peduncle. No fish with melanoma have been seen in preserved collections of X. m. cortezi or in hybrids between the inbred stock and wild fish. All the major populations studied are polymorphic for the four patterns, although there may be significant differences in their frequencies. The situation in X. m. cortezi, where the macromelanophore patterns are controlled by three unlinked loci, contrasts with the one present in X. maculatus and X. variatus, where the patterns are controlled by the same gene or supergene. Introduction The genus Xiphophorus provides excellent material for the study of evolutionary processes at various taxonomic levels, be- cause a variety of characters that can be analyzed genetically are present in related forms (e.g. pigment patterns: Anders and Klinke, 1965; Atz, 1962; Gordon, 1951; Kallman and Atz, 1966; Zander, 1962, 1969; sex determina- tion: Dzwillo und Zander, 1967; Gordon, 1952; Kallman, 1965, 1968, 1970a; Kosswig and Oktay, 1955; Peters, 1964; behavior: Clark, Aronson and Gordon, 1954; Franck, 1964, 1970; gonopodial traits: Gordon and Rosen, 1951; Sengun, 1949). Nine of the 17 described species or subspecies are polymorphic for one 'Genetics Laboratory, Osborn Laboratories of Marine Sciences, New York Zoological Society, Brooklyn, N.Y. 11224. or more macromelanophore patterns (Kallman and Atz, 1966; Rosen and Kallman, 1969). Best studied are those of X. maculatus. A very large number of crosses has shown that they are controlled by sex-linked factors that are members of the macromelanophore locus. How- ever, at least two cases among several thousand offspring are known in which two different macromelanophore genes have become linked to each other (MacIntyre, 1961; Kallman and Schreibman, 1971). There is also some evi- dence that a modifier is adjacent to the pigment gene that regulates its expression. Thus the macromelanophore patterns are controlled by a complex locus. Another interesting fact that has emerged from comparative genetic studies is that identical patterns in different populations of the same species have a different genetic basis (caused by different alleles at the major pigment locus 77 78 New York Zoological Society: Zoologica, Fall, 1971 interacting with population-specific modifiers) (Kallman, 1970b). All available evidence indicates that no macromelanophore factor is present in more than one species, but admittedly the patterns of species other than maculatiis have been poorly studied. The few crosses made with X. variants and X. milleri indicate that their macromelano- phore patterns are also under the control of sex-linked genes. The number of progeny raised are too small, however, to determine whether their macromelanophore factors are pseudo- alleles also. Because the gonosomes of macu- latus, variatus and milleri are homologous, Kallman and Atz ( 1966) suggested that all three species arose from an ancestral form (XX 99^ — XY SS) with a sex-linked macromelanophore locus. The two spotted patterns of X. hellerii, Db^ and Db=, are not associated with sex; no experiment has yet been performed that would test whether they are alleles. The chromosome that carries D/>' is not homologous to the sex chromosomes of macidatus (Gordon, 1958; Kallman and Atz, 1966). Atz (1962), Kallman and Atz (1966), and Zander (1965) pointed out that in X. montezumae cortezi the two known macromelanophore patterns. Sc (spotted- caudal ) and At (atromaculatus) were caused by different genes that were not linked. There is some evidence that Sc is located on a chromo- some that is homologous to the sex chromosome of niaciilatiis (Breider and Mombour, 1949; see also comment by Kallman and Atz, 1966). During a recent field trip to the Rio Mocte- ZLima, San Tuis Potosi, Mexico, some X. m. cortezi were collected with macromelanophore spotting that differed from At in consisting of fewer but larger markings on the flank. The present report is an account of the genetic basis of the new pattern and also of caudal-blot, Cb, the only known tailspot pattern for which X. m. cortezi is polymorphic (Gordon, 1940; Kallman and Atz, 1966; Rosen, 1960). Material and Methods The Montezuma swordtail, Xiphophorus montezumae Jordan and Snyder, is endemic to the Rio Panuco-Rio Tamesi drainage. Two sub- species are recognized (Rosen, 1960). As far as is known, X. m. montezumae inhabits the headwater streams of the Rio Tamesi (Rio Frio, Rio Sabinas, but not Rio Guayalejo) and the northern and western tributaries (Rio Salto de Agua, Rio Verde) of the Rio Panuco (Rosen, 1960, Darnell, 1962) while X. m. cortezi is restricted to the headwaters of the Rio Mocte- zuma and Rio Tempoal (Rio Calaboza) that drain into the Rio Panuco from the south. With the exception of seven fish (four from Rio Calaboza system and three from the rather dubious location “arroyo near Valles”), all specimens were collected along the Pan Ameri- can Highway between Tamazunchale, San Luis Potosi, Mexico, and a point approximately 44 km north of this town (figure 1 ) . The samples come mainly from the Rio Moctezuma, from the Arroyo Palitla that flows into the Rio Mocte- zuma north of Tamazunchale, from the Arroyo Matlapa that enters the Rio Axtla from the south, and from the Rio Axtla proper or small streams running into it. Fish of pedigree 1765 were collected in the Arroyo Palitla, 13 km north of Tamazunchale on April 21, 1965. Strain 38 has been derived from fish that were collected in the Rio Axtla, near the ferry crossing to Xilitla, in 1939. This stock has been maintained in the laboratory for 21 generations. An account of this stock has been presented by Kallman and Atz (1966). The system of raising fish and assigning pedigree numbers has been described previously (Gordon, 1950; Kallman, 1965). The four patterns with which this study is concerned are: Cb: caudal blotch, a large oval area com- posed of micromelanophores in the proximal portion of the caudal fin (figure 2). The Cb pattern is also present in X. m. montezumae and X. pygmaeus nigrensis (Kallman and Atz, 1966). Sc: spotted caudal, irregular longitudinal streaks or spots consisting of macromelano- phores in the caudal fin (figures 3 and 4). This gene often gives rise to melanomas in a strain maintained in this laboratory (figure 5), but not in natural populations. At: atromaculatus, a large number of black spots, composed of macromelanophores on flank and dorsal fin. Most spots concentrated below dorsal fin and in dorsal part of caudal peduncle (figures 3, 4, 5, 6). Cam (the newly discovered pattern): carbo- maculatus (from the Latin words carbo for coal and maculatus for spotted), relatively few but large spots on flank. This pattern differs from At in possessing fewer but larger markings. The dorsal fin is only rarely spotted and then only at the base (figures 2 and 7). Individual spots were counted on all fish on the left side under an xlO dissecting scope. All fish were preserved in 10 percent formalin in the Genetics Laboratory for future reference. Size of fish is given in mm of standard length. The distribution of the four patterns in natural populations was studied by examining the following preserved collections: Arroyo Matlapa, San Luis Potosi, Mexico. Gordon, Coronado, Gandy, April 14-15, 1939. UMMZ (University of Michigan, Mu- seum of Zoology) #124374 (collected at Comoca ) . Kallman: Inheritance of Melanophore Patterns and Sex Determination 79 Figure 1. Rio Panuco-Tamesi system showing major streams. Virtually all collections of X. monte- ziimae cortezi were made along a stretch of the Pan American Flighway (broken line) between Rio Axtla and Tamazunchale (T). Figure 2. X. montezumae cortezi, S, 1765-1 1,19 months after capture, 53 mm. The dark markings on the flank and the one large spot in the dorsal fin are caused by Catn. Note large size of the spots which often extend over several scale areas. The grayish elongate vertical bars (“parr marks”) are under nervous control and are not part of Cam pattern. Such bars are found in most Montezuma swordtails. The dark crescent shaped area in the proximal part of the caudal fin is caudal blot, Cb. 80 New York Zoological Society: Zoologica, Fall, 1971 Figure 3. X. inontezwnae contezi, S, strain 38, 20th laboratory generation, 12 months old, 35 mm. Heavy spotting on flank and in dorsal fin is At pattern. This pattern consists of relatively smaller spots than Cam. The irregular elongate streaks in caudal fin are caused by Sc, spotted-caudal. Figure 4. X. monteziiniae cortezi, 2, strain 38, 19th laboratory generation, 19 months old, 44 mm. Spotting on flank and in dorsal fin is caused by At. Black mark in caudal fin is spotted caudal pattern. Figure 5. X. montezumae cortezi, 9, strain 38, 21st laboratory generation, 19 months old, 42 mm. Spots on flank and in dorsal fin are caused by At. Large black area below anterior part of dorsal fin is the result of fusion of several smaller spots. Melanoma in caudal fin is caused by Sc. Kallman: Inheritance of Melanophore Patterns and Sex Determination 81 Figure 6. X. monteziimae cortezi, S, ped. 2202, 12 months old, 34 mm. Markings on Hank (17 spots) are attributed to At because of their small size. Caudal blot pattern is present in tail fin. Figure 7. X. montezumae cortezi, $, ped. 2202, 12 months old, 43 mm. Spotting pattern on flank is attributed to Cam, since many of the spots (16) are of relatively large size. Figure 8. X. variants variants, $ UMMZ #108673, 33 mm. Black area in caudal fin is caused by macromelanophores and resembles spotted-caudal pattern of X. montezumae cortezi- 82 New York Zoological Society: Zoologica, Fall, 1971 Gordon and party, April 14, 1939. UMMZ #124341 (collected at Matlapa). Breder, Jr., March 25, 1940. UMMZ-Station 87 of New York Aquarium Expedition. Rio Axtla and arroyo flowing into Rio Axtla, San Luis Potosi, Mexico. Gordon, Whetzel, Ross, April 20, 1932. UMMZ #108602 (collected at Axtla). Gordon and Atz, January 14, 1939. UMMZ #124174 (collected at Axtla). Breder, Jr., March 25, 1940. UMMZ-Station 84 of New York Aquarium Expedition (col- lected in small arroyo between Rio Axtla and Rio Moctezuma). Robinson, Nov. 26, 1957. UMMZ #174563 (collected 1 mile before Xilitla on road from Pan Am. Hwy. ). Arroyo Palitla, San Luis Potosi, Mexico. Gordon and party, April 13, 1939. UMMZ #124331 (collected at Palitla). Coronado, April 2, 1940. UMMZ #186323. McLane,Dec. 19, 1940. UMMZ #162142. Kallman and Kallman, April 21, 1965. Ge- netics Lab. collection. Rio Moctezuma, at Tamazunchale, San Luis Potosi, Mexico. Coronado, April 1, 1940. UMMZ #186319. Sanders, July 11, 1937. UMMZ #180036, #105682. Rio Calaboza (Rio Tempoat) system, Veracruz, Mexico. Creaser, Gordon and Ostos, May 5, 1930. UMMZ #108678 (collected from Rio de los Hules, 11 miles SW of Tantoyuca). Gordon, Creaser, Ostos, May 7, 1930. UMMZ #108679 (collected in tributary of Rio Calaboza, 20 miles S of Tantoyuca) . The frequencies of the patterns have been listed in Table 4, but only fish above 25 mm of standard length have been included. Smaller fish were deemed too young; in many the patterns were poorly developed and conceivably in others the pigment genes were not expressed. Three collections (numbers 105682, 108602, 124174) consisted exclusively of small specimens and none were included in the tables. No patterns were present in the four fish from the Rio Calaboza. The age of preserved fish cannot be deter- mined. Generally, size increases with age, but this does not hold true for males which virtually stop growing at the time of sexual maturity. Their sizes in the above collections ranged from 20 to 51 mm of standard length. It is a well known phenomenon in Xiphophorus that males (even siblings raised under identical conditions) become sexually mature at different ages (Rosen, 1960; Rosen and Kallman, 1969; Peters, 1964; Zander, 1965) and this accounts in part for the large variations in size. Results The results of all breeding experiments are listed in Table 1 and the postulated genotypes of the parental fish are given in Table 2. The Cb pattern is inherited as a dominant autosomal trait. One male, 1765-13, collected from a natural population (ped. 1860) and two Table 1. Pedigree and phenotype of parents $2 1765-1 + unknown 1765-2 + unknown 1765-3 + unknown 38 At Sc 1765-13 Cb 38 At Sc 1765-11 Cam Cb 1797-1 + 1765-11 Cam Cb 1800-1 -b 1800-12 Cb 1889b-l Cb 1860-11 At Sc 1889b-2 Cam Cb 1962-11 -f 1889b-4 + 1889a-ll At Sc 2085-1 Cb 1889a-ll At Sc 2043-1 Sc Cb 2085-12 Cam Cb 2096-U At 2085-11 Cb 2096-25 Cam 2085-11 Cb 2043-26 Cb 2085-13 Cam Cb 2043-3 + 2096-11 At 2096-3 + 2043-11 Sc Cb 2085-3 + 2085-14 -b 2202-1 At 2214-12 Cb ‘ Age in months at which fish were scored for presence or absence of Sc. - A strong Sc pattern obliterates Cb. ^ 12 2$ (2 Sc), 4 33 (3 Sc) at 8-10; 26 2$ (4 Sc), 21 33 (3 Sc) at 12-15; 3 33 (1 Sc) at 19. ' Non-expression of Sc at 12 months. “Non-expression of Sc at 17 months. “Non-expression of Sc at 15 months. Kallman: Inheritance of Melanophore Patterns and Sex Determination 83 fish bred in the laboratory, 2043-2 and 2043-11, the offspring of Cb parents, were homozygous for Cb (peds. 2258, 2277). All crosses of the type + X yielded marked and unmarked progeny in equal frequency. When both parents were heterozygous, a ratio of 3 Cb : 1 + was obtained. A well developed Sc pattern that ex- tends over much of the central portion of the caudal fin base can totally obscure or obliterate Cb, and this accounts for the absence of this pattern in some fish of peds. 1860 and 2277. At is inherited as a dominant autosomal trait (peds. 2043, 2096, 2202, 2222, 2270, 2471) confirming the earlier report by Kallman and Inheritance of Pigment Patterns in Xiphophortts montezumae cortezi. Pedigree and sex of offspring + + At Cb Sc -b Cb Phenotypes of offspring Cam + Sc + Cb + Cb + + + Cb Sc + Cb Age {month)'^ 1797 9$ 4 15 8 (incl. Sc 3), 13 19 1 except 3 $$, 19 $$ at 12 1798 99 1 5 1 8, except At $ SS 5 at 12 1800 99 5 5 5 $9, 3 33 at 6, SS 9 17 5 99, 17 33 at 12, 6 33 at 16 1860 99 27 22 4 see-’ 21 12 6 1889a 99 1 2 12 SS 1 1 1 1889b 99 2 1 1 12 SS 1 1962 99 3 9-12 3 2 2043 99 4 11 2 4 2 11 1 4 12-20 SS 6 7 1 1 4 13 2 2085 99 1 3 2 6 9 except 4 Cb, 1 + 35 6 4 2 4 99 and 1 -f ^ at 18 2096 99 4 1 2 8-12 except Cam SS 6 1 2 2 at 17 2202 99 4 2 1 1 1 1 1 1 2 3 1 1 12 33 4 5 2 1 1 2 1 2214 99 2 6 3 3 8 2 13 except 2 Cb, 4 33 2 6 5 CbCam $$ at 19 2222 99 7 6 4 6 2 10 33 5 6 5 11 1 2249 99 3 3 2 7 4 2 8 33 3 2 2 2 4 1 2258 99 9 7 2 16 33 4 1 12 1 2270 99 8 1 7 1 9 33 7 6 2277 99 10 2 18 except 7 fish 33 15 12 (incl. 2 Sc) at 7 2319 99 8 12-18 (3 99,3 33), S3 7 21 (599,4 33) 2471 99 1 8 2 2 1 2 4 10 33 3 1 1 3 84 New York Zoological Society: Zoologica, Fall, 1971 Atz (1966). The results of ped. 2043 establish that the loci for At and Cb segregate independ- ently. Under this assumption four classes of off- spring should occur in the frequency of 3 At Cb : 3 Cb : I At : I + and the actual result fits this expectation rather well (jf= = 3.11; n = 3; 0.5>p>0.3). The alternate possibility that the two loci are linked is ruled out by the presence of wild-type progeny. The Cam pattern is inherited as a dominant trait that is not associated with sex. There exists no statistically significant difference in the fre- quency of Cam males and females regardless from which parent the pigment factor was intro- duced (Cam from 2; Cam — 12 22, 16 + — 21 22, 14 $$, combined count of peds. 2085, 2249; Cam from P, $: Cam — 22 22, 14 $$; + — 24 22, 20 SS, combined count of peds. 1889 b, 2214, 2258). Crosses of the type Cam Cb \ + + give rise to four classes of off- spring indicating that the two pigment genes are not linked (ped. 1889b, 2085). The spotted patterns. At and Cam, are con- trolled by two factors occupying different loci that are not linked. Crosses of an unspotted parent with one that had inherited both factors yielded offspring in the frequency of 2 At : 1 Cam : 1 + (41 At, 17 Cam, 23 +, combined count of peds. 2096, 2202, 2471 ). By contrast, all crosses of the type spotted x unspotted which gave rise to either At or Cam progeny, but not both, yielded marked or unmarked fish in a rafio of 1 : 1 (65 Cam, 77 + , combined counf of peds. 1889b, 2085, 2214, 2249, 2258; 76 At, 79 + , combined count of ped. 2043, 2222, 2270). It must be pointed out that not one At progeny was obtained from two of the crosses (Cam x + , peds. 2214, 2258) in which the wild-type parent had come from pedigrees with At off- spring and, conversely no Cam fish were present in ped. 2222, although some sibs of the + parent were Cam. The results of these crosses do not support the hypothesis that the difference be- tween the two spotted patterns are caused by modifiers. Fish that are genotypically At Cam look like fish with just At. Two attempts were made to determine whether the more heavily pigmented fish contained both factors while the more lightly pigmented ones were merely At. Male 1889 a-11 was thought to possess both factors and this proved to be the case. However, female 2202-1, which looked just like any other At fish of pedi- grees in which Cam did not occur, was hetero- zygous for both factors. Of critical importance for establishing that At and Cam are not allelic was the demonstration that a wild-type female of ped. 2096 did not carry a spotted factor un- expressed (ped. 2277). Thus the wild-type fish of ped. 2096 cannot be attributed to nonexpres- sion of a pigment gene. The inheritance of Sc is difficult to study be- cause of its incomplete penetrance. From an inspection of Table 1 (and also of Tables III and IV of Kallman and Atz, 1966) it may ap- pear that Sc has a polygenic basis. In a sense this is true, since obviously many modifiers are involved in bringing about the pattern. How- ever, the series of crosses in which Sc was intro- duced into a X. hellerii genotype (Kallman and Table 2. Genotypes of Parents from Matings of Table 1 Pedigree $9 $$ 1797 + + + + + + + + unknown 1798* + + {Sc?) + + + + + unknown 1800 + + + + + + + + unknown 1860 At At Sc Sc + + + + + + + + + + Cb Cb 1889a At At Sc Sc + + + + + -b -b -b Cam + Cb + 1889b + + + + + + + + + + + -b Cam + Cb + 1962 + + + + + + + + + + + + + + Cb + 2043 + + + -b + + Cb + At + Sc + + + Cb + 2085 + + + -b Cam + Cb + + + + + -b -b + + 2096 + + + + + + -b + At + Sc + Cam + + + 2202 + + + -b + + Cb + At + Sc + Cam + + + 2214 + -b Sc + + + Cb + -b + H~ + Cam + Cb -b 2222 At + Sc -b + + + + + + + + + + Cb + 2249 + -f Sc + Cam + + + -b + + -b + + Cb + 2258 + + Sc + + + Cb Cb + + + -b Cam + Cb -b 2270* + + iSc?) + + + + + At + (Sc?) + + -b + + 2277* + + (Sc?) + + + + + + -b Sc + + -b Cb Cb 2319 -b + + + + + + + + + -b + + + -b 2471* At -b (Sc?) + Cam + + + + + (Sc?) + + + Cb + * Jn ttiese pedigrees one or both parents must have been heterozygoys for Sc. Kallman: Inheritance of Melanophore Patterns and Sex Determination 85 Atz, 1966) clearly shows that this pattern is due to a single major pigment gene. Sc does not appear to have been present in five related pedi- grees (peds. 1800, 1889b, 1962, 2085, 2319), since none of the 94 fish exhibited the pattern. Strain 38 is apparently homozygous for Sc. The pattern is present in about 77 percent of the fish ( 10th to 15th generation, Kallman and Atz, 1966; 16th to 21st generation. Table 3) and those that do not develop it nevertheless carry the Sc gene as shown by breeding experiments (Kallman and Atz, 1966). This was also the case with the female parents of peds. 2222, 2249, and 2258, and with one or both parents of peds. 2270 and 2471 (Table 1 ). The expression of Sc in strain 38 varies from a small elongate spot in the caudal fin to large melanomas that eventually destroy the fin (Atz, Kallman, and Nigrelli, 1963) (figure 5). This stock represents the only known example in Xiphophorus in which mela- nomas caused by macromelanophore genes oc- cur without prior hybridization or cannot be attributed to mutation or crossing over (Kallman and Schreibman, 1971). There is some evi- dence that both the incidence and also the de- gree of Sc expression increases with age. The highest percentage of Sc individuals and tumor- ous fish observed was in the three sibships of the 21st generation which were maintained for longer periods of time than any other generation (Table 3). Four males did not develop the Sc pattern until they were older than 17 months. However, Sc melanomas are not only found in older fish; large tumors may be present in fish as young as seven months. The Sc melanoma invariably arises in the proximal portion of the tail fin, but may eventually encompass the entire fin and also part of the caudal peduncle. In the most severe cases, the caudal fin sloughs off. Fish with a large melanoma have a high mortal- ity and often die several months or years sooner than their sibs without tumors. The first extant record of a Sc melanoma comes from the seventh laboratory generation. When fish of strain 38 are outcrossed to other stocks of X. in. cortezi, the penetrance of Sc becomes reduced to ap- proximately 22 to 30 percent. This estimate is based upon two pedigrees (1860, 1889a) in which all individuals were heterozygous for Sc and seven others (2043, 2096, 2202, 2214, 2222, 2249, 2258) in which one-half of the fish were expected to have inherited Sc (Table 1 ) . No association of Sc with sex is apparent. Assuming an XX $2 - XY SS mechanism for cortezi, fish of peds. 1860 and 1889a must have inherited Sc on the X chromosome. Conse- Table 3. Pigment Patterns in Xiphophorus montezumae cortezi, Strain 38 (16th to 21st Generations; Obtained from Matings of At Sc $2 x At Sc $$) Generation At At Sc Total Age^ Tumors- 22 SS 22 SS 22 $$ I6A 5® 2 1 1 13 16 15 12 1 16B 1 5 4 5 5 12 0 16C 1 1 1 3 2 4 10 0 17A 2 5 5 9 7 14 9 1 17B 1 4 5 5 5 10 0 17C 3 7 5 10 5 9-12 1 18A 3 2 4 6 7 8 9-12 3 18B 2 1 3 5 5 6 11 2 18C 1 2 5 4 6 6 1 1, 17^ 0 18D 1 — 4 4 5 4 11 2 19A 1 2 1 2 2 4 7 1 19B 2 — 8 7 10 7 12 2 19C 4 3 3 7 7 10 11 1 20A — 2 2 2 2 4 8 2 20B 3 3 1 2 4 5 8 1 20C 2 1 10 12 12 13 12 5 21A — — 6 8 6 8 16 3 21B 3 — 14 1 1 17 11 175 5 21C 1 ■ — 15 11 16 11 16 15 Total® 35 25 109 120 144 145 ' Age in months at which fish were scored for presence or absence of Sc. •This column does not include fish that were merely melanotic. “Three of these fish may have had Sc, but spots were of small size and could conceivably be part of At. at 17 month, at II months. “Four males showed no Sc at 17 months, but when rescored a year later the pattern was present in all of them. “For earlier generations, see Kallman and Atz, 1966. 86 New York Zoological Society: Zoologica, Fall, 1971 quently, in peds. 2043, 2096, and 2202, Sc should have been inherited in females only. However, the spotted-caudal pattern was present in both sexes. The difference in the percentage of Sc males and females is not statistically significant (because of the small number of Sc fish, the results of all three pedigrees have been com- bined: 29: 16 Sc of 70, 22.9 percent; 7 Sc of 61, 11.5 percent, P = 0.09). Similarly no sex linkage of Sc can be demon- strated, if the sex determining mechanism of cortezi is assumed to be of the WY 29 - YY $S type. Under this condition Sc males and fe- males are expected in peds. 2043, 2096, and 2202, but no Sc females should occur in peds. 2214, 2222, 2249, and 2258. This was not the case. If the data of all four pedigrees are com- bined, no significant difference between the fre- quency of Sc males and females is observed (99: 13 Sc of 88, 14.8 percent; $$: 6 Sc of 73, 8.2 percent; P = 0.19). Sc and Cam are not linked. The Cam pattern of ped. 2096 can be traced to 1765-11 and Sc to strain 38. It the patterns were controlled by factors on homologous chromosomes, no Sc Cam progeny should be present in ped. 2249. For similar reasons, the Sc and At loci cannot be linked (already reported by Kallman and Atz, 1966), since in peds. 2043 and 2202, some fish inherited both. According to the results of ped. 2277, male 2043-11 was homozygous for Cb and hetero- zygous for Sc. Since 2043-11 inherited one of its Cb factors from 1765-11 and the other from 1765-13 while Sc can be traced to strain 38, the loci for Sc and Cb cannot be located on homo- logous chromosomes. At and Cam patterns can be distinguished in sexually mature fish on the basis of number and size of spots. Cam fish have few large spots on the flank (from 1 to 16 in our experiments), and at most one or two large spots in the prox- imal portion of the dorsal fin. More than half of the markings of the Cam pattern, even in young individuals, are at least as large as one hexagonal unit that marks the reticulum. Often the spots extend over several scale areas. How- ever, there is a brief period when Cam is just developing during which the size of the Cam spots is identical with those of At. By contrast, the At pattern consists of from one to several dozen small spots on the flank and in the dorsal fin. In At fish that are 12 months or older, the number of spots may become so large that adja- cent spots fuse to form large irregular black patches below the dorsal fin. Such markings may become as large as those of Cam fish. This is one of the reasons why At Cam individuals look like those with just At. All pedigrees listed in Table 1 and figure 9 were scored independently for both patterns by the author and one or two laboratory assistants with identical results in all cases. Even among offspring of crosses of the type At Cam x ++ (identified by a ratio 3 spotted:! unspotted) Cam fish could be sepa- rated from At or At Cam progeny (peds. 2096, 2202, 247 1 ) . For example, the male of ped. 2202 with 16 spots was classified as Cam on the basis of large spot size and absence of spotting from the dorsal fin, while the- two males with 17 spots were At because the size of their spots was small (figures 6 and 7). For the same reason the female of ped. 247 1 with ten spots was clas- sified as Cam. It must also be pointed out that the size of the spots of all fish listed as Cam in Table 1 and figure 9 was similar to that of the fish illustrated in figures 2 and 7. Observations on the etiology of the Cam pattern in ped. 2258 has shown that the small number of large spots is not due to a fusion of several smaller spots. Although when different pedigrees are com- pared with each other, the spot number (but never their size) of the more heavily pigmented Cam fish may overlap with the least pigmented At fish, no such overlap was observed in the three crosses in which both At and Cam segre- gated. It is significant that in ped. 2096 the Cam progeny had fewer spots than the At fish, al- though the former were scored at 17 and the latter at 12 months. The number of spots that compose the At pattern increases with age. This is clearly illus- trated by fish of ped. 1860 scored at different ages and also by a comparison of 9 to 10 months old fish of peds. 2222 and 2270 with 12 to 20 months old ones (ped. 2043). No such increase has been noted in Cam fish (compare peds. 2085, 2249, 8 to 9 months, with peds. 2214, 2258, 13 to 19 months). X. m. cortezi with macromelanophore pat- terns have been repeatedly illustrated in the past but no distinction between At and Cam has ever been made. Only Zander (1969) has suggested that two slightly different spotting patterns may occur in X. m. cortezi. Photographs of fish with typical At can be found in Gordon (1951, Fig. 17, $; 1956, lower cover picture; 1957, Fig. 13, 9 and $), Kosswig (1936, Abb. 3, 9), Stowahse and Villwock (1969, Fig. la, 9), and Zander (1967, Tafel V, Abb. 18, 9). Fish with typical Cam have been illustrated only by Rosen (1960, Fig. 13, 3) and Zander (1967, Tafel V, Abb. 18, 3). The picture of X. m. cortezi on page 10 of Gordon (1956) presumably refers to At as judged by the small size of the spots, but their low number and their absence from the dorsal fin makes it a somewhat doubtful identification. The phenotype of the male illustrated by Koss- wig (1935, Abb. 1) is also doubtful because of the presence of only a single large spot. Kallman: Inheritance of Melanophore Patterns and Sex Determination 87 1 2 3 4 5 6 7 • • • ••• o 0 00 0X0 o o 00 o oox • • * 8 o . 9 1 0 o 1 1 12 ° ° j- ^ ocoo o IQ CO oo X oo c 14 15 00 o •*«« I i 1 o o* oox o 0X0 I { I I 1 I I • • 4 o o • •• I 1 I I 1 I I I 1 I I — I I I 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 $9 NUMBER OF SPOTS d'a’ Figure 9. Number of spots comprising At and Cam patterns in Xiphophorus montezitmae cortezi (laboratory broods). The ages at which the fish were scored are given in parenthesis (months). The number of spotted offspring of certain pedigrees is sometimes less than in Table 1, because some fish were used in other experi- ments or died (not preserved) before the spots were counted. 1) ped. 1797 (8-12) 2) ped. 1860,99 (8-10), $S (10) 3) ped. 1860,99 (12-15), 35 (12-19) 4) ped. 1889a (12) 5) ped. 1889b (12) 6) ped. 2043 (12-20) 7) ped. 2085 (9) 8) ped. 2096, /ft 99 (8-12), ^t 33 (12), Cam 99, 33 ( 17) 9) ped. 2202 (12) 10) ped. 2214,99 (13), 33 (19) 11) ped. 2222(10) 12) pod. 2249 (8) 13) ped. 2258 (16) 14) ped. 2270 (9) 15) ped. 2471 (10) New York Zoological Society: Zoologica, Fall. 1971 In preserved collections from natural popu- lations, two types of spotted patterns can be dis- tinguished that correspond to Cam and At of laboratory reared fish (Table 3). Both patterns are present in the Rio Moctezuma, Arroyo Palitla, A.rroyo Matlapa, and Rio Axtla (except UMMZ 174563), but their frequencies are not the same in the different collections. Whether these are true genetic differences be- tween adjacent local populations or are merely due to sampling error cannot be determined. Cam fish were absent from one arroyo flowing into the Rio Axtla but made up 28 percent of the fish from the Arroyo Matlapa. The frequency of At fish in the individual samples ranged from 11 to 40 percent. Between 25 and 32 mm of standard length, the number of spots of Cam and At patterns overlaps considerably (figure 10), but even within this range. Cam fish had usually fewer markings than those with At. Fish listed as Cam and those recorded as At with 10 or more spots have probably been correctly identified, but some of the individuals scored as At with fewer than 10 spots could conceivably have been Cam in which the pattern was just beginning to de- velop. Below 25 mm of standard length, only few fish can be classified unequivocally as to their pattern and, therefore, they have been omitted from Table 3 and figure 10. In larger (presumably older fish) the number of spots of At and Cam fish diverges strongly (figure 10). These data are in agreement with laboratory observations that with age the number of mark- ings increases in At fish only. Sc and Cb are also found in the four main locations, but Cb was absent from two col- lections. Cb appears to be most common in the Arroyo Palitla. The frequency of fish with Sc ranged from 6 to 43 percent. Discussion The genetic analysis of the spotted phenotypes of X. m. cortezi indicates that this form has three unlinked macromelanophore loci, each possessing the wild-type (unmarked) and one pattern allele. To this author, it seems unlikely that additional patterns will be discovered in the area of the Rio Moctezuma, Rio Axtla, and Arroyo Palitla, because in the extensive col- lections during the last 40 years only Cam, At, and Sc were present. The four patterns occur in all of the populations sampled. Nothing is known about the populations (if any) that live upstream from Tamazunchale in the Rios Moc- tezuma, Amayac, and Clara, or in the Rio Calaboza system. The evolutionary events are unknown that are responsible for the difference between inaculatus, variatiis, and milleri in which the numerous macromelanophore factors are all members of the same locus or supergene, and X. m. cortezi. Kallman and Atz (1966) have pointed out that no experiment exists that would indicate whether At or Sc or both are homologous to the macromelanophore factors of the above three species or are of independent 80. 70. 60 cn I- o S)50L Lu O ^40L CD :30. 20„ 10. • • • ’* * i :• • I n*- ° ° • o o o o o o o MM, STANDARD LENGTH Figure 10. Number of spots comprising At (solid) and Cam (circle) patterns in X. monte- ziintae cortezi from natural populations. There is considerable overlap in smaller fish, but above 30 mm of standard length Cam fish have fewer spots than fish with At. Larger fish with At have more spots than smaller individuals. The num- ber of Cam spots is independent of size. KciUman: Inheritance of Melanophore Patterns and Sex Determination 89 origin. To these considerations we may now add Cam. An alternate possibility mentioned by Kallman and Atz (1966) and Zander (1969) is that the macromelanophore factors of cortezi are genetically related to those of maculatus and were perhaps at one time members of the same complex macromelanophore locus. However, during the course of evolution they could have become separated through chromosomal re- arrangement and are now located on different chromosomes. Support for such a view is pro- vided by the observations of Kallman and Schreibman (1971) and MacIntyre (1961) that in tnaciilatus the macromelanophore factors can become separated through crossing-over. However, admittedly such crossover events within the locus are rare. For an understanding of the evolution of the macromelanophore systems, it is also important to determine whether any of the patterns of X. m. montezumae are controlled by genes that are identical with At, Cam, or Sc. Preliminary results obtained in this laboratory indicate that at least some patterns of X. m. montezumae are sex-linked. Zander (1969) has reported that in a complicated hybrid involving maculatus, hellerii, and montezumae, a pattern of X. m. montezumae that he called Sr, was located on a chromosome that segregated from the sex chromosome of maculatus. If Zander’s (1969) experiment can be confirmed, then four species of Xiphophorus are known with macromelano- phore loci located on the same pair of homo- logous chromosomes. Since the diploid number of Xiphophorus is 24 (Friedman and Gordon, 1934; Lueken and Foerster, 1969), Zander’s result can be taken as additional evidence for the suggestion that the macromelanophore genes of the various species can be traced to a common ancestral form. Of particular interest is the relative high fre- quency of Sc in natural populations. This gene more than any other has to be considered as potentially deleterious, since within strain 38 it may give rise to melanomas. This strain of X. montezumae cortezi represents the only known example in the genus in which atypical pigment cell growths caused by a macromelanophore gene. Sc, occurs without prior hybridization, or cannot be attributed to mutation or crossing- over. If the penetrance of Sc is as low in natural populations as in the laboratory pedigrees of Table 1, the incidence of Sc must be consider- ably higher in natural populations than appears from Table 4. Assuming a penetrance of 30 percent and no difference between homozygous and hetero- zygous individuals, about 83 percent of the fish from the Rio Axtia must carry Sc. Based upon this rather rough estimate, the frequency of this potentially injurious gene may be as high as 59 percent. Presumably a large number of modi- fiers are present in natural populations that keep the expression of Sc under control. No X. m. cortezi with melanotic fins or melanomas have been seen in preserved collections. This by itself cannot be taken as proof that fish with genotypes permitting the development of tumors do not occur in nature. Undoubtedly, such fish will be swiftly eliminated by predators as the melanoma develops and their swimming ability becomes impaired, and thus there will be no record. More convincing evidence for the absence or extreme rarity of such gene combinations comes from the analyses in the laboratory of broods from wild-caught females or from hybrids between strain 38 and wild stock. No melanomas were seen in such offspring (the data from Table 1 of this investigation and from Table 3 of Kall- man and Atz, 1 966 ) . The presence of tumors in strain 38 must be a product of selection in the laboratory. Although no deliberate attempt has ever been made to increase the incidence of melanomas, it has consistently been the practice of our laboratory assistants to choose as parents tor the following generation fish with well de- veloped Sc patterns at the age of 9 to 10 months. The only other pattern similar to Sc of X. montezumae cortezi is known from the popula- tions of X. variatus variatus inhabiting the Rio Cazones system (Rosen, 1960). Morphologically this pattern (figure 8) is indistinguishable from that of cortezi. Based upon two collections (Gordon, Atz, Whetzel, station 45, March 31, 1948; Arroyo Mariandrea at Mariandrea, Puebla, Mexico, about 10 miles W of Poza Rica on road to Apapantilla; Gordon, Greaser, Ostos, UMMZ #108673, May 11, 1930, unnamed arroyo, near Agua Fria, 12 miles S of Mia- huapan) , this pattern is present in approximately 10 percent of the population. No Sc-like pattern was seen in any other X. variatus collection (for complete list, see Rosen, 1960). According to Atz ( 1962) and Zander (1969), the expression of Sc of X. m. cortezi is suppressed when intro- duced into X. variatus [variatus from Rio Axtia (Atz) and of unknown geographic origin (Zander)]. The spotted-caudal patterns of vari- atus and cortezi, therefore, may be caused by different genes. The crucial test, however, can only be provided by introducing the factors responsible for the patterns in the two species on to common genetic backgrounds. Closely related to the problem of the poly- morphic macromelanophore (and xantho- erythrophore ) patterns is the one concerned with the evolution of mechanisms for sex de- termination. As was pointed out in the intro- duction, the gonosomes of three species, X. variatus, X. tniUeri, and X. maculatus, are 90 New York Zoological Society: Zoologica, Fall, 1971 homologous. The first two have a XX 99 — yet been found in wild populations (Kallman, XY $S mechanism, while in natural populations 1970a and unpublished). Macromelanophore of X. maculatus, three types of females. XX, and other pigment factors, however, are present WX, WY, and two types of males, XY and YY, on the X. Zander ( 1968 ) has recently discovered may occur (Kallman, 1965 and 1 1970a) . Al- two pigment factors, Vfi and FI, in two popu- though crossing over between the Y and W lations of X. pygmaeus nigrensis that have only chromosomes has been observed repeatedly in minimal phenotypic effects within their own the laboratory, no marked W chromosomes have species, but manifest themselves as strikingly Table 4. Melanophore Patterns in Wild Populations of Xiphophorus montezumae cortezi Patterns in preserved fish Cam Cb Cb CamCamCb Cb Location + Cam At Sc Cb At Sc Sc Sc At Cb Sc At Sc % A xtla Sta. 84 9$ 19 3 16 2 1 N.Y.A. <53 7 3 6 5 2 1 2 Cam 5 At 36 174563 99 9 8 14 4 Sc 26 S3 13 8 6 2 Cb 4 Mat la pa 124374 99 14 13 7 2 1 2 Cam 28 S3 23 9 2 6 3 1 1 At 17 Sta. 87 99 8 3 1 2 N.Y.A. 33 8 6 3 1 1 3 1 Sc 14 Cb 8 124341 99 7 3 6 2 33 8 2 4 3 3 1 Palitia 124331* 99 33 4 6 3 1 1 2 2 Cam 14 16242 99 13 2 5 3 4 3 1 1 At 29 33 11 2 11 9 2 2 1 3 111 Sc 23 186323 99 1 Cb 20 33 8 3 2 1 1 2 2 1 K& K, 99 5 2 1 1 1 1965 33 4 1 1 2 Mocteziima 180036 99 1 Cam 20 33 1 1 At 33 186319 99 4 1 1 Sc 20 33 1 1 2 1 1 Cb 7 *Not included in percentage calculations, since this is obviously a selected sample. This collection consisted of 319 specimens, but only 19 mature males are extant. Kalhnan: Inheritance of Melanophore Patterns and Sex Determination 91 red body patterns after introgression into X. maciilatns. Most significant is that Vfl is located on a chromosome that is homologous to the Y chromosome of maculatus and can replace it functionally. Although no numbers were re- ported, Zander's experiment suggests that in one of his nigrensis stock, sex-determination is by the XX 9$ — XY $$ mechanism, since all males but none of the females inherited FL Because of the homology of the sex chromosomes of inacii- latiis, variatiis, and milleri, it is unlikely that the same pair of undifferentiated chromosomes had evolved independently into gonosomes three times (or four times if X. pygmaeus nigrensis is included). Presumably, the ancestral species already had a XX-XY mechanism with a macro- melanophore locus. The W chromosome of niaculatiis may be a recent innovation (Kallman, 1970a). The sex determining system of X. hellerii is not well understood because of widely fluctu- ating sex ratios (Kallman and Atz, 1966; Peters, 1964). Several authors are of the opinion that sex determination in this genus has evolved from an ancestral condition in which sex determina- tion was achieved polygenically by the segrega- tion of many M or F factors scattered over many chromosomes to one with well defined gono- somes (Anders and Anders, 1963; Dzwillo and Zander, 1967; Gordon, 1952; Kosswig, 1964; Peters. 1964). According to these investigators, the original condition, or one similar to it, is still present today in X. hellerii. An alternate possibility has been suggested by Kallman (1965, 1968) and Kallman and Atz ( 1966). It is therefore of considerable interest that an early experiment (but unfortunately based upon hybrids of unknown origin with the identity of some of the relevant pigment genes in doubt) by Breider and Mombour ( 1949) suggested that the chromosome of cortezi that carries Sc may be homologous to the sex chromosomes of niaculatiis. The crosses listed in Table I, limited as they are because of the low penetrance of Sc, do not indicate sex linkage for Sc. This does not preclude the possibility, however, that the chromosome on which Sc is located is homo- logous to the gonosomes of maculatus and the other species, but that within X. m. cortezi this chromosome plays no role in sex determination. The data published previously by Kallman and Atz (1966) did not provide any evidence for or against the presence of a XX 99 — XY $$ or any other sex chromosome mechanism in X. m. cortezi. Of the forty pedigrees reared in the Genetics Laboratory only two showed a signifi- cant deviation from unity. Similarly, Kosswig (1959) did not find any significant preponder- ance pf one or the other sex. Only Zander (1965) reported some crosses in which virtually all off- spring differentiated into females (e.g. one male mated to three females sired 105 offspring, all but two females; a second male mated to the same three females gave rise to 183 females and one male). According to his interpretation, X. m. cortezi possesses an XX 99 — XY 55 sex de- termining mechanism that can easily be upset by autosomal factors. Males that sire predominantly female broods have two X chromosomes and a certain number of autosomal male determining factors that override the action of the sex chromosomes. The male determining potency of the Y chromosome of cortezi is less than that of the Y of maculatus (Zander, 1965). Zander pointed out that X. tn. cortezi may have reached a stage in the evolution of sex- determining mechanisms that is intermediate be- twen a polygenic (original) one and one in which sex is determined strictly by gonosomes. In the Sc males of his experiments, sex determination was thought to have been by sex chromosomes (XY 55), but in the males without Sc to have proceeded polygenically (XX). However, the significance of these relationships was not clear (Zander, 1965). Females with the exceptional sex genotype XY and males that are XX are known from X. maculatus and other species (see summary by Kallman, 1968), and there is no reason to assume that this condition cannot occur in cortezi as well, if indeed it has a XX-XY system. The abnormal sex ratio of 288 99 : 3 55 is certainly strong evidence that the male parents of these broods possessed a sex genotype more characteristic for females, but, curiously, when these males were mated to other females normal sex ratios were obtained. Because the sex chromosomes of X. maculatus carry dominant sex-linked marker genes, the recognition of genetic sex reversals (i.e. XX 55 and XY 99 ) presents no problem in this species. This is not the case for X. tn. cortezi and many other species. In this context it must be pointed out that significant deviations from an expected 1 : 1 sex ratio cannot a priori be attributed to genetic sex reversals as has been done by Schroder for Poecilia (1964) and Zander for X. m. cortezi. Other independently arrived cor- roborative evidence, e.g. sex-linked marker genes or chromosome analysis, is needed in each case. Kallman (1965) working with X. macu- latus found that the deviation from the expected sex ratio (ped. 1485, 23 XX 99, 44 XY 55, Table 1 5 ) was due to the deficiency of one class and not due to genetic sex reversals. To me the conclusion is not justified that XX males are not exceptional for cortezi and occur rather frequently (XX-Ausnahme-55 sind fiir cortezi keine Besonderheit und treten relativ haufig auf), since nowhere did Zander (1965) 92 New York Zoological Society: Zoologicu, Fall, 1971 indicate just how frequently broods with a significant excess of females are obtained. Noth- ing was reported about the sex ratios or breeding performance of his stock. The data of Kallman and Atz (1966) and Kosswig (1959) do not show that unbalanced sex ratios are of common occurrence in cortezi. Zander's interpretation was in part based upon the analysis of sex ratios of hybrids between cortezi and niaculatiis and cortezi and variatus. Sex determination in species hybrids of Xipho- phoriis while sometimes proceeding normally, as e.g. in maciilatus x pygmaeiis hybrids (Zander, 1968), is just as often contrary as to what one would expect from the sex chromosome geno- type of the hybrids (provided one uses species with marked gonosomes which is not the case with cortezi)- Kallman and Atz (1966) found that in tuaciikitiis x niilleri hybrids, sex de- termination was largely governed by the sex chromosomes (XX $$ XY 33) provided the maciilatus parent came from the Gp stock. But when the maciilatus parent came from stock Hp-2, more than half of the XX offspring differentiated into functional males, which upon breeding with XX females of either species yielded progeny (large numbers) with a sex ratio that approached unity, mimicking XX 2$ x XY 33 crosses. Other examples of atypical cases of sex determination are provided by Zander’s (1968) milleri-pygmaeus crosses and by the hellerii x maciilatus hybrids (summary in Table 26, Kallman, 1965). Even Zander (1965) admits that some of the results of his hybrid crosses fit into his scheme of sex determination for cortezi only with difficulty or not at all. The crosses listed in this paper do not help to clarify the sex chromosome mechanism of X. m. cortezi, because no sex-linked marker genes have been found. However, these results, as our earlier ones, clearly show that males (XX ?) that sire predominantly female broods are not of common occurrence in cortezi- Some of the crosses (Table 1) were among fish that repre- sent a pure line from Arroyo Palitla while others are hybrids between Arroyo Palitla and strain 38 (Rio Axtia). Sex ratios for strain 38 are listed in Table 3. Of all pedigrees, only two (Table 1 ) showed a sex ratio that deviated significantly (at the 0.05 level) from unity (ped. 1800: 10 22 — 26 33, 0.01 < P < 0.02; ped. 2471: 20 22 — 8 33,0.02) producing the dennis pattern, the other {R) add- ing the rays. The patterns would then be con- Turner: Genetics of Some Polymorphic Forms of Butterflies 129 Text-figure 3. The three main forms of the Ecuadorian species Heliconius timareta (Hewitson). About % times natural size. trolled by three chromosomes DR, Dr, and dr. This theory is supported by the closely related species Heliconius timareta from the eastern slopes of the Andes in Ecuador, one of whose three main forms has rays without the dennis pattern (Text-fig. 3). If dennis and ray are in- deed closely linked loci, then the recombination value is very low, as no individual showing ray without dennis has ever been found among the thousands of specimens exported from the poly- morphic populations of Guyane (Joicey and Kaye, 1917). Inheritance of color of bands The band on the forewing may be of the fol- lowing types, designated by letters in the tables; Y or “yellow”. A group of firm pale yellow marks outside and inside the cell; the scales within the marks are entirely yellow and not mixed with black (Plate-figs. 1-4). TY or “thin red with yellow”. Like “yellow,” but having a thin red band round the outside of the outermost yellow marks; the width of the red varies, but is usually no more than a series of red edgings to the outermost and most ante- rior yellow spots (Plate-figs. 5-9). S or “dusky yellow”. A group of pale yellow green marks in the same positions as those of Y ; the marks tend to be smaller than in Y, and the yellow patch in the cell is often reduced or absent. The yellow green color is produced by a mixture of black and yellow scales (Plate-figs. 10-13). TS or “thin red with dusky yellow”. Like S, but with a thin red band exterior to the yellow marks; this red band varies in width, but is nor- mally much wider than it is in TY, being a definite bar of red curving through the whole length of the area occupied by the band (Plate- figs. 14-19). Table 2. First Generation Backcrosses to Surinam Stock (See Also Brood 46 in Table 7) PARENTS RADIATE (DR) DENNIS (Dr) BROOD MATING 9 $ yK TY*’ S*" TS*' yF TY" S" TS*' Total 10 M36 6 3 S 0 0 0 1 0 0 0 2 5 TS^^Dr Y'^DR 9 1 0 0 0 0 0 0 1 11 M49 3 5' S 0 0 2 2 3 8 6 2 42 Y'^DR TS^Dr 9 0 6 1 0 1 3 6 2 12 M50 7 Surinam** S 0 0 0 0 3 3 1 1 13 TS^^Dr Y*'Dr 9 0 0 0 0 1 1 2 1 13 M51 5 3' $ 1 6 3 2 1 7 3 0 52 TS’^Dr Y'^DR 9 5 1 5 3 5 1 3 6 14 M52 5 3** S 5 0 1 0 1 2 3 2 25 TS"Dr Y*^DR 9 1 2 2 3 0 2 1 0 15 M60 Combination $ 1 0 1 0 0 0 0 0 3 of M51 and M52 9 0 0 1 0 0 0 0 0 Total* 14 15 16 11 11 23 22 15 127* Total* 56 71 127* Excluding brood 12. 130 New York Zoological Society: Zoologica, Winter, 1971 O or “absent band”. The band is virtually absent, and is represented only by a few faint green-yellow marks in the region of the cell, and a red C-shaped mark near the posterior angle of the wing (Plate-fig. 20). W or “wide red”. A broad red band covering all the area in the region of the cell; in Trini- dadian butterflies and in many of the hybrids, this is covered on the underside with white (sometimes yellow) scales (Plate-figs. 21-23). The Trinidadian butterflies were always W; all Surinamian butterflies used as parents of and backcross generations were Y; two others were S and TY, and their offspring are not listed with Fj’s or backcrosses. Early in the experiments, it became obvious that band-color was controlled by more than one locus. The TY female from Surinam (mate un- known) produced a brood (Table 1, brood 1) of Y, TY, S, and TS in roughly equal numbers, suggesting that two loci are segregating. The S female from Surinam produced a brood (Table 1, brood 2) (apparently a backcross) equally of S and Y; an S X Y cross produced a similar backcross brood (table 7, brood 39). The F^ Surinam X Trinidad hybrids (Y X W) are al- ways TS (Table 1, broods 5-9). An S female from brood 2 mated to a W Trinidadian male, produced approximately equal numbers of TS and W offspring (Table 7, brood 41 ) . From these results, and bearing in mind the findings of Turner and Crane ( 1962) and Shep- pard ( 1963), I formed the hypothesis about the inheritance of band-color illustrated in text-fig. 4. Two loci, B and N, are involved; the reces- sive allele b reduces the amount of red in the band, and the semi-dominant allele reduces the amount of red and increases the amount of yellow. W butterflies from Trinidad are of the genotype BBN^N^; Y butterflies from Surinam are bbN^'N^'; the S phenotype is produced by the heterozygote and the addition or subtraction of the red T mark from the S or the Y pattern is controlled by the substitution of B for b. The phenotypes of BBN^N^ and bbN^N“ butterflies (top right and bottom left of text-fig. 4) were uncertain, but I guessed that they would be TY and something similar to TS (not shown in text-fig. 4). Broods which emerged after the formation of this hypothesis confirmed it, except in one detail, the phenotype of bbN^N^. This genotype has been produced in two matings between butter- flies known to be BbN^N’^ (Table 6, broods 35 and 37), and turns out to be “absent-band” or O (Text-fig. 4; Plate-fig. 20). Apart from this modification all test-crosses performed conform to the hypothesis. Thus Y should be homozygous; broods 40, 42, 44, and 48 which are Y X Y matings produced 80 Y butterflies in all, and no other phenotypes (the anomalous individual in brood 50 will be dis- cussed later). An S X S mating should produce O, S, and Y; brood 47 (Table 7) has produced Table 3. First Generation Backcrosses to Trinidad (and a Similar Brood) PARENTS DENNIS (Dr) PLAIN (dr) BROOD MATING $ $ tsf TS' W'' W' tsf TS' W"' W' Total 16 M21 Trinidad 1 S 1 1 0 3 0 0 0 2 20 W^dr TS'Dr $ 2 1 0 2 1 4 1 2 17 M46 Trinidad 5' $ 3 3 4 7 3 6 3 3 61 W^dr TS^^Dr 9 3 4 3 6 4 2 4 3 18 M58 Trinidad 8 $ 0 0 0 0 0 0 0 0 1* WMr TS'DR 5 0 0 0 0 0 0 0 0 19 M35 6 Trinidad'’ $ 4 0 0 0 1 4 0 3 23t TS'^Dr W'dr 5 1 1 2 0 0 2 0 4 20 M41 5 Trinidad S 1 1 1 3 1 2 0 3 28 TS'^Dr WMr 9 3 0 2 2 1 2 4 2 21 M56 8 Trinidad S 0 0 0 0 0 0 0 0 It TS^'DR W'dr 9 0 0 0 0 0 0 0 0 Total 18 11 12 23 11 22 12 22 131 Total 29 35 34* 35t 134*tt Total 64 69 134*tt * 1 male, TSdr, not scored for F. t Includes one Wdr butterfly, not scored for F or sex. 1 1 male, TS'^DR. Turner: Genetics of Some Polymorphic Forms of Butterflies 131 /s/NnB bb Bb BB Text-figure 4. The interaction of the B and N loci in determining the color of the band. About 0 times natural size. S and Y, but the absence of O is not significant in a brood of only 6. A Y individual mated with a W butterfly known from its pedigree to be heterozygous Bb produced, as predicted, roughly equal numbers of S and TS (Table 7, brood 43 ) . Brood 52 (Table 7) is a cross between S and TS (the bracket round the S in the table indicates that yellow marks were virtually absent) ; it seg- regates 4 TY, 12 TS, and 5 W, a satisfactory approximation to the ratio 1:2:1 expected if the TS parent was homozygous BB (from its pedigree it had an even chance of being homo- zygous). The first generation backcrosses to Surinam (TS X Y) and to Trinidad (TS X W) also con- firm the hypothesis. Those to Surinam (Table 2) gave, as expected, Y : TY : S : TS in the ratio 1 : 1 : 1 : 1 (actual numbers, including brood 46 from Table 7, are 33 : 46 : 43 : 29; ^(-,3, = 5.2; P > 0.1). Those to Trinidad (Table 3) gave TS : W in the ratio 1 : 1 (actual numbers, including brood 16, are 64 : 70). The Fo broods, produced by sib-mating hybrid but- terflies (Table 4), segregate Y : TY : S : TS : O : W in the numbers (including brood 25) 4 : 27:9:38:0:17. The expected ratio from the hypothesis as shown in text-fig. 4 is 1 : 3 : 2 : 6:1: 3, or in numbers 5.9 : 17.8 : 11.9 : 35.6 : 5.9 : 17.8. This gives x'- = 12.2 for 5 degrees of freedom, which is significant at the 5 percent level. The Fo broods have therefore segregated all the expected phenotypes, except O, and the absence of this phenotype is mainly responsible for the significant deviation from the expected ratio. Reasons will be advanced later for think- ing that the genotype bbN^N^ may sometimes produce a phenotype very like TS; in that event the expected ratio is 1 : 3 : 2 : 7 : 3, which gives = 6.4 for 4 degrees of freedom, which is not significant. The segregation of the Fo broods therefore indicates that the hypothesis is prob- ably correct, but that there are some additional complications which are not yet understood. The genotype BBN^N^ has not been formed for certain in these broods; reasons will be given later for thinking that it does indeed produce the phenotype TY. Inheritance of shape of bands The variation in the red bands (W or T) has been explained fully by the loci B and N. The yellow marks in the band may be either broken up into a series of spots (Plate-figs. 1-2, 5-7, 10-11, 14-16), or joined together into a yellow patch (Plate-figs. 3-4, 8-9, 12-13, 17-20). These phenotypes are indicated in the tables by super- script letters, F for a broken band and f (fused) for a joined one. This variation is found in the phenotypes Y, TY, S, TS, and probably O (the only two individuals obtained being apparently O^). In the phenotype W (wide band), the vari- ation is found in the distribution of white (or yellow) scales on the underside of the band, which are evenly spread in phenotypes, but 132 New York Zoological Society: Zoologica, Winter, 1971 Q s 05 tA Q Q S U 9 o a z 9 ^ X J fu 1 i ^ 05 < 3 Q 12 Z C M 5 u S3 w s 9 < 05 w 3 o H T3 z < ►4 Q- Q C/5 z Z UJ Q Q O O qC CQ \D r-J Tt lo (N O'! m OOOOOOOO-hOOO OOOOOOOOOr<-iOO 0000000(NC500 OOOfSOOOmtSOO oooooooooooo oooooooooooo OOO — OrlOOOOOO ooo — o— O L» c/5 O ^ o i ^ (U c/5 S H D ^ X) 2 1 -rj ^ o -o CJ X Of Id =3^ (L> X ^ Turner: Genetics of Some Polymorphic Forms of Butterflies 133 Table 6. Various Test-Crosses PARENTS DENNIS (Dr) PLAIN (dr) BROOD MATING 9 $ TY'' Tyi TS’' TS’ W’' O’ TY’’ TY’ TS’' TS’ W’' W’ Total 35 M55 41 4V $ 0 0 0 0 0 0 0 0 0 0 0 1 3* WDr Wdr 9 0 0 0 0 0 1 0 0 0 0 0 0 36 M91 49 49 $ 1 1 1 1 0 0 2 1 2 0 0 0 7 TY^Dr TS'Dr 9 0 0 2 1 0 0 1 1 2 1 0 0 37 M92 52 52 3 0 0 0 0 1 0 0 0 0 0 0 1 7t WDr W'Dr 9 0 0 0 0 3 1 0 0 0 0 0 0 38 M95 52 52 $ 0 1 0 1 0 0 0 0 0 1 0 0 6 TS'Dr TY'Dr 9 0 2 0 1 0 0 0 0 0 0 0 0 * Includes Ig Wdr not scored for F or f. t Includes 1 ^ WDr not scored for F or f. gathered into patches separated by red scales in butterflies (Plate-flgs. 24-25). The two types are often difficult to score on W butterflies, and in some butterflies they cannot be distinguished because there are so few white scales (columns headed W ) ; TS phenotypes occasionally lack most of their yellow, but F or f can be detected on the underside as marks of a lighter brown than the background color. Band shape is controlled by a single pair of alleles (F and /), with broken bands dominant to fused. Thus the hybrids from the cross X (Table 1, broods 5-9) are always TS*^^'; the backcrosses to Surinam are entirely F (Table 2); and the backcrosses to Trinidad (Table 3) segregate 54F to 78f (or excluding the W pheno- types which are difficult to score, 29F to 33f). The Fo broods, excluding brood 25, give 59F to 3 If, or excluding the W phenotypes, 49F to 24f. The much closer correspondence to the ex- pected ratio in the backcross when W pheno- types are excluded, shows that scoring on these butterflies is unreliable. Inheritance of yellow line “Yellow line” denotes a narrow band of scales starting at the base of the forewing and extend- ing roughly along the posterior vein of the cell, towards the band (Plate-fig. 26). The subspecies H. m. nanna (with its variant population H. m. burchelli) is monomorphic for this phenotype (Plate-fig. 29), but in the present broods its ex- pression is very variable, and it may be repre- sented only by a yellow spot at the base of the wing; when it is combined with the dennis or radiate patterns, this dot is usually all that is visible. All the Surinam parents show the yellow line (in the form of the dot); it is absent in butterflies from Trinidad. As one has to use a dissection microscope to score this character, only two of each of the backcrosses, one Fg brood and a small sample from one F^ have been scored. The butterflies are divided into three classes: “++” The yellow at the base of the line is solid, with no mixture of black scales; the spot is visible to the unaided eye. “-f ” The yellow scales are mixed with black scales, producing a vague yellow spot not im- mediately apparent to the unaided eye. “ — ” The yellow line (or spot) is absent, or if present, the spot is represented by no more than half a dozen yellow scales. The results (Table 8) show that the yellow line is a character of variable expression, par- ticularly in TS phenotypes, but that it is strongly influenced by the N locus, such that but- terflies are usually “-f-f”, butterflies “ — ”, and heterozygotes variable but often The other factors influencing this character have not been identified, but appear not to include F or sex, which for the sake of simplicity have not been tabulated. Inheritance of white dots and yellow bar The “white dots” are a series of faint white markings, comprising all or any of the following (Plate-fig. 27) : (a) a series of up to five white spots on the veins near the tip of the forewing about 3 mms from the edge of the wing; (b) a white spot, rarely two, near the outer angle of the hindwing; (c) a series of paired marginal spots at the posterior of the hindwing. These marks are found only on the underside of the wings. They are not normally found in any of the races of melpomene in the Guianas, Venezuela, Trinidad, or Brazil, and I have never noticed them on any wild-caught specimen of any phenotype, although they are, of course, easily overlooked. All these marks are found in the closely related H. ethilla, and, all except the white spot on the hindwing, in various subspe- cies of the closely related H. elevatus (Turner, 1967). Table 7. Various Test-Crosses 134 New York Zoological Society: Zoologica, Winter, 1971 Total C/3 Q H C/3 bi C/3 Z H Z w Q C/3 H '•a. > H £b on H Cb C/3 H Di Q Xa W H C Eb c/3 Q < Pi Eb H > XA H Z w Pi < p- 0+ O z H < s Q O O Pi pa — 't" OOOOOOOOOOOOOOOOOOOOOOOO-^OfNOOOOO OOOOrl-ONOOOOOO^OOOOOOoOOOOOOOOOOOO ooooooooooooooooooooooooooo-^oooo OOOOr'^OOOOOOOOC’OOOOOO — OOOOOtNmO^OO 0000or^-^00<^ — OO^t^OOOO — (N — Tj-OOOOOOOOOOO — OOOOOOOOOOOOOOOOOOOOOOOOO — — — oooo OOOOOOOOOOOOOOOOOOOOOOOOOO — — oooo OOOOOoOO — OOOOOOOOOOOOOOO — — (N(n|0— oo OOOOOOOOOOOOOOOOOOOOOOOOOOOOO — oo OOOOOOOO — nOOOOOtNOOOOOOOOOOOOOOOO ooooooooooooooooooooooo — ooonoooo ooooooooooooooooooooooo — oooooooo 000000<^)f'I000000! NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable) , and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 6. Chromosome Numbers of Heliconiine Butterflies from Trinidad, West Indies (Lepidoptera, Nymphalidae). By Esko Suomalainen, Laurence M. Cook, and John R. G. Turner. Figures 1-4 121 7. The Genetics of Some Polymorphic Forms of the Butterflies Heliconius melopomene (Linnaeus) and H. erato (Linnaeus). II. The Hybridization of Subspecies of H. melpomene from Surinam and Trinidad. By John R. G. Turner. Plate-figures 1-37; Text-figures 1-4 125 News and Notes Lfnderwater Sounds of Southern Right Whales. By Roger Payne and Katharine Payne. Plate I; Text-figures 1-2 159 INDEX TO VOLUME 56 166 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx, New York 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, New York. Published August 7, 1972 ©1971 New York 2toologicaI Society. All rights reserved. ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 57 • 1972 • NUMBERS 1-10 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Simon Dresner, Editor & Curator, Publications and Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman', William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Chairman, Dept, of Ornithology: Joseph Bell, Curator, Ornithology; Donald F. Bruning, Associate Curator, Ornithoolgy; Hugh B. House, Curator, Mammalogy: James G. Doherty, Associate Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator. Herpetology; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus: Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director: H. Douglas Kemper, Associate Curator: Chritopher W. Coates, Director Emeritus: Louis Mowbray, Research Associate, Eield Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Senior Scientist: George D. Ruggieri, S.J., Director & Experimental Embryologist: Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist: Jack T. Cecil, Virologist: Paul J. Cheung, Microbiologist: Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist: Klaus D. Kallman, Eish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant: Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology: Martin F. Schreibman, Research Associate, Eish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate: George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne. Research Zoologists: Robert M. Beck, Research Fello'w ANIMAL HEALTH Emil P. Dolensek, Veterinarian: Consultants: John Budinger, Pathology: Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition: Kendall L. Dodge, Ruminant Nutrition: Robert Byck, Pharmacology: Jacques B. Wallach, Clinical Pathology: Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology: Harold S. Goldman, Radiology,- Roy Bellhorn, Paul Henkind, Alan Friedman, Comparative Ophthalmology: Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine ;T\\todorz KazimirofT, De/jf/irry; Alan Belson, Resident in Pathology ® 1973 New York Zoological Society, All rights reserved. Contents Issue 1. September 11, 1972 PAGE 1. The Heliconians of Brazil (Lepidoptera: Nymphalidae). Part II. Introduc- tion and General Comments, with a Supplementary Revision of the Tribe. By Keith S. Brown, Jr., and Olaf H. H. Mielke. Plates I-VI; Text- figures 1-12; Map 1 2. The Heliconians of Brazil (Lepidoptera: Nymphalidae). Part III. Ecol- ogy and Biology of Heliconius natterei, a Key Primitive Species Near Ex- tinction, and Comments on the Evolutionary Development of Heliconius and Eueides. By Keith S. Brown, Jr. Plates I-IV; Text-figures 1-4 41 Issue 2. January 8, 1973 3. The Hematological Parameters and Blood Cell Morphology of the Brown Bullhead Catfish, Ictalwus nehiilosus (LeSueur ). By Sheila R. Weinberg, Charles D. Siegel, Ross F. Nigrelli, and Albert S. Gordon. Tables 1-3 71 4. Histochemical Analyses of the Fluid and the Solid State of the Adhesive Materials Produced by the Pre- and Postmetamorphosed Cyprids of Balcmus ebwneus Gould. By Paul J. Cheung and Ross F. Nigrelli. Figures 1-6; Tables 1-10 79 5. Blood-Group Activity in Baboon Tissues. By Joseph V. Chuba, William J. Kuhns, and Ross F. Nigrelli. Tables 1-4 97 6. Captive Breeding of Orangutans. By John Perry and Dana Lee Horse- men 105 Issue 3. March 29, 1973 PAGE 7. Captive Propagation, A Progress Report. By John Perry, Donald D. Bridgwater, and Dana L. Horsemen 109 8. Status of Rare and Endangered Birds in Captivity with a General Reference to Mammals. By Donald D. Bridgwater 119 9. Venom Yields of the Philippine Cobra, Naja naja philippinensis. By Enrique S. Salafranca. Plate I; Text-figure 1; Tables I-II 126 News and Notes Preliminary Report: Status Investigations of Morelet’s Crocodile in Mexico. By Howard W. Campbell 135 Issue 4. April 25, 1973 10. Daily Activity Patterns and Effects of Environmental Conditions on the Behavior of the Yellowhead Jawfish, Opistognathus aurifrons, with Notes on its Ecology. By Patrick L. Colin. Plates I-V; Text-figures 1-21; Tables 1-4 137 Index to Volume 57 170 B D ^ -> '7 3 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 57 • ISSUE 1 • SPRING, 1972 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, Nev) York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Simon Dresner, Editor & Curator, Publications and Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman', William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology; Hugh B. House, Curator, Mammalogy; James G. Doherty, Assistant Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Robert A. Brown, Assistant Curator, Animal Departments: Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director; Stephen H. Spotte, Curator; H. Douglas Kemper, Assistant Curator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director and Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung, Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology; Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate; George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists; Robert M. Beck, Research Fellow ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants; John Budinger, Pathology: Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology; Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology; Harold S. Goldman, Radiology; Roy Bellhorn, Paul Henkind, Alfred Freidman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine; Theodore Kazimiroff, Dentistry © 1972 New York Zoological Society. All rights reserved. 1 The Heliconians of Brazil ( Lepidoptera : Nymphalidae). Part II." Introduction and General Comments, with a Supplementary Revision of the Tribe. (Plates I-VI; Text-figures 1-12; Map) Keith S. Brown, Jr. Centro de Pesquisas de Produtos Naturals Faculdade de Farmacia, U.F.R.J. Praia Vermelha, Rio de Janeiro ZC-82, Brazil and Olaf H. FJ. Mielke- Departamento de Zoologia Universidade Federal do Parana C.P. 756, Curitiba, Parana, Brazil The Lepidopterous tribe Heliconiini has recently become very useful in biological and genetic investigations. For this reason, it was important to obtain information on many previously little-studied species in the Amazon Basin and southern Brazil, and to clarify the systematics of the tribe in the light of this new information. There are 18 species in the tribe normally found in extra-Amazonian Brazil, with seven more appearing marginally on the borders of the Amazon Basin. They demon- strate rather little variation or subspeciation over this area of nearly four-million square kilometers. In the mountains of the southeast, especially in subtropical areas, they undergo dramatic and cyclical annual variations in abundance, reaching their peak in late summer and fall (February through June). The 37 species of Heliconiini found over the four-and-a-half million square kilometers of Amazonian Brazil are often very variable, and in many cases broken into multiple subspecies which often show parallelism in color-pattern with those of other species throughout the area; the subspecific divisions closely follow those observed in other animal groups and ascribed to Pleistocene weather changes. Many Amazonian heliconians demonstrate striking polymorphisms in single populations. In light of new biological and distribution data, and with relation to Emsiey’s sys- tematic revisions {Zoologica, 1963-1965), a total of 12 good species must be added to the tribe, and a further two must be recombined; the total number of species is now 66. A few remaining systematic uncertainties still remain, which could modify this number by two or three. Introduction The lepidopterous tribe Heliconiini (Nymphalidae: Nymphalinae; also fre- quently referred to as a subfamily, Heli- coniinae) has recently attracted much attention as a convenient and varied tool for biological and genetic investigations (M. G. Emsley, 19th ^ Part I of this series; see K. Brown, 1970. Part III: see following paper. - Contribution No. 260 from the Departa- mento de Zoologia, Universidade Eederal do Parana. Annual Meeting of the Lepidopterists’ Society, Washington, D.C., June 15-18, 1968). A long series of papers has been published by the De- partment of Tropical Research of the New York Zoological Society on many aspects of heli- conian life, taxonomy, mimicry, behavior, physi- ology, and genetics: Alexander, 1961a, 1961b; Baust, 1967; Beebe, 1955; Beebe, Crane, and Eleming, 1960; Brower, Brower, and Collins, 1963; Crane, 1954, 1955, 1957; Crane and Eleming, 1953; Emsley, 1963, 1964, 1965, 1970; Fleming, 1960; Sheppard, 1963; Swihart, 1963, 1964, 1965, 1967a, 1967b, 1968; Turner, 1 2 New York Zoological Society: Zoologica, Spring, 1972 1968a, 1971; and Turner and Crane, 1962. These papers have resulted in a thorough knowl- edge of the 14 heliconian species normally pres- ent in Trinidad. We have recently initiated biochemical studies on Brazilian heliconians (K. Brown, 1965, 1967; K. Brown and Domingues, 1970; Tokuyama et al., 1967) and thus have had to develop a similarly thorough knowledge of the species present in this area. We succeeded in delineating food-plants and observing aMeast some part of the early stages of all 18 heliconian species nor- mally present in extra- Amazonian Brazil (Ap- pendix I). Most species proved to be readily bred in captivity and reasonably resistant to disease, parasitism, and handling. A few species, however, were not tractable even under the most ideal conditions. These included the four high- flying and uncommon-to-very-rare species, Phi- laethria wernickei, Eueides pavana, Heliconiiis nattereri, and H. silvana ethra. Partial to com- plete information on early stages, either observed in nature or reared from fertile eggs expressed from wild-caught females, and food-plants is nonetheless available for these species (Appen- dix I). This paper presents a general view of the tribe in Brazil, with detailed comments on various species. It includes some specific corrections and additions to Emsley's recent papers (1963, 1964, 1965) on the Heliconiini, including com- ments on non-Brazilian species, and thus amounts to a complete supplementary revision of the tribe. The paper also contains a com- plete synopsis of the species occurring in extra- Amazonian Brazil and preliminary food plant data for them, as well as a brief list of the heli- conians of the Brazilian Amazon. Part III in- cludes a description of the biology of the key primitive species Heliconiiis nattereri, and a graphical formulation of the possible geohis- torical evolution of the genera Heliconiiis and Eueides. Three new subspecies from the central Brazil plateau (Appendix I) will be fully de- scribed in Part IV. Part V is a revision and dis- cussion of the mimetic silvaniforms (the first half of Emsley’s “munatui-group”) ; its taxo- nomic conclusions are incorporated into the nomenclature used in this part. Taxonomy The taxonomy of the Heliconiini was revised by Emsley in 1963-1965, reducing the recog- nized number of species in the tribe from 116 to 55. Useful papers have since been published by Turner (1966, 1967b, 1967c), clarifying the systematic positions and variations of Heliconius demeter and H. elevatiis, and the nomenclature of Dryas iiilia. Emsley’s papers require a few further clarifications, corrections, and additions in light of new information about Brazilian and extra-Brazilian heliconians, which brings the total number of recognized species in the tribe back up to 66 (see below and Appendix III). We also will separate the genus Eueides from Heliconius on morphological, biological, karyo- logical, behavioral, and chemical grounds. Emsley (1965) revised, where appropriate, the endings of all names in Heliconius and Eueides to masculine gender or genitive case. The inadvisability of such modifications of the endings of originally described specific names to agree with the supposed gender of an often changeable genus has been defended by Turner ( 1967d), with specific reference to the heliconi- ans. His comment on the modification of vesta to vest us (“Scandal in Temple. Vestal Virgins say fVe are just good friends") is truly classic and defends our preference, followed in this series of papers, for leaving all names as origi- nally proposed by their authors. Variation One of the principal reasons that heliconians have been so useful to biologists is that they vary extremely in bright and colorful wing- patterns. The perfectly parallel variation of the common species, Heliconius erato and H. mel- pomene, over essentially all of tropical America and through at least 20 distinct basic color- patterns and over 200 named forms (Emsley, 1964; Turner, 1970), is material to astonish the layman, confound the collector, and delight the geneticist. The variability in erato and melpo- niene expresses itself most luxuriantly at the borders of the Amazon Basin (map, page 71). Here, the blue to black ground color, with one or two red forewing bands and often a yellow hind- wing stripe, typical of the extra-Amazonian forms of these two species, encounters the radi- cally different Amazonian dennis-rayed pattern (Plate VI, figs. 63 and 64), also displayed by many other Amazonian Heliconius and Eueides species.® The complexity of forms occurring in a small area (northeastern Bolivia, central Ecuador, south-central Colombia, and French Guiana-Surinam are especially noteworthy) challenges the imagination. The transition zone between the Amazonian and extra-Amazonian color-patterns has yet to be thoroughly investi- gated in any part of Brazil, with the possible exception of the Obidos-Santarem area (Plate VI, fig. 64). In central Mato Grosso, the start of pattern plasticity has been recorded in the literature (Talbot, 1928) and documented by us. Unusual variations of erato and melpomene frequently turn up there, along with several other species Brown & Mielke: Heliconians of Brazil, Part II 3 normally confined to Amazonia and carrying the dennis-rayed pattern. However, at least 95 per cent of the populations are within normal limits of H. erato phyllis and H. melpomene biirchelli. In June, 1971, apparently monomorphic popula- tions of H. erato phyllis and of the dennis-rayed H. e. veniistus and H. melpomene penelope were traced to within 100 Km of each other in west- ern Mato Grosso, with no signs of hybridization being observed. By October, a wandering phyllis had apparently crossed the very inhospitable dory grassland between these colonies and in- troduced its genes into the northern vemistiis population, which showed over 30 percent of in- dividuals with hybrid characters. In this area, however, the subspecies may be effectively sepa- rated by the grassland ridge of the Serra dos Parecis, unlike in lowland Bolivia where they hybridize extensively. Two extra-Amazonian species {Heliconius silvana ethra and H. ethilla narcaea) show a moderate north-south variation that has resulted in the separation of weak but recognizable sub- species. Intergradation occurs, however, well into “typical” populations both north and south of any arbitrary boundary. These are the only two species that are appreciably polymorphic in extra-Amazonian Brazil. Four named forms 3 This pattern is apparent in the female of Eueides vibilia unifasciatus, in E. eanes and tales, and in Heliconius aoede, biirneyi, egeria, astraea, xanthocles, doris (red forms), elevatiis, melpomene, timareta+ (except nominate form), erato, and demeter (species marked here with a cross are not known from the Brazilian Amazon and are in all cases marginally Amazonian, from higher elevations on the eastern slopes of the Andes). Four additional Heliconius species (hierax+, himera+, clysonymus+, and ricini) show a similar though simpler black-yellow-red pattern, which easily may be confused in the field with that of the dennis-rayed heliconians. Eueides lampeto and Isabella, and the six species of silvaniform Heliconius [ismenius+, silvana, numata (= aristiona) , hecale (= quitalena) , ethilla, and pardalinus\ see Part V for clarifica- tion of taxonomy], have a related black-yellow- orange pattern which suggests partial mimetic association with the dennis-rayed species (see Emsley, 1964: 281 ). Species in these two genera known from the Amazon Basin which do not show a black-yellow-red (or orange) pattern are almost all either orange with black bars and wing bands; Eueides lybia, aliphera, and procula edias+, and Heliconius metharme, wallacei, hecuba+, heurippa+, luciana, cydno+, herma- thena, telesiphe+, charitonia+, sara, leucadia, antiochus, and congener^. See map, figures, and Appendix II. of silvana ethra may be found in a single popu- lation in Espirito Santo, and six named forms plus dozens of minor individual variants of ethilla narcaea can be found in southern Minas Gerais. Appreciable variation is also evident in some populations of Amazonian heliconians marginal in central Mato Grosso (Appendix I, B and Plate VI). The total amount of poly- morphic variation, however, is small, especially in relation to that observed in Amazonian popu- lations, particularly silvaniforms (see Part V of this series). Zoogeography While heliconians usually conform to major zoogeographic barriers, they are relatively strong flyers and we frequently have observed them crossing unfavorable terrain, ascending moun- tains, or apparently moving deliberately from one area to another. They are therefore not as useful as, say, the Ithomiinae for detection of finer zoogeographic boundaries (Fox, 1967). In extra-Amazonian Brazil (map), the only barrier noticeably separating the heliconians is the divide formed by the high southeast coastal mountains, a geohistorically significant boun- dary affecting many groups of plants and ani- mals. This restricts the tropical Philaethria dido, Eueides vibilia, Heliconius melpomene nanna, and H. silvana ethra and largely restricts H. Sara apseudes to the coastal plains and foot- hill canyons north of Santa Catarina (where the high mountains are close to the ocean), and sub- stantially restricts Dione moneta to the Parana- Paraguay River Basin, Santa Catarina, and Rio Grande do Sul. The two subtropical species Eueides pavana and Heliconius besckei are na- tive to the coastal and adjacent mountains which form this divide. Isolated and undifferentiated colonies of besckei (and of many other south- eastern mountain butterfly species, some of which have evolved into recognizable subspe- cies) may be found in the Brasilia area, northern Goias, and isolated highlands in northeastern Brazil. The exceedingly rare and declining primitive species Heliconius nattereri (Parts I and III of this series) is confined to a very few select areas of undisturbed extensive virgin for- est in eastern Brazil (Littoral-median region = northern Espirito Santo, eastern Bahia, and pos- sibly eastern Minas Gerais, Alagoas, Sergipe, and Pernambuco). Other than the species men- tioned above and marginal species (see below), the remaining nine heliconian species may be expected in nearly all parts of extra-Amazonian Brazil, with the exception of a very few high, cold, or excessively dry regions. Some species, notably Dione juno, Dryadula phaetusa, and Philaethria wernickei, while widespread, tend to be very strongly localized. 4 New York Zoological Society: Zoologica, Spring, 1972 Altitudinal transitions seem to present only imperfect barriers to many Heliconian species in extra-Amazonian Brazil, certainly with much less importance than as implied in Emsley (1964, 1965). Agraidis vanillae maculosa flies from sea level to the highest areas in southern Brazil (near 3000 meters). In the genus Heli- coniits, the Andean species telesiphe was men- tioned by Emsley (1965) as the only species normally flying at elevations above 1300 meters. In southern Brazil, however, H. besckei is found from sea level (locally) to its mecca at 800 to 1600 meters. From there it ranges upward in summer to more than 2000 meters. In late sum- mer, even H. erato phyllis can be found breed- ing at nearly 2000 meters elevation, many kilo- meters from the nearest valley below 1300 meters. Many additional Heliconius species (hierax, heciiba, heurippa, timareta, cydno, himera, and clysonymus, as well as some local races of erato and melpomene) also have been found by us and by other collectors in the 1300-2500 meter range on the Amazonian slopes of the Andes in Bolivia, Peru, Ecuador, and Colombia. However, it does seem that heights above 2500 meters are impassable for Heliconius other than telesiphe (though not for A graulis or Dione ) . Cyclic Annual Variations in Abundance Many field observations suggest that most heliconians undergo great annual variations in abundance that are partially but not wholly re- lated to extensive new growth on the passiflora- ceous foodplants, and possibly accompanied by appreciable range expansions. They commonly appear in late summer and fall in areas where they do not survive or are drastically reduced in numbers in winter. A similar pattern has been observed in the two North American subspecies of Agraulis vanillae, and has been suggested for Dione moneta in Texas (Gilbert, 1969). Several areas of intermediate elevation have progressively increasing numbers of tropical heliconians from January (mid-summer) into late May or June, followed by disappearance (or great reduction) during the winter and spring. Areas at 600 to 1200 meters elevation in the coastal mountains, such as Curitiba (Parana), P-etropolis (Rio de Janeiro), and Santa Teresa (Espirito Santo), provide good vantage points for observing this. In Curitiba, at 900 meters elevation, Heliconius sara apseudes and H. erato phyllis begin appearing only in summer; by early fall they are frequent, but they seem to disappear with the first winter frosts. Both species fly all year around in the neighboring lowland areas; no definite information is available suggesting seasonal diapause mechanisms in Heliconius species in Brazil. H. silvana robigus and H. sara apseudes appear only in January and may be found only through May on the seaward slope of the coastal mountains in Petropolis, an area 1000 meters in elevation and without winter frosts. In relatively warm Santa Teresa, at 600 to 800 meters elevation and above a rich tropi- cal tableland area, Eueides vibilia, Heliconius melpomene nanna, H. sara apseudes, H. silvana ethra, and H. nattereri are encountered prin- cipally from January through June; they some- times are very common in March and April, depending upon the year. These five species have actually been observed by K. B. moving up and down the seaward face of the mountains; and they appear first each summer at lower eleva- tions near where larger streams run down the mountain-face. The increased abundance of the mountain species Eueides pavana and Heliconius besckei at sea level near the foothills in winter, when they are not commonly encountered on the colder mountaintops, may result from a dimin- ished upward movement of individuals from these populations during the cooler months. The Chapada de Guimaraes in central Mato Grosso seems to have an invasion of Amazonian species from lowland northern Mato Grosso in fall. The marginal species mentioned in the sec- tion below all have been found in this region rarely or not at all from September to February, commonly in May to July. Finally, Prof. Dr. Heinz Ebert of Rio Claro, Sao Paulo, has observed that the southwestern species Dione moneta is common in central Sao Paulo in March to June, hut very rare or absent there during the rest of the year. Whether local populations simply build up in fall, or the species invades from the south or west in re- sponse to unusual weather conditions or popu- lation pressure, has yet to be established. A general discussion of the annual variation of butterfly frequencies in Brazil has also been published recently by Dr. Ebert (1970). Marginal Species in Extra-Amazonian Brazil The borders of the Amazon Basin in north- eastern and central Brazil (map) are not well- marked zoogeographical barriers, being visible essentially as a gradual change from humid for- est to more dry and open vegetation, and a number of normally Amazonian species of heli- conian (Appendix I, B and II and Emsley, 1965) cross them into adjacent areas of the extra- Amazonian region. Two areas have received special attention. One is the Cuiaba region of central Mato Grosso. Although in the basin of the Paraguay River, this region has a large influx of Ama- Brown & Mielke: Heliconians of Brazil, Part // 5 zonian Lepidoptera, especially in the highlands ringing Cuiaba from the north (Rosario Oeste, Melguira, Serragem, Nobres, Tombador, Quebo) around through the northeast (Chapada de Guimaraes, Buriti) to the east (Sao Vicente). Amazonian heliconians found by us to be mar- ginal in this area include Eueides vibilia imi- fasciatiis, Heliconius sara thamar, H. inelpo- mene burchelli, H. wallacei flavescens, H. ethilla as a new subspecies (see Part IV of this series), H. xanthocles melete, and Eueides Isabella isa- bella. The first three of these species also are found locally as far as the Parana drainage in central and western Goias, and H. sara thamar has been captured in the San Francisco River drainage in northwest Bahia (Rio Sapao; speci- mens in the Carnegie Museum). In the northwesternmost tributaries of the Paraguay River in central-western Mato Grosso, between the swampy Pantanal around Caceres and the high grassy ridge of the Serra dos Pare- cis, is an eastward extension of the north Boli- vian rain forest, which is in turn contiguous with the general Amazonian forest (Hylaea) down the Rio Guapore. Here the flora and fauna are very strongly Amazonian. In the heliconians, H. erato remains as the southern subspecies phyllis; but the Amazonian species H. numata (many variants), H. silvana as subspecies mirus, H. aoede as a new subspecies (Part IV), //. bur- neyi, and El. leucadia were all discovered in a few hours’ collecting on the upper Rio Branco, a major tributary of the Rio Caba^al, in June 1971. The last species was also found in the upper Rio Jauru to the west, only a few kilo- meters from the upper Rio Guapore, which flows to the Amazon within the Hylaea. Addi- tional north Bolivian species which might be found with more intensive collecting in the Caba^al-Jauru area include Philaethria dido (known from coastal extra-Amazonian Brazil but not from the interior), Eueides lybia (pos- sibly observed already on the Rio Branco), Heli- conius doris, H. hecale (members of the sisy- phus subspecies complex), and H. elevatus perchlora. A further four species have been recorded as occurring in the “Cuyaba-Corumba River Sys- tem,” a rather ill-defined area which may or may not be restricted to the Paraguay drainage in central Mato Grosso (not the Corumba River in southern Goias, however): Heliconius ricini, H. astraea as a new subspecies (Part IV) (one in the British Museum, fide J. R. G. Turner, and one in the Kaye collection, now part of the Allyn collection), H. elevatus schmassmanni (which may only be a variation of Neustetter’s aquilina) , and H. demeter eratosignis (for the last two, see Joicey and Talbot, 1925 ) . We have searched in essentially all habitats in central Mato Grosso without locating any of these spe- cies. They would not invade from the north- west where the lowland Hylaea extends into extra-Amazonian Brazil, since here they were not found south of the Pimenta Bueno area in southeastern Rondonia (map). In this region, still 400 Km north of the Cabagal-Jauru forests, Heliconius erato makes a transition from the dennis-rayed, open-forewing-banded form, ama- zona, to the reduced dennis-rayed, compact- forewing-banded form, venustus (Plate VI, fig. 63). In view of the known parallelism of forewing band modifications in Amazonian dennis-rayed Heliconius (Appendix II), it would be expected that demeter, astraea, and elevatus would sim- ilarly acquire a compact square forewing yellow patch in areas adjoining the northwestern Para- guay Basin. Indeed, the Pimenta Bueno popu- lation of elevatus already shows about 50 per cent of the subspecies perchlora with this com- pact band. The four species may invade from the northeast into the extreme upper Rio Cuiaba, in a forested highland which gives birth also to five major Amazonian rivers, or into the north- eastern Pantanal, where highly suitable lowland forest habitat exists. However, in northeastern Mato Grosso, the Hylaea is 400 Km north of the Cuiaba river system, connected with it only by sparse riparian forests very poor in Helico- niini. We are inclined to regard the specimens labelled “Cuyaba-Corumba River System” as originating from northern Mato Grosso or east- ern Rondonia; until these species are confirmed in extra-Amazonian Brazil, they will remain on the hypothetical list. Here we place also H. anti- ochus, which lives commonly in northeastern Mato Grosso, even well south of the Hylaea in typical erato phyllis/ melpomene burchelli ter- ritory (the dry southeastern Amazon). The second area is in northern Ceara, where a number of mountain ranges (Serras de Ibia- paba, Uruburetama, Maranguape, and Baturite) break the palm-grassland plain and provide oases for unusual butterfly life. Also included in this marginal region is the area of Dom Pedro in southern Maranhao, but not northwestern parts of Maranhao which are decidedly Ama- zonian. Dr. Dmitro Zajciw collected these areas in 1962-1963, and donated many of his speci- mens to the Museu Nacional in Rio. His mate- rial includes Heliconius erato phyllis, H. melpo- mene burchelli, H. ethilla near eucoma, and H. ricini. The last three are not known from further southeast along the Brazilian coast. Other species known from central Maranhao and probably present in the extra-Amazonian portion, which has not yet been visited by the 6 New York Zoological Society: Zoologica, Spring, 1972 authors, are Eiieides lybia and i. isabella, and Heliconius doris, wallacei, biirneyi, numata, sara thamar, and antiochus. Two Andean butterflies have been recorded in Misiones in Argentina (Hayward, 1951) and might eventually be found in western Parana or Santa Catarina: Heliconius numata (= aristi- ona, see Part V of this series) splendidus and Dione glycera. To our knowledge, however, nei- ther of these has yet been captured in Brazil. Some Specific Comments on the Species Philaethria dido and P. wernickei, two large black-and-green species very similar on the dor- sal wing surface, are clearly distinct on the ven- tral surface and have consistently different male genital valves (Text-figs. 1 and 2). Furthermore, they are sympatric at least over all of the lower and middle Amazon Basin {wernickei as sub- species pygmalion) and along the east coast of Brazil as far south as Rio de Janeiro (the Museu Nacional in Rio contains long series of both species from a number of localities). They may be readily distinguished, at times even in flight, by the under surface of the hindwing (Plate I, figs. 1 and 2) : dido has a much redder ground- color (wernickei is black or gray), large silvered intervenal marginal spots (wernickei has only a series of faint submarginal whitish streaks), and a long white costal stripe limited by the subcostal vein (wernickei has a short white streak which drops below the vein and covers the black upper border of the green area). The flight of dido tends to be higher, more rapid, and less interrupted than that of wernickei; both species may be frequently encountered on hill- tops. The mature larva of wernickei is much deeply more deeply and richly colored than that of dido (Beebe, Crane, and Fleming, 1960), with a dark brick- red head and prolegs (orange in dido)\ the pupae of the two species are nearly identical. Agraulis lucina (Felder), a singular form from the upper Amazon and Andean slopes, should be separated from A. vanillae, with which it has been treated as conspecific in the past. The former possesses a dramatically dif- ferent color-pattern on both wing surfaces and its wings are of a distinctly different shape from those of A. vanillae (Plate I, figs. 3 and 4). The two species occur sympatrically in much of the Brazilian upper Amazon and on the Andean slopes of Peru, Ecuador, Bolivia, and Colombia. In the areas where they commonly fly to- gether, occasional specimens of vanillae (catella Stichel) show a coagulation of the dark mark- ings on the dorsal wing surface and a reduction of the silvering on the ventral surface, looking thereby somewhat like lucina (Plate I, figs. 3 and 4). However, catella retains the light forewing apex, broader hindwing, and dark distal spot in hindwing space Rs-Mj, all typical of vanillae. In these same areas, occasional specimens of lucina are more heavily silvered ventrally, add- ing to the impression that the two species inter- grade. Indeed, they may well interbreed occa- sionally where they meet, though they occupy different ecological habitats; vanillae lives in open areas and second growth, while lucina lives in forest clearings. The overall behavior of lucina in the field is closer to that of Dione juno or D. moneta, which it also mimics in color-pattern, than to shallowly notched shorter evenly curved Text-figure 1. Philaethria dido, male, left genital valve, external. Text-figure 2. Philaethria wernickei or P. w. pygmalion, male, left genital valve, external. Brown & Mielke: Heliconians of Brazil, Part II 1 that of vanillae. The egg of lucina, expressed with difficulty from the female, is noticeably smaller and more spherical than that of vanillae, and a number of expressed eggs failed to hatch. This suggests that the eggs of lucina may be laid in clusters, as are those of jiino (but not those of vanillae). Male genitalia of lucina could be consistently distinguished from those of vanillae (including catella) by the form of the process on the inner face of the valve. In lucina, the upper flange of the process is flared upward and narrowly serrated, while the lower posterior edge bears no teeth. In vanillae, the upper flange is curved inward and heavily serrated, and the lower posterior edge is usually denticulate (Text-figs. 3 and 4). Eueides pavana, poorly represented in mu- seums, is not at all rare in southeastern Brazil, occurring frequently in all the mountain area and sparsely down foothill canyons to the out- wash plains at sea level. Its tendency towards high flight makes it somewhat difficult to capture. Two color phases of the female exist, evi- dently in nearly equal numbers. One resembles the orange male, and the other is a pale straw color. Intermediates with the forewing light- colored and part or all of the hindwing orange (Plate I, fig. 5) also may be encountered. The species is sympatric with Eueides vibilia over much of the low-altitude part of its range (up to 800 meters, including the city of Rio de Janeiro); the dimorphic female of vibilia is quite similar in appearance to pavana, and both participate in the mimetic complex of Actinote spp. (distasteful Acraeinae), being almost in- distinguishable from these on the wing (Plate I, fig. 5). Eueides vibilia is also sympatric with the morphologically very similar Eueides lampeto in the Guianas (/. copiosus, Plate I, fig. 6), the Brazilian Amazon (/. copiosus and /. lampeto), and at various points on the eastern slopes of the Andes. As both species are quite localized and not frequently taken by commercial collec- tors, their micro-sympatry is not easy to prove. The two differ appreciably in size {lampeto is appreciably larger), wing-shape, and color- pattern. All races of lampeto demonstrate a large diffuse dark spot at the inner angle and an inward-directed dark triangle at the margin of forewing space Cuj-Cuo (a wing area very use- ful in taxonomy both in heliconians and itho- miines) and heavy dark intervenal postcellular marks on the hindwing, lacking in vibilia. No intermediates are known from areas where the species fly near each other. The egg of E. lam- peto carbo (Coroico, Bolivia, 1600 m) was to- tally dissimilar to that of E. vibilia unifasciatus from nearby Mato Grosso and Rondonia. The egg of lampeto is larger, creamy yellow instead of white with a pink cap as in vibilia, closely resembles that of E. isabella, and is laid singly rather than in large rafts as in vibilia. The ap- pearance and individual feeding-pattern of the solitary first-stage larva of lampeto are very simi- lar to those of gregarious vibilia larvae; both may be distinguished by the light-colored rather than black head from all other known first-stage Eueides larvae except E. aliphera. Thus, we regard lampeto and vibilia as closely related but distinct species. Text-figure 3. Agraulis vanillae, male, left genital valve, internal, and detail of upper flange of process. Text-figure 4. Agraulis lucina, male, left genital valve, internal, and detail of upper flange of process. New York Zoological Society: Zoologica, Spring, 1972 Heliconius egeria and H. astraea, represented in long series in the Museu Nacional, Rio, con- sistently show great differences in the male genital-valves (Text-figs. 5 and 6; see also Ems- ley, 1965), and usually in wing-shape and color-pattern. They apparently are sympatric at least at Sao Paulo de Olivenga on the upper Amazon and at Manicore on the Rio Madeira, where they do not intergrade morphologically, and so probably should be treated as separate species. Some middle and upper Amazonian and Venezuelan egeria (e. hyas Weymer) look very much like south-middle Amazonian astraea (which lack a name, having been assigned to e. hyas in the past; see Part IV for description) (Plate III, fig. 11); egeria usually has a more pointed forewing, with the hindwing rays more abbreviated and vein Cu red (in astraea, always black). However, as the color-patterns of the two are quite variable, and sometimes closely approach each other where the species fly to- gether, examination of the genitalia is advisable for positive identification. It is worthy of note that the Amazonian dennis-rayed heliconians {Eueides tales and eanes, and Heliconius aoede, burneyi, egeria, astraea, xanthocles, elevatus, and demeter) may be found flying together with red-forewing-band extra-Amazonian subspecies of erato and mel- pomene, not only in central Mato Grosso (e. phyllis and m. burchelli with aoede new subsp., burneyi, xanthocles inelete, and possibly astraea new subsp., elevatus schmassmanni, and demeter eratosignis in the northern part of the state), but also in Peru (e. favorinus and m. amaryllis with aoede cupidineus, xanthocles melior, burneyi huebneri, and elevatus pseudo- cupidineits, and Eueides tales calathus at Tingo Maria), north-central Colombia (e. guarica and m. melpomene with burneyi lindigii and xan- thocles flavosia and polymorphs, and Eueides eanes and tales cognatus above Villavicencio) , and in northern Para, southeastern Venezuela and adjacent Guyana (e. hydara and m. melpo- mene with aoede near astydamia and a. aoede, b. burneyi and b. catherinae, e. egeria and e. hyas, X. xanthocles, x. vala, and v. subsp. incog., elevatus barii, e. tumatumari, and e. roraima, and demeter bouqueti and d. beebei, and Euei- des tales surdus in Obidos to Itacoatiara and northward, Bolivar and western Guyana, and coastal Surinam and Guyane; see also map, fig. 64, and Masters, 1969; many of these species are polymorphic in this area for dennis only/dennis- ray). Upper Amazonian subspecies of xan- thocles {x. melete, x. melittus, x. melior, and .r. flavosia, plus several additional named and unnamed morphs) are unusual in lacking the subapical yellow forewing band typical of the lower Amazonian and Orinocan x. xanthocles and V. vala {x. paraplesius and the southeast Venezuelan race, probably new, are intermedi- ate, showing partial fusion of the two yellow bands). They also possess many of the minor characters (Emsley, 1965) of H. aoede: paired intervenal submarginal white spots, a longer an- Text-figure 5. Heliconius egeria, male, left genital valve, internal. Text-figure 6. Heliconius astraea, male, left genital valve, internal. Brown & Mielke: Heliconians of Brazil, Part II 9 terior red basal spot, and a longer yellow costal stripe on the ventral surface of the hindwing; and prominent yellow lateral dots and interseg- mental annuli on the abdomen. The males, however, have typical xanthocles genitalia and broad, rounded forewings, while sympatric sub- species of aoede (in southwestern Brazil, low- land Peru, and Venezuela) have the genitalia and deep triangular forewing, with an excep- tionally broad androconial area on the costal half of the upper hindwing, typical of that spe- cies. Females in these areas may be quite diffi- cult to distinguish. In Mato Grosso, males of xanthocles melete have a rapid, fluttery flight, often quite high above the ground, and cover a large area in their promenading. The species is thus somewhat reminiscent of H. nattereri (see Part I and Part III), another primitive Heliconius occurring in extra-Amazonian Brazil, although in xanthocles the sexes are identical, though differing in flight habits.^ It was not possible to express fertile eggs from the short abdomens of either H. aoede or H. xanthocles females, suggesting that the eggs may be laid all at once, and the caterpillars may be gregarious. This is so in H. antiochus, H. Sara, and Eueides vibilia, other species with short abdomens, which give no fertile expressed eggs. Some Notes on the Melpoinene-GKOVP Four species must be added to the inelpo- mene-group as defined by Emsley in 1965 as part of the '‘nutnatiis-group” One of the two Brazilian species (H. besckei) was mentioned as probably distinct in Emsley’s 1965 revision; later experiments have confirmed its specific status. The second species (liiciana, which may be shown to be conspecific with elevatits), is unusual in that it escaped detection until 1960 (Lichy). The tropical and subtropical habitat prefer- ences of H. melponiene and H. besckei, respec- tively, in southern Brazil, do not often permit their occurrence at the same locality. Three areas where they have been found flying to- gether are central Espirito Santo (in river valleys and up the mountain slopes to at least 1000 meters, principally in late summer), the foothill canyons near Rio de Janeiro (where melpomene nanna is very scarce), and the Brasilia area in the central plateau (where besckei flies with m. burchelli). The two may also be found to- gether over much of southern Mato Grosso and in northern Goias and other isolated mountain areas in northeastern Brazil. Typical besckei also have been recorded as far west as Santa Cruz de la Sierra in Bolivia, where they fly with a polymorphic population of melpomene in which the predominant subspecies amandus is partially infused with genes from the Amazonian penelope. In all of these areas, no signs of intermediate characters have been found in many dozens of melpomene and besckei examined. Eggs of besckei from Petropolis ( 1000 meters, near Rio de Janeiro) were bred through to adults on Passifiora sidaefolia. The egg, larva, and pupa were very similar to but distinct from those of melpomene (see Beebe, Crane, and Fleming, 1960). It should be noted, however, that these early stages are subject to much permissible variation, and that striking differences were ob- served by the first author in the size, color, and patterns of the eggs, larvae, and pupae of geo- graphically separated but indubitably conspe- cific populations and subspecies of Heliconius melpomene, H. erato, H. wallacei, and many silvaniform Heliconius. Furthermore, male besckei showed no response to virgin females of H. melpomene flagrans in Trinidad (Emsley, 1970); indeed, they showed no reaction to any exclusively red-banded heliconians, but indulged in social chasing with red-and-yellow banded H. erato phyllis reared there from a previous shipment from Rio. In life, H. besckei tends to have a higher and more fluttery flight, but also with more plan- ing, than H. tnelpomene. The tip of the male genital valve in besckei is elongated, silvani- form rather than melpomeneform, similar to that observed in the closely allied and evidently allopatric species H. elevatus (Turner, 1967b). The ventral hindwing costal streak and basal spot complex are so different between besckei and elevatus, however, that it is highly unlikely that they could be conspecific. They may even- tually be found flying together in central or southwestern Mato Grosso, meeting-ground of the Amazonian and southeastern forms of mel- pomene which these two species resemble in their respective ranges. The existence of closely parallel species in the melpomene group and the sara group, H. cydno — H. sapho’’^ and H. pachinus — H. hewitsoni, suggested the possibility of a melpomene-linkQd species parallel to the ^urn-linked H. antiochus. This species, luciana Lichy, was finally discov- ered in southern Venezuela in the late 1950s. A single female of luciana, from near Boa Vista in the Brazilian territory of Roraima, on the southern slope of the Venezuelan highlands, is present in the collection of the Museu Nacional, Rio. The wing-pattern of this specimen is very similar to that of the sympatric and common H. antiochus alba on the dorsal surfaee. How- ever, there are elements in common with cydno and elevatus (both in the melpomene group) 10 New York Zoological Society: Zoologica, Spring, 1972 *A further presumably quite primitive heli- conian species observed to have very similar large-scale promenading behavior is Heliconiiis hecalesia formosus in Panama. ® We have observed the complete sympatry of Heliconiiis heiirippa with H.m. melpomene (Plate II, fig. 8) in the Rio Negro area above Villavicencio, Colombia, where they are among the 32 species of heliconian present (of a total of 51 so far recorded, with six more expected, in Colombia — a very high percentage of the 66 species in the tribe). We are grateful to Dr. E. W. Schmidt-Mumm and his brother Helmut of Bogota for opportunities to visit the latter’s property on the Rio Negro in 1969 and 1971. Here, H. heiirippa and H. melpomene fly to- gether in the altitude range 600 to 1600 meters, with melpomene found principally on the forest edge at lower elevations, and heiirippa prin- cipally in clearings within the forest at higher elevations. However, they are frequently ob- served together. The eggs of heurippa (expressed from females) had more vertical ridges (17-18) than those of melpomene', the caterpillars and pupa, reared from these eggs showed many small but consistent differences from the correspond- ing early stages of melpomene. We judge heurippa to be a good “splinter species,” prob- ably originally arising from a yellow-banded ancestor of melpomene (see Emsley, 1964). The red outer band of heurippa is very incon- spicuous in the field, and not likely to be useful in courtship recognition (see below, in text). Heliconiiis cydno, a closely related species which also could be regarded as involved in the ancestry of heurippa, and which has nearly identical field behavior with the latter species, lacks the red basal dot pattern on the ventral surface of the hindwing shared by melpomene and heurippa, having in its place a variable U-shaped red-brown marking across the middle of the wing. H. cydno is absent from the re- stricted area where heurippa flies above Villa- vicencio, but could reach it, as have H. erato giiarica, m. melpomene, and charitonia bassleri, presumably by going around to the north in southwestern Venezuela, or across several low passes in the southeastern Colombian cordillera. H. cydno in near-typical forms is definitely present on the southeastern slopes of the Vene- zuelan cordillera at Barinitas, and (together with another central valley species, H. ismenius, which has also been recorded near Villavicen- cio) on the Amazonian slopes of the eastern Colombian cordillera above Florencia. In the central valleys of Colombia, the morphologically very close cydno and melpomene are common, fly together, and occasionally hybridize, pro- ducing little-known intermediate forms; these either strongly resemble melpomene {rubellius, seitzi', K. B. took one of these in Victoria, Cal- das, on Jan. 21, 1971, in normal courtship with a female melpomene, not recognizing the hybrid until it was in the net), or have the double yellow-and-red forewing band as in heurippa and strongly resemble cydno in the field (wer- nickei, emilius). All of the hybrids have at least part of cydno's U-shaped red-brown mark on the ventral hindwing (this is variable enough in the parent populations of cydno to admit its near absence in a hybrid, however); all show reduced but clear red basal spots intermediate between the three large dots of melpomene and the lack of spots of cydno. No hybrids are known to us from the heurippa area to the east of the eastern cordillera, and heurippa does not show hybrid characters other than the double- colored band. The polymorphic yellow-banded species Heli- contius timareta is completely sympatric, with no signs of intergradation, with the very differ- ent H. melpomene plesseni (and occasionally, with some of its intergrades to H. m. aglaope) in eastern Ecuador between 1000 and 1800 meters (Plate II, figs. 9 and 10). The field be- havior of timareta is very similar to that of heurippa and cydno', it flies fairly high above the ground, and indulges extensively in repeti- tious promenading over set courses. It is found much more inside the steep pre-montane forest than is the sympatric melpomene, which prefers edges and riverbanks. H. timareta should be re- garded as another distinct “splinter species,” isolated at moderate elevations in the eastern Ecuadorian river valleys, and possibly closer systematically to cydno than to melpomene. Reproductive isolation from melpomene almost surely takes place by a color-courtship mecha- nism (see below, in text). ® The species regarded as sapho in Emsley (1965) is divisible into at least two fully sym- patric species (Plate II, fig. 7) with dramatically different flight habits and behavior. One, repre- sented by sapho and probably by leuce, occurs from southern Mexico to western Ecuador, flies high and slowly, frequently visits flowers, and is quite attached to one place both when feeding and when occupying a territory. The other, rep- Brown & Mielke; Heliconians of Brazil, Part II 11 resented by eleuchia and primularis, and prob- ably eleusinus and ceres, occurs from central Panama (Colon) to western Ecuador, and flies low, rapidly, and in a straight line, not stopping at flowers or remaining over long periods in one area. We are grateful to Dr. E. W. Schmidt- Mumm of Bogota for detailed information on the sympatry and habits of these species in Colombia; we have fully confirmed his observa- tions in Panama, Ecuador, and in museum col- lections. Dr. Tarsicio Escalante of Mexico also provided key information on the field behavior of H. sapho leiice. Where sapho and eleuchia fly together (Panama, central Colombia, and western Ecuador), they show no intergradation and rarely occupy the same habitats in the for- est; in these areas, the latter species invariably has a shorter red costal streak and anterior red basal spot on the ventral surface of the hind- wing, in relation to those of the former. Where only one form in the complex is known (leuce from Mexico to Costa Rica, and eleusinus and its yellow morph ceres along the west coast of Colombia), this form shows the field behavior of sapho but the shorter red spots of eleuchia. Tentatively, leuce is placed with the former spe- cies; a short series from Narino in extreme southwestern Colombia, present in the Instituto Oswaldo Cruz in Rio, strongly suggests inter- gradation of primularis with eleusinus through ceres and varieties; we thus tentatively place these three forms together with eleuchia. A further very different-appearing and allo- patric form which flies east of the Andes in Colombia, Ecuador, and northern Peru, H. con- gener, has the field habits and shortened red basal spots of eleuchia. If its reported chromo- some number (33) is correct (de Lesse, 1967; presently being reconfirmed), it should stand as a good species. The allopatric H. hewitsoni, known only from the “Chiriqui” faunal region in southern Costa Rica and northwestern Pa- nama, seems to merit its presently accepted spe- cific status. There are thus most probably four species in the sapho-complex, apparently still in rapid evolution as the most recent major group in the Heliconiini. Heliconius cydno also shows a separation into forms (c. chioneus and c. cydnides) which closely resemble sapho and eleuchia in Colom- bia. We have little field experience with the forms related to cydnides, and cannot com- pletely eliminate the possibility that cydno may eventually be divisible into more than one spe- cies when more information becomes available, although this seems unlikely. In addition, cydno has some very unusual related forms in Colom- bia (c. hermogenes in the upper Magdalena valley, c. weymeri and its form gustavi in the Cauca Valley) which do not resemble members of the i'ap/m-complex, but approach other spe- cies of Heliconius (notably hecalesia and erato chestertonii) flying in the same areas. Kaye (1917) argued for the separation of these forms, which also frequently show a diminished or absent U-shaped red-brown mark on the ventral hindwing surface, from cydno. In defense of the unity of the species cydno, the following facts are presented: ( 1 ) weymeri and gustavi are evi- dently conspecific with cydnides and cydno zelinde, since complete intergradation among all of these forms is evident in series taken west of the cordillera northwest of Cali where low passes permit them to meet and mix (some forms illus- trated in Holzinger & Holzinger, 1968) ; and, (2) the intergradation of c. cydno and hermogenes can be seen in many intermediate specimens known from the middle Magdalena valley, and hermogenes apparently meets and intergrades with weymeri in select areas of the central cor- dillera. Thus, present evidence suggests the existence of but a single, if highly variable, spe- cies, cydno, in this complex. An additional member of Emsley’s sara- group, H. hygiana, is evidently interfertile with H. clysonimus. A polymorphic population exists northwest of Cali, Colombia, at high altitudes on the Pacific slope of the western cordillera, which includes occasional specimens of near- typical clysonymus and hygiana, a number of intermediates in color and pattern, and several unique endemic forms as well; morphologically, members of this population are nearer hygiana, but intermediate characters can be seen. It is probable that these two species, which have identical and quite singular field behavior among members of the genus, should be combined in spite of their appreciable morphological differ- ences (Emsley, 1965; Holzinger and Holzinger, 1970). H. hygiana occurs from central-western Colombia through western Ecuador, at moder- ate to high elevations; H. clysonimus is known from similar elevations from Costa Rica to east- ern Venezuela and southern Ecuador, but is sparse in central Colombia. It has been found on the inner face of the western cordillera near Cali, and in eastern Narino; in these areas, where it can cross the western cordillera through passes below 2000 meters, it can meet and ap- parently occasionally interbreed with hygiana. The two are perhaps best regarded as “semi- species,” very closely related in an evolutionary sense and not yet with perfect reproductive iso- lation in spite of long and nearly complete geo- graphic isolation. For more details on the inter- mediate population northwest of Cali, see Holzinger and Holzinger, 1970. 12 New York Zoological Society: Zoologica, Spring, 1972 on the ventral surface of the hindwing. In par- ticular, there are a yellow streak under vein Sc-Rj shared in the genus only by elevatiis, and part of the unusual U-shaped red-brown bar of cydno (Text-fig. 7). This female was dis- sected; the bursa copulatrix has signa (Text- fig. 8) placing the species clearly within the tuelpomene-gxou^ (Emsley, 1965). The meta- pretarsus (Text-fig. 9) has paronychial proc- esses nearly equal in length, and the abdominal processes (Text-fig. 10) are narrow, strongly curved at the base, and recurved near the outer tip, further confirming the placement of the spe- cies near cydno and elevatiis in the nielpomene- group. In January 1970, K. B. examined the type- series of hiciana (two pairs) in the Facultad de Agronomia, Universidad Central de Venezuela, Maracay, courtesy of Dr. Francisco Fernandez Yepez of the Facultad. Although dissection of a male was not performed, the tip of the valve was examined under a 80x microscope and proved to be typically silvaniform, very similar to that of H. elevatiis, but not like the abbrevi- ated tip of the valve in cydno. A most unusual series of hiciana was taken by Mr. Harold Skinner of La Victoria, Vene- zuela, in April 1968 at Mantecal on the upper Rio Cuchivero in Bolivar, Venezuela, well north of the type-locality of the species. No two speci- mens of this series are alike; included are the typical white-banded form, a variety of yellow- banded forms with very variable forewing band shape and spots, and variable markings at the base of the hindwing, and one specimen which even has long yellow rays on the hindwing. We illustrate on Plate III (figs. 12-15), in addition to the type-series of Heliconiiis hiciana, eight specimens from this series taken by Mr. Skinner. In early 1970, a party of six collectors, in- cluding Mr. Skinner and Dr. Fernandez Yepez, returned to Mantecal and captured a further fourteen hiciana, all yellow-banded. According to Dr. Fernandez Yepez, the species flies quite high above the ground and is difficult to capture except when it descends to flowers. Sr. Fran- cisco Romero R., another member of the party, described the males as flying at more than ten meters of height above the ground, descending only occasionally to chase other passing Heli- conius or species with similar flight or appear- ance. In February 1970, K. B. was privileged to obtain through the kindness of Mr. Skinner a single male (Plate III, fig. 15) from the Man- tecal series of hiciana. The genital valves of this specimen (Text-fig. 11) are very close to those of elevatiis, but show a somewhat less elongated Text-figure 7. Heliconiiis hiciana, paratype female in the Museu Nacional, Rio de Janeiro, from Boa Vista, Roraima, hindwing, ventral, schematic. 13 Brown & Mielke: Heticonians of Brazil, Part II Text-figure 8. Heliconius liiciana, same female, bursa copulatrix. Text-figure 9. Heliconius luciana, same female, metapretarsus. Text-figure 10. Heliconius luciana, same female, abdominal process. 14 New York Zoological Society: Zoologica, Spring, 1972 general form, a narrower dorsal process, and no apical tuft of bristles regarded as typical in elevatus genitalia (Turner, 1967b). However, as these characters have all been shown to be vari- able in the silvaniform valves, though the thick- ness of the dorsal process is usually reliable, we also compared female genitalia of the two species. The form of the signa on the bursae copulatrices is indistinguishable in the two, but elevatus possesses a vulvar plate markedly more lobed at the corners, and abdominal processes (Text-fig. 12) thicker, less curved at the base, and less recurved near the tip than those of luciana. These minor morphological differences between the two species are accompanied by significant differences in both major and minor elements of color-pattern between luciana and the nearest races of elevatus {tumaturnari and roraima'. Turner, 1967b, and Masters, 1969). H. luciana, in contrast to elevatus, possesses in- clined rather than vertically arranged elements in the fore wing subapical band; no dennis, but a yellow hindwing bar; usually a yellow stripe along the forewing cubitus; usually several red basal dots on the ventral hindwing; small or absent postcellular yellow elements on the fore- wing; and in most specimens a light submarginal spot in forewing space Cui-Cu2, an element which we have found significant in correlation of the silvaniforms (see Part V). All of these facts lead us to regard luciana as specifically distinct from elevatus. However, at least one yellow-banded luciana has been taken at San Juan de Manapiare, 100 Km southwest of Man- tecal (map), which could easily be interpreted as an intermediate between typical H. luciana from farther south and H. elevatus roraima from farther east. Both species are presently so little-known that they have not been found flying in the same area; luciana has been found in Venezuela to the west of areas occupied by elevatus, and the latter species has not yet been captured in Roraima, Brazil. Thus, until more collecting in intermediate areas or interbreeding can be performed, we cannot completely elimi- nate the possibility of luciana being conspecific with elevatus. Further specimens of luciana cap- tured in 1971 in central Venezuela and Boa Vista conform to previous patterns, not casting new light on the problem; a population was dis- covered in Bolivar with equal representation of yellow and white-banded individuals. It is of considerable interest that these five parallel species to melpomene within its same group {H. timareta, heurippa, elevatus, luciana, and besckei) apparently retain yellow as a court- ship-release color, while red is distinctly the important color in melpomene (Emsley, 1964, 1970; Crane, 1957). The first three of the spe- cies have bright yellow forewing bands in all known forms, and maintain these, in the first with complete suppression of red (in the nomi- nate form), in spite of being sympatric with red-banded forms of melpomene. The rare luciana, also sympatric with red-banded erato and melpomene in all its known localities, exists in yellow-banded and white-banded morphs. Either color is probably equally effective in courtship release, as they have similar reflect- ance and are interchangeable in many silvani- form heliconians such as ismenius and hecale Text-figure 11. Heliconiiis luciana, male from Mantecal, Rio Cuchivero, Bolivar, Venezuela, H. Skinner leg., tip of genital valve (at right), compared with valve tip of Heliconius elevatus tumaturnari (at left), from north of Obidos, Para (the latter has the dorsal process, normally curved inward toward the dorsal midline, straightened out for comparison). Brown & Mielke: Heliconians of Brazil, Part II 15 (Part V). H. besckei, as mentioned above, seems to respond socially to yellow but not to red (Emsley, 1970). Yellow is presumably a more ancient color (Emsley, 1964), typical of the genus Heliconiiis and present in all of its members (chemical composition 3-hydroxy-L- kynurenine; K. Brown, 1967, and Brown and Domingues, 1970). This color predominates in the male of the most primitive Heliconiiis, H. nattereri. Thus, these five species parallel to melpomene may have been “left behind” in an evolutionary sense when melpomene appeared as a widespread and red-responding species, or they may have developed independently from the more primitive silvaniforms, which at least the last three resemble morphologically more than they do melpomene. The Silvana-GKOvv in Extra-Amazonian Brazil Our use of silvana and ethilla as species names, and the former as a group name for the silvaniforms of older authors, rests on data which is detailed in the fifth part of this series. With respect to the latter, a cross of narcaea from Rio de Janeiro with Trinidadian ethilla revealed good fertility in the offspring of the Fo backcross to the latter, thus confirming their conspecificity as suggested by morphological studies (Emsley, 1965). The polymorphism of ethilla narcaea in Rio was clarified by rearing eggs obtained from a female of the rare dark form satis; five adults, including three narcaea and two satis, were ob- tained from seven eggs. We thus believe that satis is merely a dark color-variant with a single gene or closely linked genes controlling its three constant color-pattern differences from narcaea (Al-locus? — Turner, 1968 and 1971, and Shep- pard, 1963). In the cooler interior of Brazil, narcaea locally intergrades smoothly to the striking form polychrous, which has an excess of yellow on both wings, with almost complete suppression of orange. A very few areas are known where polychrous is nearly monomorphic, but it usu- ally flies together with narcaea and interbreeds freely with it. Kaye (1917) mentioned the existence of an unusual brand on the inner margin of the ventral surface of the forewing of male robigus and ethra, absent in Amazonian silvana, and on this basis, coupled with the extremely elongated genital valves of the first two forms, separated these from silvana. We have verified in the col- lection of the Museu Nacional that this brand is present in all ethra and robigus, and also, to a varying degree, in over half of all Amazonian silvana. The form of the genital valve in Ama- zonian silvana is also extremely variable, fre- quently being as elongate as those of the south- ern subspecies. Both the brand and the genitalia are variable characters in many subspecies of H. numata, hecale, and ethilla. The caterpillar and chrysalis of some ethra are noticeably dif- ferent from those of Amazonian and Bolivian silvana, but these differences do not surpass those observed in geographically isolated sub- species of other Heliconiiis. In spite of the es- sentially complete geographical isolation of ethra from silvana, we regard the possibility of reproductive isolation between the two as very small, and thus maintain them for the present as a single species. Recent breeding results have suggested that, in spite of many constant differences in pattern, behavior, and early stages, the silvaniform spe- cies numata and silvana may interbreed freely in some areas of the Amazon Basin, where they are sympatric and common over nearly six mil- lion square kilometers. In this paper, and until more extensive field and insectary experiments can be completed, tbe two species are still main- tained as distinct. Text-figure 12. Heliconiiis elevatus tiimatumari, female, Obidos, abdominal process. 16 New York Zoological Society: Zoologica, Spring, 1972 An Explanatory Note on Materials AND Methods We have based our conclusions on examina- tion and study in vitro, with standard biological dissection methods, of all known heliconians, and extensive field experience with 59 of the 66 species recognized in the tribe. A biological rather than narrowly morphological definition of the species is advanced and, in the systematic ordering of these species, considerable weight has been placed upon in loco observations of adult behavior and micro-sympatry, and on characters of the early stages where these are known. Field observations and breeding were realized in Panama, Jamaica, Colombia, Vene- zuela, Trinidad, Guyana, Ecuador, Peru, Bo- livia, and all areas of Brazil except the extreme northeast and the upper Rio Negro. Dr. Wood- ruff W. Benson also contributed additional field information from Costa Rica and Guyana. Complete Heliconiini collections were examined in the Museu Nacional in Rio de Janeiro; the Facultad de Agronomia in Maracay, Venezuela; the Allyn Collection (including the W. J. Kaye collection) in Sarasota, Florida; the Carnegie Museum in Pittsburgh, Pennsylvania; the Cor- nell University Entomology Department in Ithaca, New York; and the U.S. National Mu- seum in Washington, D.C. Partially complete collections were studied of each of the authors, and of the Departamento de Zoologia in Curi- tiba, the Museu Goeldi in Belem, the Institute Oswaldo Cruz in Rio, the Universidade Federal Rural of Rio de Janeiro, P. Gagarin and the late R. F. d’ Almeida in Rio, L. W. Harris in Lima, E. W. Schmidt-Mumm in Bogota, G. Small in the Canal Zone, F. Romero and H. Skinner in Venezuela, and W. Benson from Costa Rica and Guyana, among others. Further discussion of methodology is presented in Part III. Summary 1. The taxonomy, variation, and zoogeogra- phy of heliconians are discussed, with particu- lar reference to the 18 species regularly occur- ring in extra-Amazonian Brazil. 2. Significant cyclical annual variations in the abundance of species are noted, especially in the more' subtropical areas and on the margins of the Amazon Basin; some possible mechanisms for these variations are discussed. 3. The following systematic changes are sug- gested in the tribe Heliconiini as defined and re- vised by Emsley (1963, 1964, 1965), based upon morphological study of museum speci- mens, field observation, breeding experiments, and proof of gross sympatry over large areas with or without evident intergradation: a. Philaethria wernickei, and its Amazonian subspecies P. w. pygmalion, are separated from P. dido. b. Agraulis lucina is separated from A. va- nillae] the apparently transitional form A. V. catella may result from occasional hybridization, but appears to be true vanillae. c. Eueides lampeto is separated from E. vi- bilia; the relationship of the latter to E. pavana is discussed. d. Heliconius astraea is separated from H. egeria; where the two are sympatric, they are often nearly indistinguishable in color- pattern, but differ morphologically. e. Heliconius heurippa, H. timareta and its forms, and H. besckei are separated from H. melpomene. f. Heliconius luciana is added to the list of species, and fully discussed; its relation- ship to H. elevatus, still not perfectly de- fined, is explained. g. Heliconius eleuchia with its subspecies primularis, and probably with eleusinus and its yellow morph ceres, are separated from H. sapho. Heliconius congener is also regarded as separate from H. sapho. h. The species H. hygiana and H. clysonymus appear to be interfertile where they occa- sionally meet in western Colombia; the two are combined to form a single spe- cies clysonymus, though hygiana may be best regarded as a nearly isolated semi- species. 4. The probable use of yellow as a courtship- release color by five Heliconius species, closely parallel to the red-responding H. melpomene, is suggested. 5. A complete synopsis of known and hypo- thetical extra-Amazonian heliconian species in Brazil is presented, with behavioral and food- plant data and geographical distribution given for each species (Appendix I). 6. A brief synopsis of the heliconian species known from Amazonian Brazil is presented (Appendix II) ; the approximate divisions of the Amazonian subspecies of erato are defined, with indication of hybridization zones (map). 7. A brief summary of the systematic con- clusions of this paper and of Part V (on the silvaniforms) , which brings the number of spe- cies recognized in the tribe Heliconiini up to 66, is presented, with indication of the systematic problems still imperfectly resolved in the tribe (Appendix III). Brown & Mielke: Heliconians of Brazil, Part II 17 Acknowledgments The authors are indebted to the Museu Na- cional, Rio de Janeiro, and especially to Prof. Alfredo Rei do Rego Barros, for assistance and encouragement as well as permission to work intensively in the very large Heliconius collec- tion under his care. Grateful acknowledgment for financial assistance (to K. B.) is also due to the National Science Foundation (U.S.A.), grants numbers GB 5389X and GB 5389X1, the Brazilian Banco Nacional de Desenvolvi- mento Economico, and the Conselho de Pesqui- sas e Ensino para Graduados of the U.F.R.J.; and (to K.B. and O.M.) the Brazilian Conselho Nacional de Pesquisas, of which both are fel- lows. Many persons freely contributed data to the development and refinement of this paper; special mention should be made of the contribu- tions of Dr. Michael G. Emsley (Fairfax, Vir- ginia), Dr. John R. G. Turner (York; England) , Dr. Heinz Ebert and his son Karl Ebert (Rio Claro, Sao Paulo), Mrs. Jocelyn Crane Griffin (New York City and Trinidad, W.I.), Dr. Lee D. Miller of the Allyn Museum of Entomology (Sarasota, Florida), Dr. E. W. Schmidt-Mumm (Bogota, Colombia), Dr. Tarsicio Escalante (Mexico), Dr. Francisco Fernandez Yepez and Sr. Francisco Romero R. (Maracay, Venezuela), Mr. Harold Skinner A. (La Victoria, Vene- zuela), Mr. Gordon Small (Balboa, Panama Canal Zone), Prof. P. M. Sheppard, F. R. S. (Liverpool, England), Dr. H. K. Clench and the late Dr. R. M. Fox (Carnegie Museum, Pitts- burgh, Pennsylvania), Dr. J. G. Franclemont (Cornell University, Ithaca, New York), Dr. H. Holzinger (Vienna, Austria), Dr. T. C. Em- mel (Gainesville, Florida), Sr. Jorge Kesselring (Joao Pessoa, Paraiba), Dr. C. M. Biezanko (Pelotas, Rio Grande do Sul), Dr. Dmitro Zajciw (Rio), and Sr. Celio A. A. Domingues (Rio). Dr. Woodruff W. Benson of Chicago, Illinois, now with the C.P.P.N. in Rio, con- tributed greatly in discussions and corrections of the manuscript at various stages. All drawings and photographs are by K. S. Brown, Jr., with enlargements prepared by Dr. Jose Antonio Pires Ferreira, with the exception of Plate I, figs. 1 and 2 (by Olaf Mielke). Literature Cited Alexander, A. J. 1961a. A study of the biology and behavior of the caterpillars, pupae and emerging butter- flies of the subfamily Heliconiinae in Tri- nidad, West Indies. Part I. Some aspects of larval behavior. Zoologica, 46; 1-26. 1961b. Part II. Molting, and the behavior of pupae and emerging adults. Zoologica, 46; 105-122. Baust, j. G. 1967. Preliminary studies on the isolation of pterins from the wings of heliconiid but- terflies. Zoologica, 52; 15-20. Beebe, W. 1955. Polymorphism in reared broods of Heli- coniiis butterflies from Surinam and Trini- dad. Zoologica, 40; 139-143. Beebe, W., J. Crane, and H. Fleming 1960. A comparison of eggs, larvae, and pupae in fourteen species of heliconiine butter- flies from Trinidad, West Indies. Zoo- logica, 45; 11 1-154. Biezanko, C. M. 1969. Letter to Keith S. Brown, Jr., including manuscripts awaiting publication; Heli- coniidae da Zona Sueste do Rio Grande do Sul; Heliconiidae da Zona das Mis- soes do Rio Grande do Sul; (with A. Ruf- finelli ) Contribuigao ao conhecimento de lepidopteros do Estado de Santa Catarina. Brower, L. P., J. V. Z. Brower, and C. T. Collins 1963. Experimental studies of mimicry. 7. Rela- tive palatability and Mullerian mimicry among neotropical butterflies of the sub- family Heliconiinae. Zoologica, 48; 65-84. Brown, K. S., Jr. 1965. A new L-a-aminoacid from Lepidoptera. Journ. Amer. Chem. Soc., 87; 4202. 1967. Chemotaxonomy and chemomimicry; the case of 3-hydroxy-kynurenine. Systematic Zool., 16; 213-216. 1970. Rediscovery of Heliconius nattereri in eastern Brazil. [The Heliconians of Brazil (Lepidoptera; Nymphalidae). Part I]. Entomol. News (Philadelphia), 81; 129- MO. Brown, K. S., Jr., and C. A. A. Domingues 1970. A distribuigao do amino-acido 3-hidroxi- L-quinurenina nos Lepidopteros. Anajs Acad. Bras. Ciencias, 42; Suplemento, 211-215. Crane, J. 1954. Spectral reflectance characteristics of but- terflies (Lepidoptera) from Trinidad, B.W.I. Zoologica, 39; 85-115. 1955. Imaginal behavior of a Trinidad butterfly, Heliconius erato hydara Hewitson, with special reference to the social use of color. Zoologica, 40; 167-196. 1957. Imaginal behavior in butterflies of the family Heliconiidae; changing social pat- terns and irrelevant actions. Zoologica; 42; 135-145. Crane, J., and H. Eleming 1953. Construction and operation of butterfly insectaries in the tropics. Zoologica, 38; 161-172. 18 New York Zoological Society: Zoologica, Spring, 1972 Ebert, H. 1970. On the frequency of butterflies in eastern Brazil, with a list of the butterfly fauna of Pogos de Caldas, Minas Gerais. Joum. Lepidopterists’ Soc., 23 (1969), Supple- ment 3, 1-48. Emsley, M. G. 1963. A morphological study of imagine Heli- coniinae (Lep.: Nymphalidae), with a consideration of the evolutionary relation- ships within the group. Zoologica, 48: 85-130. 1964. The geographical distribution of the color- pattern components of Heliconius erato and Heliconius melpomene, with geneti- cal evidence for a systematical relation- ship between the two species. Zoologica, 49: 245-286. 1965. Speciation in Heliconius (Lep.: Nympha- lidae): morphology and geographic dis- tribution. Zoologica, 50: 191-254. 1970. An observation on the use of color for species-recognition in Heliconius besckei (Nymphalidae). Journ. Lepidopterists’ Soc., 24: 25. Fleming, H. 1960. The first instar larvae of the Heliconiinae (butterflies) of Trinidad, W.I. Zoologica, 45: 91-110. Fox, R. M. 1967. A monograph of the Ithomiinae (Lepi- doptera), Part III. The tribe Mechanitini Fox. Mem. Amer. Entomol. Soc., number 22. Gilbert, L. E. 1969. The biology of natural dispersal: Dione moneta poeyii in Texas (Nymphalidae). Journ. Lepidopterists’ Soc., 23: 177-185. Hayward, K. J. 1951. Catalogo sinonimico de los rhopaloceros Argentinos, excluyendo “Hesperiidae.” Acta Zool. Lilloana, 9: 85-281. Holzinger, H., and R. Holzinger 1968. Heliconius cydno gerstneri, n. ssp. und zwei neue Formen von H. cydno cydnides STGR. (Lep. Nymph.). Zeitsch. der Arbeitsgemeinschaft dsterr. Entomologen, 20: 17-21. 1970. Heliconius hygianus fischeri (FASSL) comb, nov., eine Subspecies aus West- Colombien (Lep. Nymph.). Zeitsch. der Arbeitsgemeinschaft dsterr. Entomologen, 22: 33-41. JoiCEY, J. J., AND G. Talbot 1925. Notes on some Lepidoptera with descrip- tion of new forms. Ann. Mag. Nat. Hist. (9), 16: 633-653. Kaye, W. J. 1917. A reply to Dr. Eltringham’s paper on the genus Heliconius. Trans. Ent. Soc. Lon- don, 1916: 149-155. DE Lesse, H. 1967. Les nombres de chromosomes chez les Lepidopteres Rhopaloceres neotropicaux. Ann. Soc. Ent. France (N.S.), 3: 67-136. Lichy, R. 1960. Documentos para servir al estudio de los lepidopteros de Venezuela. IV. Una espe- cie nueva del genero Heliconius Kluk (Rhopalocera, Nymphalidae): Heliconius luciana sp. nov. Rev. Fac. Agronomia, Univ. Central Venez. (Maracay), 2: 20-44. Masters, J. H. 1969. Heliconius liecale and xantliocles in Vene- zuela. Journ. Lepidopterists’ Soc., 23: 104-105. Sheppard, P. M. 1963. Some genetic studies of Mullerian mimics in butterflies of the genus Heliconius. Zoologica, 48: 145-154. SWIHART, S. L. 1963. The electroretinogram of Heliconius erato (Lepidoptera) and its possible relation to established behavior patterns. Zoologica, 48: 155-164. 1964. The nature of the electroretinogram of a tropical butterfly. Journ. Insect Physiol., 10: 547-562. 1965. Evoked potentials in the visual pathway of Heliconius erato (Lepidoptera). Zoo- logica, 50: 55-60. 1967a. Neural adaptations in the visual pathway of certain Heliconiine butterflies, and re- lated forms, to variations in wing colora- tion. Zoologica, 52: 1-14. 1967b. Maturation of the visual mechanisms in the neotropical butterfly, Heliconius sarae. Journ. Insect Physiol., 13: 1679-1688. 1968. Single unit activity in the visual pathway of the butterfly Heliconius erato. Journ. Insect Physiol., 14: 1589-1601. Talbot, G. 1928. List of Rhopalocara collected by Mr. C. L. Collenette in Mato Grosso, Brazil. Bull. Hill Mus., Witley, 2: 192-220. Tokuyama, T., S. Senoh, T. Sakan, K. S. Brown, Jr., and B. Witkop 1967. The photo reduction of kynurenic acid to kynurenine yellow, and the occurrence of 3-hydroxy-L-kynurenine in butterflies. Journ. Amer. Chem. Soc., 89: 1017-1021. Brown & Mielke: Heliconians of Brazil, Part II 19 Turner, J. R. G. 1966. A rare mimetic Heliconius (Lepidoptera: Nymphalidae). Proc. Royal Ent. Soc., London (B), 35: 128-132. 1967a. Some early works on heliconiine butter- flies and their biology (Lepidoptera, Nym- phalidae). J. Linn. Soc. (Zool.), 46: 255- 265. 1967b. A little-recognized species of Heliconius butterfly (Nymphalidae). Journ. Research Lepid., 5: 97-112. 1967c. The generic name of Papilio iulia Fabri- cius, sometimes called the Flambeau (Lepidoptera, Nymphalidae). The Ento- mologist, 100: 8. 1967d. Goddess changes sex, or the gender game. Systematic Zool., 16: 349-350. 1968a. Natural selection for and against a poly- morphism which interacts with sex. Evo- lution, 22: 481-495. 1970. Mimicry: a study in behavior, genetics, ecology and biochemistry. Science Prog- ress (Oxford), 58: 219-235. 1971. The genetics of some polymorphic forms of the butterflies Heliconius melpomene (Linnaeus) and H. erato (Linnaeus). II. The hybridization of subspecies of H. mel- pomene from Surinam and Trinidad. Zoo- logica, 56: 125-157. Turner, J. R. G., and J. Crane 1962. The genetics of some polymorphic forms of the butterflies Heliconius melpomene Linnaeus and Heliconius erato Linnaeus. I. Major genes. Zoologica, 47: 141-152. APPENDIX I A Synopsis of the Heliconians of Extra-Amazonian Brazil In polymorphic species, only significant color- morphs with well-established names are included in this synopsis. Ranges are given for extra-Amazonian Brazil only. Abbreviations used for states are those standardized for use in Brazil: BA, Bahia; CE, Ceara; DF, Distrito Federal; ES, Espirito Santo; GO, Goias; GB, Guanabara (city of Rio de Janeiro, formerly the Distrito Federal before the creation of Brasilia); MA, Maranhao; MG, Minas Gerais; MT, Mato Grosso; PB, Paraiba; PR, Parana; PE, Pernambuco; RJ, Rio de Ja- neiro; RS, Rio Grande do Sul; RN, Rio Grande do Norte; SC, Santa Catarina; SP, Sao Paulo. See map. Indications of preferred flowers are : R = red {Lantana, Gurania, red Bidens, Passiflora coc- cinea, Poinsettia, Bromeliaceae ) ; M = magenta {Passiflora kermesina)\ B = blue (Stachytar- pheta, many Eupatorium) ■, Y = yellow (Oxy- petalum, yellow Bidens, many Compositae); W = white (many Eupatorium, Orchidaceae, Compositae). The larval food-plants represent a very pre- liminary list. The passifloraceous species have been identified by the authors, by Dr. W. W. Benson, by Dr. C. M. Biezanko, and by Appa- ricio P. Duarte, following Killip’s recent revision and Masters in Martius, Flora Brasiliensis, and more recent studies in Brazil, especially by Sacco; and by Dr. Stephen S. Tillett of Barqui- simeto, Venezuela, following Killip and his own investigations. Tentative identifications are in- dicated with a question mark; several of the species used as food-plants by Brazilian helico- nians are undoubtedly new. All the information on food-plants for Santa Catarina and Rio Grande do Sul is taken, with permission, from Biezanko, 1969. Food-plant records are based on field observations, usually on several occa- sions and in different areas, of feeding larvae and ovipositing females. For earlier works on the immature stages of Brazilian heliconians, see Turner ( 1967a). A. Normally Extra-Amazonian Heliconians Philaethria Billberg, 1820. dido (Linne, 1763) (Plate I, fig. 1). Eastern coastal lowlands in forest at least from PB to RJ ; not certain if present in the extreme south or in the interior (possibly marginal in central MT). Very localized and rather rare south of ES. Hilltops, though not strongly. Flowers W, Y, B, rarely R. Cater- pillars solitary: Passiflora mucronata (ES) ; refused P. alata, P. speciosa (whose close relative P. vitifolia is accepted in Colombia and Panama), P. violacea, P. jileki, and Tetrastylis ovalis (ES). 20 New York Zoological Society: Zoologica, Spring, 1972 wernickei wernickei (Rober, 1906) (Plate I, fig. 2, and Plate IV, fig. 16). Entire area in forest and on edges, commoner southward (but rare in coastal RS), very rare in cen- tral plateau except in central MT (where flies in cerrado near gallery forests) ; usu- ally quite localized. Intergrades perceptibly to w. pygmalion (Fruhstorfer, 1912) northward. Hilltops. Flowers W, Y, B, R. Caterpillars solitary: Passiflora sidae folia (GB, RJ), P. coerulea (SC, RS), P. siib- erosa (RS), P. elegans (RS), P. mansii (MT). Dryadula Michener, 1942. phaetusa (Linne, 1758) (Plate IV, figs. 20 and 21). Entire area in open country (fields, marshes, and scrublands), strongly localized but common where found. Flow- ers W, Y, R. Caterpillars solitary: Passi- flora mucronata (GB), P. misera (BA), P. mansii (MT). Agraulis Boisduval and LeConte, 1833. vanillae (Linne, 1758) maculosa (Stichel, 1907) (Plate IV, figs. 17 and 18). Entire area, common to abundant, in open coun- try only or in large cultivated areas within the forest. Flowers R, B, W, Y. Caterpillars solitary but tolerant: Passiflora ichthyura (ES), P. mucronata (GB), P. edulis (GB, BA, DF), P. odontophylla (?) (ES), P. kermesina (ES), P. speciosa (ES), P. vio- lacea (ES), P. quadrangularis (GB, ES), P. coerulea (SC, RS), P. mansii (MT). Dione Hiibner, 1819. juno juno (Cramer, 1779) (Plate IV, fig. 19). Entire area in forest clearings, but very rare in interior plateau; quite localized but common where encountered. Hilltops. Flowers R. Caterpillars strongly gregarious with coordinated behavior: Passiflora edu- lis (GB, RJ, BA), P. alata (GB, ES), P. speciosa (ES), P. odontophylla (?) (ES), P. coerulea (SC, RS). juno suffumata Hayward, 1931. Isolated pop- ulations in the Brasilia area; to he expected elsewhere in south-central Brazil (described from Paraguay). Both fore- and hindwings heavily suffused with black from margins inward; some specimens in populations in central MT tend towards this suffusion. Flowers R, B. Caterpillars gregarious: Passiflora cornuta (DF), P. alata (DF). moneta moneta Hiibner, 1825 (Plate IV, fig. 22). Southwestern MG, western SP and PR, and SC and RG, common in late sum- mer and fall in open country and cleared areas within the forest. Perches in after- noon, and roosts at night on projecting tips of grass in open areas. Flowers R, W, Y. Caterpillars solitary: Passiflora violacea (SP). Dryas Hiibner, 1807. juno juno (Cramer, 1779) (Plate IV, fig. 23 and 24). Entire area in all habitats, less common in interior plateau. Flowers R, W, Y, B. Caterpillars solitary and canni- balistic: Passiflora organensis (GB), P. truncata (GB, SP), P. sidae folia (GB), P. misera (BA), P. capsularis (GB, SP), P. quadrangularis (RS), P. coerulea (RS), P. edulis (GB, RS). Eueides HuhmT, 1816. aliphera aliphera (Godart, 1819) (Plate IV, fig. 28). Entire area, strongly localized, almost always encountered very near its food-plant. Flowers W, Y. Caterpillars solitary but tolerant: Passiflora violacea (GB, RJ, SP, MG, ES, BA, DF) , P. coeru- lea (RS), P. quadrangularis (RS), P. cocci- nea (MT), P. sidae folia (GB). vibilia vibilia (Godart, 1819) (Plate I, fig. 5). Very local on coastal plain in deep forest, up to moderate elevations in the coastal mountains and river-valleys, south at least to PR. Observed in apparent unidirectional migration in late summer in ES, from the valley of the Rio Doce SE over the moun- tains towards the coast. Flowers W, Y, B. Caterpillars gregarious: Passiflora odonto- phylla (?) (ES). pavana Menetries, 1857 (Plate I, fig. 5). Lo- cally common in forest in the Serra do Mar and Serra da Mantiqueira at 600 to 1500 meters elevation, from central ES and cen- tral MG south to SC, also locally down to foothills and outwash plains at sea level. Flowers W, Y. Caterpillars solitary: Passi- flora sidae folia (RJ). isabella (Cramer, 1781-2) dianasa (Hiibner, 1806) (Plate IV, fig. 25). Entire area in forest, uncommon. Becomes plastic north- ward, showing an increasing percentage of yellow and/or divided subapical elements on the forewing (intergradation to i. isa- bella). Hilltops. Flowers W, Y, rarely R, B. Caterpillars solitary but tolerant: Passi- flora edulis (GB, RJ), P. alata (GB), P. odontophylla (?) (ES). Heliconius Kluk, 1802 (1780?). nattereri C. and R. Felder, 1865 (illustrated in Part I and Part III of this series). Very rare and local in large tracts of virgin for- est in the “Amazonian island” of lowland BA, ES, and possibly eastern MG. Prefers steep, humid areas where its foodplant is giving abundant fresh growth. Female = Brown & Mielke: Heliconians of Brazil, Part 11 21 fruhstorferi Riffarth, 1898. Flowers R, M, B, rarely W. Caterpillars solitary but toler- ant; Tetrastylis ovalis (ES). silvana (Cramer, 1781) ethra (Hiibner, 1827- 31) (Plate V, figs. 34-37). Coastal belt from PE south to central ES in deep pri- mary forest, rarely up river valleys and foothill canyons to 900 meters elevation. Includes form brasiliensis Neustetter, 1907, and many additional minor varieties. Quite localized. Flowers R, W. Caterpillars soli- tary; Tetrastylis ovalis (BA). silvana robigus Weymer, 1875 (Plate V, fig. 38). Coastal belt from southern BA (over- lapping with ethra) to SC (occasional) in deep forest, rarely up to 900 meters eleva- tion in foothill canyons, very rare and lo- calized southward. Flowers R, W. Cater- pillars solitary; Passiflora alata (GB), P. sidaefolia (GB), P. rhamnifolia (GB). ethilla Godart, 1819 narcaea Godart, 1819 (Plate V, figs. 40-43, 45, 46). Entire area in many habitats (but prefers forest), north to southern BA and DF, northwest to eastern MT, and south to western RS. Form sutA Weymer, 1884 (fig. 45) appears very rarely in all populations (commoner locally, up to five percent of populations, in ES, MG, RJ, and GB). Form polychrous C. and R. Felder, 1865 (figs. 43 and 46) is commoner in the central plateau and in SP, predominant in some populations west- ward. Ridgetops. Flowers R, M, B, rarely W, Y. Caterpillars solitary; Passiflora si- daefolia (GB), P. alata (GB, DF), P. ker- inesina (GB, ES), P. jileki (ES), P. rham- nifolia (GB), P. nitida (DF), P. corniita (DF), Tetrastylis ovalis (ES). ethilla flavomaculatus Weymer, 1894 (Plate V, fig. 39). Coastal belt from PB south to southern BA. Flowers R. Caterpillars soli- tary; Passiflora recurva (PE), P. kermesina (PE). besckei Menetries, 1857 (Plate V, fig. 47). Mountains in forest above 700 meters ele- vation from central ES (possibly BA and PE, locally), northern GO, and southern MT (at lower elevations) south to western RS, also occasionally down foothill can- yons to outwash plains at sea level. Flowers R, B, rarely W, Y. Caterpillars solitary to semi-gregarious, tolerant; Passiflora sidae- folia (RJ), P. villosa (RJ), P. coerulea (SC), P. organensis (RJ). melpomene (Linne, 1758) nanna Stichel, 1899 (Plate V, fig. 48). Eastern coastal belt in forest from RN south to ES, rarely to GB and occasionally to SC, also up mountains and river valleys to 1000 meters elevation. Flowers R. Caterpillars solitary; Passiflora alata (young, soft, shaded plants) (ES), P. misera (ES, BA), P. violacea (BA), Tetrastylis ovalis (ES). erato (Linne, 1758) phyllis (Fabricius, 1775) (Plate V, figs. 50 and 51). Entire area in all habitats, common. Becomes plastic (in- cluding form phyllides) in central MT. Forms artifex and cohaerens appear nor- mally in all populations. Adults roost com- munally at night, in groups of three to 20 individuals. Flowers R, B, rarely W, Y. Caterpillars solitary and cannibalistic; Pas- siflora tnmcata (GB), P. organensis (GB, RJ, ES), P. jileki (ES), P. violacea (BA), P. misera (ES), P. sidaefolia (GB), P. alata (SC, RS, but rejected in GB), P. capsularis (GB, ES, GO), P. coerulea (RS), Tetra- stylis ovalis (ES). Sara (Fabricius, 1793) apseudes (Hubner, 1806) (Plate IV, fig. 32). Coastal plain in Oceanside hammocks, forest, and second growth, from PB to SC, up mountains, common; very sparse in interior of Sao Paulo (Loreto) and MG (Belo Horizonte). Adults roost communally at night in groups of up to 40 individuals. Flowers W, Y, R, B. Caterpillars strongly gregarious; Passi- flora mucronata (GB, ES), P. sidaefolia (GB), P. rhamnifolia (GB), P. edulis (RJ), Tetrastylis ovalis (ES). B. Marginally Extra-Amazonian Heliconians Eiieides Hubner, 1816 vibilia (Godart, 1819) unifasciatus Butler, 1873 (Plate IV, figs. 26 and 27). Marginal, locally abundant in fall and early winter, in forest and scrub in central MT and southwestern GO. Populations include a few percent of v. vibilia and many interme- diates. Flowers W, Y, B. Caterpillars gre- garious with coordinated behavior; Passi- flora mansii (MT). isabella isabella (Cramer, 1781-2) (Plate VI, figs. 52-55). Marginal, local, in central MT; to be expected in MA, CE. Polymorphic. Flowers W, B, R. Heliconius Kluk, 1802 (1780?) aoede (Hubner, 1809-13) manuscript sub- species, K. Brown (see Part IV of this series). One specimen from the Rio Branco, tributary of the Rio Cabagal (Paraguay drainage) in west-central MT, a male taken in the afternoon of a cloudy day flying together with very similar-appearing itho- miines in heavy riparian forest, 400 meters elevation. Flowers R, W. Caterpillars probably gregarious. 22 New York Zoological Society: Zoologica, Spring, 1972 wallacei Reakirt, 1866 fiavescens Weymer, 1890 (Plate IV, fig. 29). Marginal in for- est in central MT, to be expected in MA. One record of form parvimaculata Riffarth, 1900 from SC may be a labeling error. Flowers R, W, B. Caterpillars semi-gregari- ous: Passiflora coccinea (MT); wallacei is closely associated with this species and its very close relatives throughout its range in the Amazon and Orinoco Basins. burneyi (Hubner, 1827-31) near burneyi (Plate VI, fig. 60). One specimen known from Caceres in west-central MT; one male observed for over 30 minutes on high yel- low flowers on the Rio Branco, tributary of the Rio Cabagal, in June 1971. To be expected elsewhere in central MT and in MA. xanthocles Bates, 1862 melete C. and R. Felder, 1865 (Plate IV, fig. 30). Regular in fall and winter in forests by streams in highland central MT. Flowers R. Cater- pillars probably gregarious. silvana (Cramer, 1781) Weymer, 1894 (Plate VI, fig. 59). Marginal, well-estab- lished in west-central Mato Grosso (Rio Branco/ Rio Cabagal, 400 meters). numata (Cramer, 1780-82) siiperioris Butler, 1875 and many forms near this (Plate VI, figs. 56-58). Marginal, common in west- central MT (Rio Branco/ Rio Caba§al, 400 meters); to be expected elsewhere in cen- tral MT. Caterpillars solitary; Passiflora coccinea (MT), P. glandulosa (MT), ac- cepted P. tricuspis (MT). ethilla Godart, 1819 eucoma (Hubner, 1827- 31) (Plate V, fig. 44). Marginal in western CE and southeastern MA (D. Zajciw); may appear in west-central MT (Rio Caba- ?al). ethilla manuscript subspecies, K. Brown (see Part IV). Marginal but regular, frequent in fall, in deep forest near streams in high- land central MT. Males promenade in small clearings. Flowers R, B. Caterpillars soli- tary; Passiflora cornuta (MT), P. glandu- losa (MT). melpomene (Linne, 1758) burchelli PoxxMon, 1910 (Plate V, fig. 49). Borders of Amazon Basin, in forest and cerrado, in MA, CE, GO, DF, and MT; becomes plastic, occa- sionally even with dennis and ray, in cen- tral MT. Flowers R, B. Caterpillars soli- tary: Passiflora cornuta (DF), P. mansii (MT), probably P. tricuspis (MT). ricini (Linne, 1758) (Plate IV, fig. 32). Marginal in MA and CE; possibly marginal but unlikely in central MT (“Cuyaba- Co rumba River System” ) . Sara (Fabricius, 1793) thamar (Hubner, 1806) (Plate IV, fig. 33). Locally com- mon in forest and second growth on the borders of the Amazon Basin in MT, cen- tral and southern GO, DF, and extreme northwestern BA (Rio Sapao). To be ex- pected in MA and CE. Flowers W, Y, R, B. Caterpillars gregarious; Passiflora man- sii (MT). leucadia Bates, 1862 pseudorhea Staudinger, 1896 (Plate VI, fig. 61). Marginal in west- central MT (Rio Branco/ Rio Cabagal and upper Rio Jauru). C. Hypothetically Extra-Amazonian Heliconians The following species are either tenuously recorded from extra- Amazonian Brazil, with no recent and reliable confirmation, or else occur commonly in the indicated areas adjacent to extra-Amazonian Brazil, or have been recently recorded from these areas by reliable authori- ties. None has yet been captured by the authors in the area under consideration. Dione glycera (C. and R. Felder, 1861). Misi- ones, Argentina (Hayward, 1951). Eueides lybia lybia (Fabricius, 1775). Maran- hao, Rondonia; possibly seen on the Rio Branco, MT, in June 1971. Heliconius astraea Staudinger, 1896 manuscript subspecies, K. Brown (see Part IV). Ron- donia and northern MT; “Cuyaba-Corumba River System.” Heliconius doris doris (Linne, 1771) and form delila (Hubner, 1813). Maranhao, northern Mato Grosso and Rondonia. Heliconius numata (Cramer, 1780-82) splen- didus Weymer, 1894. Misiones, Argentina (Hayward, 1951). Heliconius hecale (Fabricius, 1775) sisyphus Salvin, 1871 and variants. Northern Bolivia west of the Rios Cabagal and Jauru. Heliconius elevatus Noldner, 1901 schmass- manni Joicey & Talbot, 1925 {1 = aquilina Neustetter, 1925) and H.e. perchlora Joicey & Kaye, 1917. Rondonia and northern MT; “Cuyaba-Corumba River System.” Heliconius demeter S\.a.\iAingQV, 1895 eratosignis Joicey and Talbot, 1925. Southeastern Ron- donia; “Cuyaba-Corumba River System.” Heliconius antiochus (Linne, 1767) alba Rif- farth, 1900. Northeastern Mato Grosso (com- mon), Maranhao. Brown & Mielke: Heliconians of Brazil, Part If 23 APPENDIX II A Brief List of the Heliconians of Amazonian Brazil The Amazonian region of Brazil is so little- explored that it would be most premature and foolhardy to present a definitive list or a com- plete synopsis of the subspecies at this time. A large network of highways, now under construc- tion and to be finished by the mid-1970s, will permit a far more thorough investigation of Amazonian heliconians by the end of this dec- ade. This list presents only a preliminary tally of the species and ranges known to date, with indications of the principal subspecies present where these are reasonably well defined. Mar- ginally Amazonian species like besckei are omitted. Many Amazonian heliconian populations are noted for their polymorphism. This phenome- non is perhaps most marked in Heliconius numata, discussed in detail along with other Amazonian silvaniforms in Part V of this series. Some examples of polymorphism in Amazonian heliconians are illustrated on Plate VI. The Amazonian area of the map has been divided and patterned according to present in- formation on the interaction of the subspecies of Heliconius erato in the Amazon Basin. Rela- tively monomorphic areas are indicated by pure patterns, blend zones (see Plate VI, figs. 63 and 64) by overlapping patterns; much of the Bra- zilian upper Amazon is a blend zone for three major subspecies, and the named form lativitta Butler 1877 is a typical hybrid from this area which shows signs of influence of all three of these subspecies (amazona, emma, and the re- ductimacula-donatia-venustiis complex). The various color-patterns of erato in the Amazonian area are closely followed by those of the other dennis-rayed heliconians (Eueides tales and eanes, Heliconius aoede, burneyi, egeria, astraea, xanthocles, elevatus, Amazonian melpomene, and demeter). However, some startling excep- tions to this generalized parallelism are known, presumably due to individual diflferences in the genetic mechanisms by which each species achieves the desired patterns. Where erato and melpomene are red-banded in the north-central and southeastern parts of the Amazonian basin, the other dennis-rayed spe- cies may exist in unchanged form. They also may be replaced by closely related species (like the substitution of luciana in the north and besckei in the south for elevatus), or may be absent (as in most of Amazonian Goias). The definition of the blend areas necessarily is ap- proximate until detailed studies can be made along the new roads. Evidence accumulated over 50 years also indicates that both the position and the composition of these hybrid zones is con- stantly changing, in dynamic equilibrium with the monomorphic zones which give rise to them and with natural selection phenomena which vary within them from year to year. Philaethria dido (Linne, 1763) (Plate I, fig. 1). Entire area except dry southeastern Amazon (erato phyllis area), quite frequent in heavy forest and clearings; very high flyer. Philaethria wernickei (Rober, 1906) pygmalion (Fruhstorfer, 1912) (see Plate I, fig. 2). Upper Rio Negro and Uaupes, and southern Rondonia eastward through entire middle and lower Amazon. Dryadula phaetusa (Linne, 1758) (Plate IV, figs. 20-21). Entire area, localized in open or marshy areas. Agraulis vanillae (Linne, 1758) (Plate IV, figs. 17 and 18; Plate I, figs. 3-4). Entire area, but rare and local westward where following species flies. Principally nominate subspecies, except in southern Amazon [v. maculosa (Stichel, 1909)]. Agraulis lucina Felder, 1862 (Plate I, figs. 3-4). Upper Amazon only, from Uaupes, Tefe, and eastern Acre westward, in forest clearings. Dione juno juno (Cramer, 1779) (Plate IV, fig. 19). Entire area though quite localized, in forest clearings; southwestern populations show appreciable variation in dark markings. Dryas iulia iulia (Fabricius, 1775) (Plate IV, figs. 23-24). Entire area, frequent in all habitats. Eueides aliphera aliphera (Godart, 1819) (Plate IV, fig. 28). Entire area, very localized, com- mon along streams. Eueides vibilia (Godart, 1819) (Plate I, fig. 5; Plate IV, figs. 26-27). Nominate subspecies rarely encountered in lower and middle Ama- zon; V. unifasciatus Butler, 1873 locally com- mon in upper Amazon. Intermediates with partial forewing subapical bands are common in populations of both subspecies in the Ama- zon Basin. Eueides lampeto Bates, 1862 (Plate I, fig. 6). Nominate subspecies very local and rare in upper Rio Solimoes (above Tefe) ; 1. copiosus 24 New York Zoological Society: Zoologica, Spring, 1972 Stichel, 1906 recently discovered north of Obidos. Eueides eanes earns Hewitson, 1861. Not rare in extreme western Amazonas and Acre. Eueides Isabella Isabella (Cramer, 1781-2) (Plate VI, figs. 52-55). Entire area but quite local; strongly polymorphic, especially south- westward. Eueides lybia lybia (Fabricius, 1775). Entire area but rarer westward and southward; local in dryer areas and secondary forest, always found very near its food-plant. Eueides tales (Cramer, 1775-6). Locally com- mon in forest and second growth; the rather variable subspecies pythagoras Kirby, 1900 (dennis-ray), tales and surdus Stichel, 1903 (dennis only), aquilifer Stichel, 1903 (con- densed FW yellow patch), and calathus Stichel, 1902 (FW yellow band distal to cell) follow the erato variations indicated in map. Not known outside the Flylaea in the dryer southeastern Amazon (Goias). Heliconiiis metharme (Erichson, 1848). Very local, from western Para northwestward and southwestward to Uaupes, Benjamin Con- stant and western Acre. Heliconiiis aoede (Hiibner, 1809-13). Entire area except dryer southeastern Amazon. Sub- species aoede (dennis-ray), astydamia (Erich- son, 1848) (dennis only), faleria Fruhstorfer, 1910 (partially coagulated FW yellow band), lucretius Weymer, 1890 (condensed FW yel- low patch) and a new subspecies with reduced dennis (see Part IV), and bartletti Druce, 1876 (FW yellow band distal to cell) closely follow erato variations (Map I). Generally uncommon and local, in heavy moist forest. Heliconiiis wallacei Reakirt, 1866 (Plate IV, fig. 29; Plate VI, fig. 62). Entire area, com- mon wherever Passiflora coccinea and re- lated species grow, many habitats. Usually flavescens Weymer, 1890; white-banded forms [clytia (Cramer, 1775-6) and elsa Riffarth, 1899] and w. wallacei are more fre- quent in the northern Amazon; polymorphic populations (colon Weymer, 1890; parvima- culata Riffarth, 1900, and many other forms) occur in the lower middle Amazon. Heliconiiis biirneyi (Hiibner, 1827-31) (Plate VI, fig. 60). Entire area except extreme southeast, rather localized; very high flyer. Subspecies biirneyi (dennis-ray), catherinae Staudinger, 1885-8 (dennis only), ada Neu- stetter, 1925 (partly coagulated FW yellow band with reduced subapical elements), and hiiebneri Staudinger, 1896 (condensed and reduced FW yellow patch and wider HW rays) occupy areas roughly corresponding to erato variations, though much discrepancy from parallelism is seen and the overall varia- tion of burneyi is less (see Map I). Heliconius egeria (Cramer, 1775-6) (Plate III, fig. 11). Rare and local, from Belem west in heavy forest to Uaupes, western Amazonas and northern Rondonia; very high-flyer. The rayed subspecies hyas Weymer, 1884 is a variable element in many populations, pre- dominant in the Rio Madeira region; its northern form with a more compact FW band, asterope Zikan, 1937, is found on the upper Rio Negro. Heliconius astraea Staudinger, 1896 (Plate III, fig. 11). Rare and local in southwestern and extreme western Amazon, in heavy forest or along rivers; habits as in egeria and burneyi. Nominate subspecies in western Amazonas above Tefe; new subspecies (see Part IV) in the Rio Madeira area, south to well beyond the range of egeria hyas. Heliconius xanthocles Bates, 1862 (Plate IV, fig. 30). Entire area except extreme south- eastern Amazon. Subspecies vala Staudinger, 1885-8 (dennis-ray), xanthocles (dennis only), paraplesiiis Bates, 1867 (partly coagulated FW yellow band), melete Felder, 1865 (con- densed FW yellow patch), and melittus Stau- dinger, 1896 (FW band distal to cell) closely follow erato variations (Map I). Heliconius doris (Linne, 1771). Entire area, principally along major rivers. Forms delila (Hiibner, 1813), metharmina Staudinger, 1896, and amathiisiiis (Cramer, 1777) occur in all populations, but are commonest in far western and southwestern Amazon. Essen- tially no green forms have been found in the Brazilian part of the Amazon Basin. This spe- cies has been placed by recent authors in a separate subgenus (Eapariis Billberg, 1820). Heliconius silvana (Cramer, 1781) (Plate VI, fig. 59; Plate V, figs. 34-37). Entire area except southeast Amazon, quite common. Grades smoothly into subspecies minis Wey- mer, 1894 in southern Rondonia; some east- ern (Belem) specimens seem to grade towards ethra (Hiibner 1827-31); far western speci- mens have larger subapical spots on the FW (as does the sympatric numata aurora). Heliconius niimata (Cramer, 1780-2) (Plate VI, figs. 56-58). Entire area except extreme southeast, locally abundant, highly polymor- phic. Forms which may deserve weak sub- specific rank include siiperioris Butler, 1875 (middle Amazon southward), aurora Bates, 1862 (far west), eii phone Felder, 1862 (south- Brown & Mielke: Heliconians of Brazil, Part II 25 west), and zobrysi Fruhstorfer, 1910 (south- east); silvaniformis Joicey and Kaye, 1917 is a strong element in far eastern populations. Some southwestern specimens have a black suffusion on the distal half of the FW as in silvana minis. Dissimilar specimens in which the yellow has been entirely replaced by orange occur in all populations; the extreme of these is arcuella Druce, 1874, commoner westward; the orange form of superioris is isabeUinus Bates, 1862; of zobrysi is seraphion Weymer, 1894. The very dark hindwing of nominate niimata appears in all populations, but is commoner northeastward. Heliconiiis ethilla Godart, 1819 (Plate V, figs. 39-46). The subspecies eucoma (Hlibner, 1827-31) and its dark variety niimismaticus Weymer, 1894 occupy almost the entire area, except for the southeast (e. narcaea Godart, 1819 and its form polychrous Felder, 1865), southwest (nebulosa Kaye, 1916), and south- central Amazon (new subspecies, see Part IV). The Guianian subspecies thielei Riffarth, 1900 may appear in the northeastern and north-central parts of the Amazon Basin. Heliconius hecale (Fabricius, 1775). Entire area except southeastern and south-central Ama- zon. Strongly fragmented into locally differ- entiated populations with apparently rather limited gene-ffow, which may be regarded as good subspecies: novatus Bates, 1867 (Belem; erroneously rechristened schiilzi)', xingiiensis Neustetter, 1925 (lower Rio Xingu); paraen- sis and latus Riffarth, 1900 (Obidos area); vetiistus Butler, 1873 (north of Obidos into Guianian highlands) ; metelliis Weymer, 1894 (near Santarem); fortiinatus Weymer, 1884 (north of Manaus); spuriiis Weymer, 1894 (south and east of Manaus, a minor element as far east as eastern Para) ; sulphureiis Wey- mer, 1894 (Rio Negro); enniits Weymer, 1890 (south and west of Manaus); nigro- fasciatus Weymer, 1894 (Rondonia and Acre); sisyphus Salvin, 1871 and forms con- cors, jonas, etc., Weymer, 1894 (extreme west and southwest); humboldti Neustetter, 1928 (extreme west north of the Solimoes), and probably many more to be discovered. Heliconius pardaiiniis Bates, 1862. Principally from extreme western Amazonas {parda- iiniis) east to Rondonia and Manaus (form lucescens Weymer, 1894 commoner), pos- sibly to Obidos and Santarem; also southwest to Acre {maeon Weymer, 1890 and dilatiis Weymer, 1894). Heliconius elevatus Ndldner, 1862. Entire area except southeast and north-central Amazon, but extremely rare and localized. Subspecies barii Oberthiir, 1902 (dennis-ray), roraima Turner, 1967 and tumatumari Kaye, 1906 (dennis only, the former with a condensed FW yellow patch), aqiiilina Neustetter, 1925 and (or=?) schmassmanni Joicey and Talbot, 1925 (partly coagulated FW yellow elements), perchlora Joicey and Kaye, 1917 (condensed FW yellow patch), and elevatus (FW band mostly distal to cell) follow fairly well the divisions of erato (see Map I). Heliconius luciana Lichy, 1960 (Plate III, ffgs. 12-15). Known only from near Boa Vista in northern Roraima, where sympatric with the very similar and abundant antiochus and un- common wallacei elsa. Heliconius melponiene (Linne, 1758) (Plate II, figs. 8 and 10; Plate V, fig. 49). Entire Ama- zon Basin, locally abundant but often absent from large areas. Subspecies thelxiope (Hlib- ner, 1806) (dermis-ray), meriana Turner, 1967 (dennis only), madeira Riley, 1919 (partly coagulated FW yellow band), vicina Menetries, 1857 (condensed FW yellow patch), penelope Staudinger, 1897 (same with reduced dennis), aglaope Felder, 1862 (FW band distal to cell), melponiene (red fore- wing band), and burchelli (red forewing band and yellow hindwing stripe) fairly closely accompany the corresponding variations of erato (Map I), with a few notable exceptions in and near hybridization zones. Heliconius hermathena Hewitson, 1853. Very rare and local in northern central Amazon from Santarem (or perhaps Belem?), Maues, and Manicore to the far west (Sao Gabriel, Rio Negro). At Faro, occurs principally as subspecies vereatta Stichel, 1912, almost identical in color-pattern to melponiene mel- pomene; transitions are known between the nominate and mimetic subspecies. Heliconius erato (Linne, 1758) (Plate V, figs. 50-51; Plate VI, figs. 63-64; Map I). Entire Amazon Basin, common. Major subspecies amazona Staudinger, 1896 (dennis-ray), amalfreda Riffarth, 1900 (dennis only), es- trella Bates, 1862 (dennis-ray with reduced forewing band), reductiniacula Bryk, 1953 (condensed FW yellow patch), venustus Salvin, 1871 (same with reduced dennis), eninia Riffarth, 1901 (FW band distal to cell), hydara Hewitson, 1867 (red forewing band), and phyllis (Fabricius, 1775) (red forewing band and yellow hindwing stripe) are represented with approximate ranges and blend areas in Map I. Heliconius ricini (Linne, 1758) (Plate IV, fig. 31). Maranhao and Amapa westward to Roraima (Boa Vista) and Rondonia. 26 New York Zoological Society: Zoologica, Spring, 1972 Heliconius demeter Staudinger, 1896. Almost entire area of Hylaea (excludes southeastern and north-central Amazon), but extremely local; at times common where found. Sub- species bouqueti Noldner, 1902 (dennis-ray, with males imitating egeria), beebei Turner, 1966 (dennis only), eratosignis Joicey and Talbot, 1925 (partly coagulated FW yellow band and clearer rays), and demeter (FW band mostly distal to cell) closejy follow the variations of erato (see Map I). Heliconius sara (Fabricius, 1793) thamar (Hilbner, 1806) (Plate IV, fig. 33). Entire area, common in many habitats. Heliconius leiicadia Bates, 1862 (Plate VI, fig. 61). Entire area from Maranhao to Uaupes, Benjamin Constant and Acre, always local and very much less frequent than sara. The nominate subspecies, with a white HW bor- der, predominates over pseudorhea Stau- dinger, 1896 only in some populations north- westward. Heliconius antiochus (Linne, 1767) alba Rif- farth, 1900. Entire area except extreme south- east, commoner at the borders of the Hylaea in Mato Grosso and in Roraima. Eorm zobeide Butler, 1869 is most frequent in the lower middle Amazon; salvinii Dewitz, 1877 may be found in extreme northeastern Roraima. APPENDIX III Systematic Changes, and Remaining Uncertainties The new systematic arrangement of the sil- vaniform Heliconius is presented in Part V of this series. A total of six species is recognized {ismenius, silvana, niimata, hecale, ethilla, and pardalinus) , two more than those recognized by Emsley ( 1965 ) and with hecale much expanded. The largest uncertainties that still remain in the revision of this extremely complicated mimetic group, other than the placement of certain little- known subspecies, are the relationships of Heli- conius numata to the H.n. aristiona and H.n. aulicus complexes, of H. silvana to H.s. ethra, and of these two species to each other; and of the northern H. hecale group of subspecies to the H.h. quitalena complex of the Amazon Basin. The following species are added by the pres- ent paper to Emsley’s lists of 1963, 1964, and 1965, defining the tribe Heliconiini: Philaethria wernickei (separated from P. dido). Agraulis lucina (separated from A. vanillae). Eueides lampeto (separated from E. vibilia) . Heliconius astraea (separated from H. egeria). Heliconius besckei (separated from H. mel- pomene). Heliconius heurippa (separated from H. mel- pomene). Heliconius timareta (separated from H. mel- pomene ) . Heliconius luciana (added, provisionally being maintained separate from H. elevatus). Heliconius eleuchia (separated from H. sapho). Heliconius congener (separated from H. sapho). The two species Heliconius hygiana and H. clysonymus are recombined, the latter name taking precedence over the former. A number of taxonomic uncertainties still exist in the tribe. We have seen no specimens of the Peruvian Dione miraculosa Hering, 1926; from its original description, it may be a good species, isolated in southwestern Peru on the Pacific slope of the Andes. Eueides procula Doubleday, 1848 and E.p. edias Hewitson, 1861, while morphologically distinguishable, are connected in western Venezuela by a graded series (E.p. luminosus Stichel, 1903) and are probably conspecific. The situation of the Eueides lybia complex, however, is less clear; E. lybia lybia, E. 1. olympia (Eabricius, 1793) ant/ E. 1. lybioides Staudinger, 1876 are allo- patric, not connected by graded series, and mor- phologically distinguishable. While we favor maintaining them together, they may prove to be not interfertile. Heliconius hecuba, with which we have very limited field experience, may be separable into two sympatric species, though apparent intergrades are known in col- lections; the extremes of variation between hecuba Hewitson, 1857 at one end and cassandra Felder, 1862 at the other end of a sympatric population are quite far apart in many ways. Finally, until H. hecalesia and H. longarena are found flying together, the considerable possibil- ity that they may be conspecific (linked by H. h. gynaesia) cannot be eliminated. Brown & Mielke: Heliconians of Brazil, Part II 27 EXPLANATION OF PLATES 28 New York Zoological Society: Zoologica, Spring, 1972 Plate I Figure 1. Philaethria dido, Rio de Janeiro, ven- tral surface of hindwing, twice life size. Black, red, and green. Figure 2. Philaethria wernickei, Curitiba, Pa- rana, ventral surface of hindwing, twice life size. Black and green. Figure 3. Upper left (upside down): Agraulis vanillae maculosa, Xapuri, Acre. Upper right: Agraulis vanillae catella, Xapuri, Acre. Lower: Agraulis lucina, Alto Rio Jurua, Acre (identical to specimens from Xapuri). All in the Museu Nacional, Rio. Dorsal, life size. Black and orange. Figure 4. Six Agraulis from near La Merced, Junin, Peru, all ventral, life size. Orange, yellow, silver, and black. Left row: three variations of A. vanillae macu- losa. Lower right: A. vanillae catella (note orange FW apex). Upper and middle right: A. lucina. We also have specimens of lucina from this area with as much ventral silvering as the catella illustrated. All in the collection of G. Harris, Lima, Peru. Figure 5. Upper left: Eueides pavana, male, Xerem, Rio de Janeiro. Middle left: Eueides pavana, orange female, Petropolis, Rio de Janeiro, 900 meters. Lower left: Eueides pavana, intermediate fe- male, Parque Nacional de Itatiaia, Rio de Ja- neiro (900 meters). Upper center: Eueides pavana, yellow female, Belo Horizonte, Minas Gerais (1100 meters). Middle center: Eueides vibilia vibilia, female, Conceigao da Barra, Espirito Santo. Lower center: Actinote pyrrha (Fabricius) (Acraeinae), male, Rio de Janeiro. Upper right: Eueides vibilia vibilia, male, Con- ceigao da Barra, Espirito Santo. All dorsal, three-quarters life size. Black, yellow, and orange. Figure 6. Types (upper male, lower female) of Eueides nigrifulva Kaye = E. lampeto copiosus Stichel, Potaro River, British Guyana, dorsal, one-half life size. Black and orange. From the Allyn Museum of Entomology. Plate I 30 New York Zoological Society: Zoologica, Spring, 1972 Plate II Figure 7. Upper left; Heliconius congener Weymer, 1890, Abitagua, Oriente, Ecuador (1100 meters). Iridescent blue and yellow. Middle left: Heliconius sapho sapho (Drury, 1782), Victoria, Caldas, Colombia. Iridescent blue and white. Lower left; Heliconius eleuchia{2) eleusinus Staudinger, 1885-8, highway from Medellin to Quibdo, northwestern Colombia. Black and white with reduced blue iridescence. Upper right; Heliconius eleuchia eleuchia Hew- itson, 1854, Victoria, Caldas, Colombia. Irides- cent blue, yellow forewing bands, white hind- wing border. Middle right: Heliconius eleuchia prinmlaris Butler, 1869, Santo Domingo, western Ecuador. Iridescent blue and yellow. Lower right: Heliconius hewitsoni Staudinger, 1875, Agua Buena, Puntarenas, Costa Rica. Deep blue-black and yellow. All dorsal, two-thirds life size. Figure 8. Left: Heliconius melpomene inelpo- mene, Rio Negro, Meta, Colombia, 900 meters. Black and red. Right; Heliconius heurippa Hewitson, 1854, Rio Negro, Meta, Colombia, 900 meters. Black, red, and yellow. Both dorsal, two-thirds life size. Figure 9. Heliconius timareta Hewitson, 1867, polymorphic population from the Rio Topo be- tween Banos and Puyo, Ecuador, 1400 meters. Identical forms fly in the Abitagua, upper Santa Clara, and upper Rio Arajuno areas at 1000 to 1300 meters, and up the slopes of the valley of the Rio Pastaza to 1800 meters, in heavy humid forest. Upper left; nominate form (black and yellow). Upper right: form richardi Riffarth, 1900 (black, yellow, and red). Lower: two forewing band variants of form contiguus Weymer, 1890. Black, yellow, and red. All dorsal, four-fifths life size. In the Carnegie Museum, Pittsburgh. Figure 10. Intergradation of forms of Helico- nius melpomene in eastern Ecuador. All forms may be found flying together on the northern escarpment of the Abitagua highlands, such as near Santa Clara (600-1000 meters), and Ara- juno (500-1000 meters). The last two are found widely at higher elevations on the Rio Pastaza, such as at the Rio Topo, and elsewhere in east- ern Ecuador at comparable levels. The first sub- species is widespread in the upper Amazon Basin of Brazil, Peru, Ecuador, and Colombia. The width of the blend zone near Santa Clara does not exceed twenty kilometers in horizontal and 500 meters in vertical dislocation. Upper left: H. melpomene aglaope Eelder, 1862, from near Arajuno. Black, yellow, and orange. Upper center: form adonides Niepelt, 1908. Yellow forewing bands, orange dennis and rays. Upper right; form isolda Niepelt, 1908. White forewing bands, red dennis and rays. Lower left: form niepelti Riffarth, 1907. Red and white forewing bands, red dennis. Lower center: H. melpomene plesseni Riffarth, 1907. Red and white forewing bands. Lower right: form pura Niepelt, 1907. White forewing bands. All dorsal, two-thirds life size. Brown & Mielke: Heliconians of Brazil, Par! // 31 Plate II 32 New York Zoological Society: Zoologica, Spring, 1972 Plate III Figure 1 1 . Heliconius egeria and astraea in Brazil. Upper left and middle left: H. astraea astraea, Sao Paulo de 01iven§a, Amazonas; typical form. Lower left: H. astraea, new subspecies (see Part IV), Manicore, Rio Madeira, Amazonas. Upper right: Heliconius egeria hyas, typical form, Maues, Amazonas. Middle right: Heliconius egeria asterope, upper Rio Negro, Amazonas (typical egeria genitalia). Lower right: H. egeria egeria, typical, Manicore, Rio Madeira, Amazonas. Identical specimens are known from Sao Paulo de Olivenga. All dorsal, one-half life size. Black, yellow, and red. In the Museu Nacional, Rio de Janeiro. Figure 12. Type-series of Heliconius luciana, extreme upper Orinoco River, Territorio Ama- zonas, Venezuela (2°11' N., 64°12' W.; Raudal “Los Tiestos”). Upper left: holotype male. Upper right: allotype female. Lower left: paratype female. Lower right: paratype male. All dorsal, two-thirds life size. Black and white. In the Facultad de Agronomia, Maracay, Vene- zuela (with permission of Dr. Francisco Fer- nandez Yepez). Figure 13. Six specimens of Heliconius luciana from the variable population at Mantecal, Rio Cuchivero, Bolivar, Venezuela. All dorsal, two- thirds life size. Black and yellow except for middle right specimen, which is black and white. Taken from a painting by the collector, H. Skin- ner of La Victoria, Venezuela, with his per- mission. Figure 14. Heliconius luciana, male, Mantecal, Rio Cuchivero, Bolivar, Venezuela, dorsal, two- thirds life size. Black and yellow, In the Facul- tad de Agronomia, Maracay (donated by H. Skinner) . Figure 15. Heliconius luciana, male, Mantecal, Rio Cuchivero, Bolivar, Venezuela. Ventral, two-thirds life size. Black, yellow, and red. Brown & Mielke: Heliconians of Brazil, Part II 33 Plate III St H 34 New York Zoological Society: Zoologica, Spring, 1972 Plate IV Heliconians known from extra-Aniazonian Bra- zil and not illustrated on other plates. Primitive genera, Eiieides, primitive Heliconius, and most advanced Heiiconiiis (inra-group). Fig. 16: Black and green. Figs. 17-24, 26: Black and orange. Fig. 25 : Black, orange, and yellow. Fig. 27: Black, orange, and dull yellow. Fig. 28: Black and orange. Figs. 29, 32, and 33: Iridescent blue and yellow. Figs. 30 and 31 : Black, yellow, and red. All dorsal, two-thirds life size. Figure 16. Philaethria wernickei wernickei, Rio de Janeiro. Figure 17. Agraitlis vaniliae maculosa, male, Itanhem, Bahia. Figure 18. A. vaniliae fnaculosa, female, Para- opeba, Minas Gerais. Figure 19. Dione juno jimo, male, Xerem, Rio de Janeiro. Figure 20. Dryadula phaetusa, male, Rio Ma- ranhao, Distrito Federal. Figure 21. D. phaetusa, female, Barbacena, Minas Gerais. Figure 22. Dione moneta moneta, male, Rio Claro, Sao Paulo. Figure 23. Dryas iiilia iulia, large male. Canal Sao Simao, Goias. Figure 24. D. iulia iulia, small female, Para- opeba, Minas Gerais. Figure 25. Eueides isabella dianasa, male, Paracatu, Minas Gerais. Figure 26. Eueides vibilia unifasciatus, male, Alto Gargas, Mato Grosso. Figure 27. E. vibilia unifasciatus, female, Alto Gar?as. Figure 28. Eueides aliphera, male, Concei§ao da Barra, Espirito Santo. Figure 29. Heliconius wallacei flavescens, male, Chapada de Guimaraes, Mato Grosso. Figure 30. Heliconius xanthocles melete, male, Sao Vicente, 90 km. E. of Cuiaba, Mato Grosso. Eigure 31. Heliconius ricini, male, Dom Pedro, Maranhao. Figure 32. Heliconius sara apseudes, male, Belo Horizonte, Minas Gerais. Figure 33. Heliconius sara thamar, male, Bra- silia, Distrito Federal. Brown & Mielke: Heliconians of Brazil, Part 11 35 31 Plate IV 36 New York Zoological Society: Zoologica, Spring, 1972 Plate V Heliconians known from extra-Aniazonian Bra- zil and not illustrated on other plates (con- tinued). Genus Heliconiiis: silvana, melpoinene, and erato groups. Figs. 34-46: Black, yellow, and orange, with white subapical spot on forewing in 40-43, 45, and 46. Figs. 47-5 1 : Black, yellow, and red. All dorsal, two-thirds life size. Figure 34. Heliconiiis silvana etlira, male, Conceigao da Barra, Espirito Santo. Figure 35. H. silvana ethra, form brasiliensis, male, Recife, Pernambuco. Figure 36. H. silvana ethra, variant, male, Con- cei§ao da Barra. Figure 37. H. silvana ethra, form brasiliensis, variant, female, Conceigao da Barra. Figure 38. H. silvana robigus, male, Rio de Janeiro. Figure 39. H. ethilla flavoinaciilatiis, female, Recife, Pernambuco. Figure 40. H. ethilla narcaea, light male, Rio de Janeiro. Figure 41. //. ethilla narcaea, dark male, Santa Teresa, Espirito Santo. Figure 42. H. ethilla narcaea, female, Santa Teresa. Figure 43. H. ethilla narcaea, form polychroiis, male, Loreto, Sao Paulo. Figure 44. H. ethilla eiicoma, male, Ubajara, Ceara. Figure 45. H. ethilla narcaea, form satis, male, Rio de Janeiro. Figure 46. H. ethilla narcaea, form polychroiis, female, Loreto, Sao Paulo. Figure 47. Heliconiiis besckei, male, Brasilia, Distrito Federal. Figure 48. H. melpoinene nanna, male, Santa Teresa, Espirito Santo. Figure 49. H. melpoinene biirchelli, male, Rio Maranhao, Distrito Federal. Figure 50. H. erato phyllis, male, Rio de Janeiro. Figure 51. H. erato phyllis, form artifex Stichel, 1899, male, Rio de Janeiro. Brown & Mielke: Heliconians of Brazil, Par! II 37 Plate V 38 New York Zoological Society: Zoologica, Spring, 1972 Plate VI Heliconians marginal in extra-Amazonian Bra- zil, not illustrated on other plates; polymorphism in Amazonian heliconians. Figs. 52-60, 63, 64: Black, yellow, and orange to red. Figs. 61 and 62: Iridescent blue and yellow. All dorsal, two-thirds life size. Figures 52-55. Eueides Isabella Isabella, vari- ants from a single polymorphic population, all taken during 90 minutes’ collecting on June 2, 1971, over a ten-meter radius in Sao Vicente, 90 Km. E. of Cuiaba, Mato Grosso, 600 meters. The total sample is 21 specimens. Figures 56-58. Heliconius numata near superi- oris (upper male, middle and lower females), three variants from a fairly stable population, June 7, 1971, 17 Km. N. of Salto do Ceu, upper Rio Branco (tributary of the Rio Cabagal), west-central Mato Grosso, 400 meters. Figure 59. Heliconius silvana mirus, near typi- cal, male, Salto do Ceu, Mato Grosso, June 7, 1971. Figure 60. Heliconius biirneyi near typical burneyi, male, Pimenta Bueno, Rondonia; iden- tical to specimen observed near Salto do Ceu, Mato Grosso, 400 meters, June 7, 1971. Figure 61. Heliconius leucadia pseudorhea, male, Patrimonio Novo, upper Rio Jauru, west- central Mato Grosso, 600 meters, June 10, 1971. Figure 62. Variation in the population of Heli- conius wallacei flying just north of Obidos, Para. The second form from the top, plus com- binants with a wider band, represents the bulk of the population. All taken in July 1970. Iri- descent blue and yellow. Forms with white in- stead of yellow forewing bands, with all the illustrated band shapes, form up to one-quarter of the populations of wallacei in the northern middle Amazon; they may be still more abund- ant northward, in areas where the white-banded Heliconius antiochus is the predominant species of the genus. This polymorphic population of wallacei may be found as far southwestward as the Manaus area. Figure 63. Polymorphic hybrid population of Heliconius erato from Riozinho, 28 Km. down the Rio Machado from Pimenta Bueno, Mato Grosso, all taken in August 1970. The sub- species which meet here are amazona (upper left) from the northeast and venustus (lower right) from the south. A little farther down- stream, emina also joins the gene pool from the west, producing a continuous series of polymor- phic populations all the way down the Rio Ma- deira to Manaus, up the south bank of the Rio Negro to Barcelos, and westward to Sao Paulo de Olivenga (see map). Six principal forms can be recognized, for scor- ing members of populations along the Cuiaba- Porto Velho highway between Vilhena and the town of Rondonia, as follows: (1) amazona Staudinger, 1896. Very open yel- low band, full orange dennis and rays. (2) (hybrid). Very open yellow band, dennis restricted to three lines and much redder. (3) form constricta Joicey and Kaye, 1917. Yellow band closed down but still encircling much black, dennis orange and complete. (4) (hybrid). Same, with dennis red and re- duced. (5) form donatia Fruhstorfer, 1910. Forewing yellow band almost totally compacted but still enclosing a small black spot or bar at or extend- ing out from the end of the cell; dennis usually red and reduced. (6) venustus Salvin, 1871. Forewing yellow patch compact, without black in center, and somewhat reduced distally; dennis red and re- duced to three lines. The eight specimens in Fig. 63 would be scored 1—2 — 1—4 — 4 — 3/5 — 5 — 6. Analysis of some populations: f form — \ total Km. Elev.(m.) Locality 1 2 • 3 4 5 6 sample -200 700 South and west of Vilhena — — — 3 12 15 0 600 Vilhena, frontier MT/RO — — — — 7 12 19 70 400 Km. 70, Vilhena-Pimenta Bueno — — — 1 3 4 8 81 350 Km. 81, Vilhena-Pimenta Bueno — — — 3 25 19 47 190 320 Pimenta Bueno 1 3 — 3 2 1 10 220 300 Riozinho, 1970 season 3 4 8 5 6 1 27 Riozinho, 1971 trips 16 12 5 8 9 3 53 240 290 Km. 48 east of P. Bueno 3 2 3 1 — — 9 Brown & Mielke: Heliconians of Brazil, Part II 39 Plate VI 40 New York Zoological Society: Zoologica, Spring, 1972 Figure 64. Polymorphic hybrid population of Heliconiiis erato just north of Obidos, Para. All taken in the same area in July 1970 except cybelellits (taken in the area in 1931). The ap- proximate names of the forms are as below; the form dryope, with a hyclara entire red band and dennis, is not illustrated; all forms except hydara and amalfreda may be considered as hybrid recombinants. intermediates here are known as viculata and rubrizona intermediate, callista hydara Hewitson, 1867 belticopis Joicey and Kaye, 1917 callycopis (Cramer, 1777) (Fig. F) elimaea (Erichson, 1848) cybelellits Joicey and Kaye, 1917 Helena Riffarth, 1907 typical form has a more irregular FW band than this specimen CO rain Butler, 1877 amalfreda Riffarth, 1900 The approximate abundance of the forms in the 1970 population is (starting with the most common; total sample about 150 specimens); amalfreda — hydara, viculata, rubrizona — Helena and varieties — belticopis, elimaea, and coralii — dryope — callycopis — cybelellits. In the Obidos area, Heliconius melpomene is abundant and practically monomorphic as m. melpomene, with a black ground color and a broad red band on the forewing; a very few individuals have been taken with hybrid char- acters (dennis, or a mixed red-and-yellow fore- wing band). The first five illustrated forms of erato, and e. dryope, are very similar in flight to m. melpomene, appearing black with two bright red areas. The last three forms illustrated closely resemble the common sympatric H. bitrneyi catherinae and H. tales. Like these latter species, they fly higher, and are encountered more away from the streams, than the red- banded erato forms which fly lower and more slowly, with melpomene in the areas near per- manent water. This double Mullerian mimicry in both behavior and pattern has apparently helped to stabilize an extremely large hybridiza- tion zone between erato hydara and e. amal- freda. covering almost the entire northern half of the lower middle Amazon (see map). This hybrid zone also extends south across the river to Santarem and Maues, where hydara, which is apparently able to cross the river, meets not with amalfreda but with the rayed amazona; these latter two forms, as well as dennis-rayed subspecies of aoede, melpomene, and demeter, seem to find an impenetrable barrier in the Amazon/ Negro Rivers between the Ilha de Marajo and Barcelos. The stronger-flying spe- cies {Eiteides tales and Heliconius bitrneyi and egeria) cross the river occasionally (like erato hydara and m. melpomene) , producing popu- lations polymorphic for rays in these species on both sides. In Itacoatiara, on the north bank of the Amazon 200 Km downstream from Manaus, the hybrid population of erato is near its west- ern limit. A sample of 13 specimens caught and another 15 seen indicates almost equal abundance of all of the forms illustrated, plus dryope. This suggests that this population may be composed principally of individuals of hy- brid parentage, rather than the backcrosses of rarer hybrids to the parent subspecies as in Obidos. In Itacoatiara, the m. melpQinene popu- lation shows greater signs of hybridization than in Obidos, but no yellow-banded individuals were seen or taken in a total sample of over 50. Many specimens, however, had signs of white or yellow in the underside, and one in the up- perside, of the forewing red band. About one- quarter of the individuals also showed a dentate red line across the postdiscal area of the hind- wing, possibly caused by a gene related to that which transforms melpomene rays from trian- gular to nail-shaped. In the Manaus area, erato is monomorphic as e. amalfreda. Across the Rio Negro, only three Km away, the western dennis-rayed popu- lation with a polymorphic forewing band is found; there seems to be no gene exchange across the Rio Negro here. 2 The Heliconians of Brazil ( Lepidoptera : Nymphalidae). Part III. Ecology and Biology of Heliconius nattereri, a Key Primitive Species Near Extinction, and Comments on the Evolutionary Development of Heliconius and Eueides (Plates I-IV; Text-figures 1-4) Keith S. Brown, Jr. Centro de Pesquisas de Produtos Naturals Faciildade de Farmacia, UFRJ Rio de Janeiro ZC-82, Brazil The nearly unknown primitive butterfly species Heliconius nattereri, purported to be sexually dimorphic and seen but three times in this century, was relocated in steep humid primary forest near Santa Teresa, Espirito Santo, Brazil. A large colony was encountered in 1970, and a basic ecological and biological study was concluded during the late summer and fall expansion of the species (from mid-February to mid-May). The com- plete and accentuated sexual dimorphism was confirmed (the female was described and has been known as H. fruhstorferi) . The juvenile and adult characters indicate the species to be very primitive in its genus, probably near the evolutionary line from Dryas iiilia to the silvana, inelpoinene, erato, and charitonia groups of Heliconius. The egg, pupa, and adult morphology are very near those of the silvaniforms, while the larval stages show relationships to more primitive and more advanced heliconian species. The adult males of nattereri promenade rapidly in strong sunlight on fixed paths in the upper levels of the forest, while the mimetic females are retiring and may rarely be observed on flowers, especially Lantana caniara and Gurania. The specialized adult behavior and restriction to a rare food-plant also used by other common, adaptable, and aggressive members of the genus are evidently the factors responsible for the apparent decline of nattereri since its discovery in the last century; its probable extinction is being greatly accelerated by man’s destruction of its virgin habitat, with the creation of grossly disturbed areas in which nattereri's more flexible relatives thrive. Any scheme for the radial evolution of the genera Heliconius and Eueides must take into account the apparent relationships of nattereri, which show very little connection to Eueides or the hierax, aoede, wallacei, xanthocles, or doris groups of Heliconius. A very ancient bifurcation in the evolution line in the latter genus would place the above- mentioned groups in a “primitive” branch, and nattereri with the silvana, erato, niel- pomene, charitonia, sara, and sapho groups in an "advanced" branch. This evolution can then be graphed in such a fashion as to suggest the simultaneous appearance of certain characteristic mimetic color-patterns in members of diverse groups in the two genera. Most of the evolution of the group probably took place in the mid-Tertiary, especially the Miocene, though many subspecies were certainly formed during the Pleisto- cene. Some excellent examples of convergent, divergent, and parallel evolution are evident in the 55 species recognized in the two genera. Two unusual red-banded species, Heliconius hennathena and H. telesiphe, are appar- ently little protected by Mullerian mimicry; they appear to be highly specialized to restricted biotopes (Amazonian sandy cerrado pockets and Andean pre-montane forest, respectively), and probably represent valuable evolutionary relics like nattereri, in need of intensive ecological and biological study. Some generalizations on the distribution and behavior of relatively primitive and relatively advanced species of heliconians can be made, which cast some light on the evolution of the various subgroups of species. Much additional ecological fieldwork will be necessary in order to make these suggestions, as graphed and discussed, more firm and useful, in understanding tropical evolutionary trends on the larger scale. 41 42 New York Zoological Society: Zoologica, Spring, 1972 Introduction IN THE PREVIOUS PAPERS in this Series (K. Brown, 1970; K. Brown and Mielke, 1972), we have discussed the history and recent rediscovery of Heliconius nattereri\ described' the tribe Heliconiini in Brazil; offered general comments on the species in the tribe; and pre- sented a supplementary revision to the papers published by Emsley (1963, 1964, 1965). The detailed study of the near-extinct primitive spe- cies H. nattereri, discussed below, has indicated a fundamental reformulation of the evolutionary scheme advanced by Emsley (1965) for ther genera Heliconius and Eiieides. This reformula- tion is presented here as a descriptive graph, and discussed with relation to the 55 species recognized in this series of papers as belong- ing to the two genera. Ecology and Biology of Heliconius nattereri At the time of Emsley’s revision (1965), H. nattereri Eelder and Eelder, 1865, was the least-known of the heliconians, with the pos- sible exception of H. luciana, omitted from that - revision through an oversight (see Part II of this series). Emsley noted the existence of "less than eight specimens of each sex" of nattereri and indicated that the two sexes, very different in appearance, had not been captured together. The presumed female, H. fruhstorferi Riffarth, 1899, was placed in his revision as a simple color-form of nattereri, unassociated with sex. At the beginning of this research program, we verified the existence of 13 males, all nat- tereri, and eight females, all fruhstorferi, in European collections, and none in American museums or the Allyn (ex Kaye) collection. At the beginning of our project, two additional males with accurate data were discovered in the Museu Nacional in Rio; one from Agua Preta, near llheus, Bahia ( = Fazenda Sao Joao, present-day town of Uruguca, north of Itabuna and well inland from llheus), September 1928, collected by E. May; and one from Santa Teresa, Espirito Santo, May 19, 1928, collected by E. Conde (from the Julius Arp collection). Edu- ardo May ( 1939) also mentioned seeing another male in the Uruguca area in 1928, and observ- ing a high-flying male near the Corrego Sabia north of Colatina, Espirito Santo, in October 1936. No further specimens could be discovered in other Brazilian or Latin American collections, and no living collector could be found, even in Santa Teresa, who had seen the species alive in recent years. In 1966, we instructed a resident insect col- lector in Santa Teresa, Claudionor Elias, to keep a lookout for nattereri. In the following year, he succeeded in collecting two males, on March 15 and June 8, now deposited in the collection of the Departamento de Zoologia in Curitiba, Parana. We then concentrated our work in the Santa Teresa area, and were able, over three years, to make a basic biological and ecological study of the species, and collect and breed several dozen specimens. Of these, we have placed three pairs and an additional female with unexpanded wings (mechanical difficulty during emergence from the pupa) in the Museu Nacional in Rio de Janeiro; one pair each in the Allyn collection (Sarasota, Florida), the Facul- tad de Agronomia (Universidad Central de Venezuela, Maracay), and the Carnegie Mu- seum, Pittsburgh; one pair to O.H.H. Mielke (Curitiba, Parana) and to Dr. J. R. G. Turner (York, England) ; and one male each in the col- lections of Harold Skinner (La Victoria, Vene- zuela), Francisco Romero R. (Maracay, Vene- zuela), Dr. E. W. Schmidt-Mumm (Bogota, Colombia), Jorge Kesselring (Joao Pessoa, Pa- raiba), Gordon Small (Panama Canal Zone), and Dr. Helmuth Holzinger (Vienna, Austria). We are attempting to place the remaining speci- mens in all important Heliconius research col- lections around the world."' Ten males and three females were also collected by K. Ebert in February and March of 1970, and are in the H. Ebert collection in Rio Claro, Sao Paulo, and a further pair was taken in Santa Teresa by C. Callaghan of Rio in April 1971. We have travelled in the territory presumably once occupied by nattereri in southern Bahia, northern Espirito Santo, and eastern Minas Gerais, but saw very little habitat suited to its demands (see below), and no further individuals of this species in areas other than Santa Teresa. Thus, all of the observations in this paper are drawn from field work in six colonies of nat- tereri discovered in the Santa Teresa area, at 500 meters to 900 meters elevation in dense primary Amazonian-type forest. A total of somewhat over 250 individuals has been ob- served during the three years of the study. Heliconius nattereri is the only member of its genus to demonstrate strong sexual dimorphism (Plate I, figs. 1-6). The Guianian H. demeter bouciueti Noldner, however, is moderately di- morphic in color-pattern and behavior, like nattereri (fide W. W. Benson; see Turner, 1966). The two sexes of nattereri occupy only poorly overlapping micro-habitats in the forest, and were evidently found together for the first time in our work in 1968. The number of specimens now known is sufficient to confirm a complete association of the color-forms with sex. This contrasts with the symmetrical distribution in both sexes of the two color-morphs of H. ethilla * If interested, please contact the author for details. Brown & Mielke: Heliconians of Brazil, Part III 43 narcaea {narcaea and satis, discussed in Part II), sympatric with and quite similar to the female of H. nattereri, in Santa Teresa. However, Turner (1968a) suggested that these morphs may eventually become wholly associated with sex, at least for the correspondingly dimorphic Trinidadian H. ethUla ethilla. The black, yellow, and orange female of nattereri has a slow, casual flight except when startled, when it either mimics Mechanics flight if mildly disturbed, or flies rapidly and directly upward and away if frightened. It joins very effectively the most common south Brazilian black-yellow-orange mimetic complex, whose principal distasteful members are silvaniform Heliconius (see Part V of this series) and itho- miines such as Mechanitis lysimnia and po- lymnia. A number of little-known ithomiine species from Bahia and Espirito Santo, notably Hypothyris eiiclea laphria and H. daetina, Hya- lyris fiammeta, and Napeogenes xanthone, ap- proximate in color-pattern the female of natte- reri more than they do any other Heliconius species. A notable resemblance to the female of nattereri, even to identical red basal markings on the ventral hindwing surface, is achieved by the sympatric but much more widespread Bate- sian mimic, Disniorphia astyocha (Plate I, figs. 7 and 8). Even more similar in markings, but much smaller in size, is an unusual variant of Phyciodes (Eresia) lansdorfi very near to form jacinthica (Plate II, fig. 10); this form was originally described, and today is principally known, from the nattereri faunal region. Two other Batesian mimics in this subgenus, Phyci- odes (Eresia) eiinice esora and the nearly un- known Bahian P. (E.) erysice, are also very similar in color-pattern to the female of H. natte- reri (Plate II, fig. 10). The females of nattereri apparently leave the deep forest to visit flowers a number of times daily — once in very early morning or on cloudy days at first clearing, and once or twice near midday (Table 1 and Graph 1 ), when they stay in the shadows and the undergrowth, approach- ing the flowers warily and leaving them quickly when satisfied. They seek out the food-plant (see below) to lay eggs in late morning and early afternoon, and may very rarely be seen other- wise, flying through the woods high above the ground or sunning on leaves near flowers or the food-plant (Table 1). The yellow-and-black male of nattereri, when in high rapid flight through the forest, looks very much like a windblown dead leaf, due to a very fast, shallow, and irregular wingbeat. It also has been confused occasionally in life (by the author) with a number of smaller sympatric ithomiines (especially Scada reckia, Napeo- genes yanetta and sidplnirina, and Aeria olena), and with three day-flying Dysschematid (= Pe- ricopid) moths, all of which have a much wider distribution than nattereri today (Plate II, fig. 11 ). The very rare and localized “splinter spe- Table I. Field Behavior of Heliconius nattereri (Total numbers of observations over four years) Hour (Day in Santa Teresa in March is 5:15 AM to 6:00 PM) Visiting and feeding on flowers Male Typical promenade Sunning on leaves Visiting and feedingon flowers Female Inspecting and laying on T. oralis Other (sunning, flying) Male Courting Female or other species 7:30- 8:00 1 - - 5 — 1 - 8:00- 8:30 1 - - 4 - - - 8:30- 9:00 11 - 6 - 1 - 9:00- 9:30 20 1 - 3 - 1 - 9:30-10:00 25 5 - 3 - 1 - 10:00-10:30 40 10 - 7 1 1 2 10:30-11:00 42 12 1 6 2 3 - 11:00-1 1:30 27 16 1 1 1 2 3 1 11:30-12:00 21 20 - 7 4 4 1 12:00-12:30 14 23 1 5 _ 2 - 12:30- 1:00 15 33 1 8 4 3 - 1:00- 1:30 14 22 2 7 - 2 - 1:30- 2:00 5 16 3 1 - 1 - 2:00- 2:30 0 16 1 2 - - - 2:30- 3:00 1 9 2 - - 1 - 3:00- 3:30 - - - 1 - - - 3:30- 4:00 - 1 - - - - 4:00- 4:30 - - 1 - - - - TOTALS 237 183 16/436 76 13 24/113 4 44 New York Zoological Society: Zoologica, Spring, 1972 cies” of Pierid, Perrhybris fiava, is known today only from the steep areas around Santa Teresa; the bright yellow male is very similar to nattereri males in flight, while the female, which we have not observed in nature, is very much like the female of nattereri in color-pattern (Plate I, fig- 9). Males of nattereri visit flowers profusely in mid-morning (Table 1 and Graphs 1 and 2), and may pass over them or occasionally stop later in the day. Most of their time in the heat of the day is spent in promenade (Table 1 and Graph II), usually high above the ground on a set and repeatable path day after day, with the area covered estimated in four separate cases as over 50,000 square meters. The frequency for passing a fixed observation point in one direc- tion is almost invariably fifteen minutes (in bright weather). The males always fly in the brilliant sun, and thus usually occupy the middle or upper story in dense forest; they rest on leaves during periods of cloud shade. In mid-afternoon, they may land with open wings on a sun-bathed leaf, usually high in a tree or vine (Table 1 and Graph 2). Both sexes are observed most easily at flowers in the morning. The preferred flowers, when available, are the introduced but widespread red-and-yellow composite blooms of Lantana cainara ( Verbenaceae) . Native red Gurania seilowiana, a cucurbitaceous vine also contain- ing many flowerlets in a single head, is very frequently visited when in flower inside the woods. Where poinsettias {Euphorbia pulcher- ritna) of the less ornamental sort, with mul- tiple yellow flowerlets surrounded by the red leaves, are introduced into the woods, they rapidly become the meeting and focal point for all local Heliconius, including nattereri', the heavily-visited blooms are produced from May through November. Other flowers occasionally visited by nattereri include small white orchids, magenta flowers of Passi flora kermesina (Plate I, fig. 6), blue and violet Eupatorium species, and red and yellow bromeliads. The males tend to visit or pass over flowers in the morning, before the high promenade period, at precise 15-minute intervals, suggesting that they may already be in a lower preliminary promenade, knowing that females also come to flowers dur- ing this period. Assuming from the breeding program (see below) a sex ratio near unity, the much larger number of observations of males over females on flowers attractive to both (Table Graph 1. Flower-visiting of male and female Heliconius nattereri, plotted as percentage of each sex in the total individuals seen at flowers during each hour of the day’s activity. Solid line: females; dotted line: males. Dashed line: median percentage of females in all the observations of nattereri at flowers. The females predominate in early morning, males in mid-morning. Data taken from Table 1. Brown: Heliconians of Brazil, Part III 45 1 ) may be due to the higher liquid requirements of the males, which promenade for hours in the midday sun, in relation to the females, which stay within the shaded woods and indeed have been observed almost exclusively during their furtive visits to forest-edge flowers. The sexes are most likely to encounter each other in nature on or near preferred flowers in mid- to late morning; we have no evidence that virgin females produce any scent that could at- tract males from any distance. Males of any age will try to court young females, and also occasionally both sexes of the quite similar H. ethilla narcaea, when they notice them in the wild, but normally do not court in captivity. The aerial phase of the courtship is as seen in most heliconians (Crane, 1955, 1957), and the fanning by the male starts at the front or side, proceeds to the back, and returns to the front, over the stationary female with wings held half- open. Thus, the overall courtship is very similar to that of Heliconius melpomene (Crane, 1957 ) . In February to May 1970, H. nattereri was found to be truly common in one area on the edge of an enormous tract (over 100 Km^) of virgin forest east of Santa Teresa. This area (Table 2) included an exceptional abundance of flower food coupled with many food-plants in vigorous growth, in a setting of forests, clear- ings, and second growth which made observa- tions and photography extremely easy. The ex- ceptional population density present, with con- sequent frequent encounters between individu- als, caused unusual quantitative (but not quali- tative) changes in the species’ normal behavior: males flew lower and more slowly, and prome- naded near the ground over relatively smaller areas which were shared occasionally with other males, and both sexes visited low flowers, even in open areas, persistently and fearlessly. That the genetically determined behavior of butter- flies is modified very often by environment and population density has been observed in a num- ber of families and regions. The modification in this colony of the habits of nattereri to produce, from a wild, high-flying, inflexible, and nearly impossible-to-observe species, a heliconian much more like the common and adaptable species. Graph 2. Distribution of the activities of male nattereri during the day, plotted as total number of sight- ings in each half-hour period (data from Table 1). Upper (solid) line; total sightings in all activities; the dip around midday may be real and is definitely not connected with decreased activity on the part of observers. Dotted and dashed line; flower visiting (peak in mid-morning). Dotted line: typical high prome- nading (peak in early afternoon). Dashed line: resting and sunning on leaves (peak in mid-afternoon). 46 New York Zoological Society: Zoologica, Spring, 1972 w Ph z o o u u o CC < Z z* § 5 d Q w s J | = s hJ * H ^ O tu S W H ■ c/5 >- z o u o u UJ S a z o p c > lai UJ c/5 n [X] O U ffl > 3 •z: o u x: dj ' it: 5 “I OfS‘or^O^OO’^'OO^r^r^'^DoooooOr-.ro^r^'O^^OC^'^’*:^• (fNi(N(Nr4io vo'or^or4r^'omoaNr4rn'^oo(Nr^^o\’— 'rofoo\r4^r4v^Oror^r^ ^^^^‘or^oo(N'sOf^ |\cc^o<^'sO’OrfTj-t-< CN rn I I I I I I -I OM5'0'Ot~'\Ot^O-HCMCMCMiOVOt^CM"CM'C)-';l-M5mmvO - ^(N'sD’— 'r^m%.S2 ' O .S ^ & G $ dj ■ o ?:? S <3 "O •X X DC 2 X C/D g)S O ^ X) <1^ ^ o O 1/5 c c «3 D, X O d X d a ° 6^ ^ So oo ^ &0 ■£ S -S to (/) to e CO O C o h O ^ ^ § :r ^ u X ^ u d P C d •« 00 ^ C CO O X « 8 co” ^ B S- O Cd ■ a >■ (U 'W 8 - I u- >< u- C c ! H W H W C/D X . ^ o 't: oj 0 > ^ - S “ o £ 1 « 2 O i,- C3 £ 8-o« CU 2 (1> i S i'^ S5 ^2 a M S3 « 3.S ^ c«- « ^ ^ O a> •rt u (L> ^ — H 00 ^ ^ *n 1 'O 2-. 8 o 5 ^ g 'd u P CO a > Jr X a 0/ c/3 — ^ 3 0/ ^ i-s tt-i s-c ° o- « _ 3 a c ^ <« S 2 S o a = " *- s: "3 cd a (U d X SB o E5 1/5 hJ t/5 S X J3 ^ Cd X (U OO ^ 3 ■<-- O t(_i O E c d ,0 CO •> 2 S d ^ (U (U X X H H S < oa U Q w fc a ffi H, i4 j S Z 5 0 O ^ r- « X -U tl. ~ ‘ CO ooooooooooooo OS ^oor^CN*/^'/^cS'o Grand^ QO NOftTEd typical hybrid 1 latlTltta t^'Hardcoi ^PERNAAABUCg IToNDaNiA; phyllls. •CMpada d», Oruyuoa*Bi Dnpattemed area 1b extra-Amas \ CltlBB and collecting areae indicated by heavy dots. State boundarlee Southeastern mountain divide Pattema in the Amazonian arei occupied by the illustrated aj of Heliooniue erato . Overlapping o^ two or more pa probable locations of bland o zones (see Figures 63 and 64) Crosses in squares are points usable statlstloal sample of erato is available ■ Rio Cabacal - Rio Branco - Rio Jauni area: a south- ward and eastward extension of the Hylaea (general Amazonian forest), where H. nuaata '^riations,^^^^ silvana miras. Hi burnt subsp. nov. Ts« ^ ^ Farthest south Coiumba River SysTt plB , Hi riolnli '* venustus fAIhlAS cidRAIS GROSSO Conoei^ j^SPm'lTft SANTO leyl, hT leuoadla and H. aoede "Ivy fly in the t'araguey basin. lonflroed locality for the "Cuy^ba- ' " spaoleBi H, demeter eratoBlg~ 1-^, ^1 H, elevatuB BChTnaeamanni. and ^ B^traea subspT nov, (see Part iv; , Poorly explored areas where these same four forms may eventually be found in the Oulaba drainage. Areas where beeokei and melpomene are sympatrio. Hew eubspecles of ethllla (see Part IV), Philaethria dido and P, wernlokel sympatrio. H, ethllla eueona in extra-Amazonlan region. Eueides Isabella Isabella outside the Amazon Basin, Pure populations of Dione .^uno suffumata. Known looalitles for Helloonlus luoiana. Brazilian localities for Clone moneta aoneta. Heliooniue nattererl seen in this century. H, riclnl outside the Amazon Basin, H, Sara tbamar In eitra-Amazonlan Brazil, Eueides vlbilia uDlfaaolatuB outside Amazon Basin. Heliooniue wallacei flaveso*Tip outside Amazon ares- Helloonlus xanthooles melete outside Amazon area. ~^S^Pwt.T»apaT< « 'Xef 4a ^Jolnvllle t NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable), and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to ne editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 1. The Heliconians of Brazil (Lepidoptera: Nymphalidae). Part II. Introduc- tion and General Comments, with a Supplementary Revision of the Tribe. By Keith S. Brown, Jr., and Olaf H. H. Mielke. Plates I-VI; Text- figures 1-12; Map 1 2. The Heliconians of Brazil (Lepidoptera: Nymphalidae). Part III. Ecol- ogy and Biology of Heliconius natterei, a Key Primitive Species Near Ex- tinction, and Comments on the Evolutionary Development of Heliconius and Eueides. By Keith S. Brown, Jr. Plates I-IV; Text-figures 1-4 41 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx, New York 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, New York. Published September 1 1, 1972 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 57 • ISSUE 2 • SUMMER, 1972 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Simon Dresner, Editor & Curator, Publications and Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman', William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Curator, Ornithology; Joseph Bell, Associate Curator, Ornithology; Donald F. Bruning, Assistant Curator, Ornithology; Hugh B. House, Curator, Mammalogy; James G. Doherty, Assistant Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Robert A. Brown, Assistant Curator, Animal Departments; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director; Stephen H. Spotte, Curator; H. Douglas Kemper, Assistant Curator; Christopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Director and Pathologist; George D. Ruggieri, S.J., Assistant Director & Experimental Embryologist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung, Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology; Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate; George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists; Robert M. Beck, Research Fellow ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants: John Budinger, Pathology; Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology; Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology; Harold S. Goldman, Radiology; Roy Bellhorn, Paul Henkind, Alan Friedman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine; Theodore Kazimiroff, Dentistry © 1972 New York Zoological Society. All rights reserved. 3 The Hematological Parameters and Blood Cell Morphology of the Brown Bullhead Catfish, Ictalurus nebulosus (Le Sueur) (Tables 1-3) Sheila R. Weinberg,’ Charles D. Siegel,- Ross F. Nigrelli,^ and Albert S. Gordon^ Department of Biology, Graduate School of Arts and Science, New York University, New York, New York 10003, and the Osborn Laboratories of Marine Sciences, Brooklyn, New York 11224. The present research indicates that the various procedures utilized in the study of the hematopoietic systems of the higher classes of the vertebrates are applicable, with some modifications, for the study of the hematology of the poikilotherms. A description is presented of the peripheral blood cell constituents, with an emphasis on the distinction of the various stages of development of the erythrocytic and leucocytic series, reflected by; changes in the nuclear chromatin, location, size and shape of the nucleus, size and shape of the cell, and the state of cytoplasmic basophilia. One of the aims of this study was to examine some aspeets of the erythrocytic system in the normal, bled, and mechanically stressed bullhead catfish. Values were determined for red cell, white cell, reticulocyte, differentials, hematocrit, hemoglobin, and the erythro- cyte corpuscular constants. Statistical analyses indicate a significant difference in the red blood cell count, white blood cell count, reticulocyte count, and mean corpuscular volume after bleeding. The evidence supports the view that there is a physiological control, possibly hormonal in nature, responsible for blood cell formation in different species of fish. Introduction There is a need for additional informa- tion concerning the morphological and physiological characteristics of the blood of different fishes. The suitability of the catfish as an experimental animal in hematology and immunology has been clearly demonstrated in investigations conducted by Dawson (1935) and more recently by Chuba, et al. (1968); Wiener, et al. (1968); Kuhns, et al. (1969); and Baldo and Boettcher (1970). The purpose of the present investigation was to obtain more quantitative data relating to the hematological parameters and morphology of catfish blood. Careful selection was made of methods previously used and in some cases a few modifications were instituted to improve upon specific procedures. ’ Pre-doctoral Trainee, National Institutes of Health Training Grant 2-T01-HE05645-06, National Heart and Lung Institute, National Institutes of Health, United States Public Health Service. - Supported by research grant 5-R01-HE03357-14, National Heart and Lung Institute, National Institutes of Health, United States Public Health Service. ^ Supported by Scaife Family Charitable Trusts, a grant to the New York Zoological Society. The blood cell values obtained from these fish may reliably be considered as representing the normal blood picture of the brown bullhead with the question as to a possible difference from wild catfish being unlikely. In this connection, it has been demonstrated that there were no significant differences in the numbers of erythro- cytes, hematocrit percentages, and hemoglobin levels between the blood of wild and hatchery- reared lake trout (Piper and Stephens, 1962) and silver salmon fingerlings (Katz and Don- aldson, 1950). Materials and Methods Aquaria and Fish A normal group of brown bullhead catfish ranging in size from six to eight inches and weighing less than 50 grams were maintained in fiber glass tubs. A 365-gallon tub was found to be sufficient to support 30 brown bullhead catfish, if it was equipped with a filter system and water pump to maintain a constant flow of water, and two air lines for constant bubbling of air through the closed water system. The tem- perature, oxygen and pH of the aquaria were measured and recorded daily in the morning. 71 72 New York Zoological Society: Zoologica, Summer, 1972 The temperature was maintained between 22° to 25° C, the oxygen content was above 5 ppm, and the pH was maintained between 6.3 to 8.5. These levels have been recommended by several fish hatcheries (Bureau of Sport Fisheries and Wildlife, 1970; Darragh Company, 1970; and Ralston Purina Company, 1970). The catfish were maintained on a diet of freshly ground beef, and fed daily in the morning. However, the fish were starved for a period of at least 24 hours prior to handling, either for the purpose of mechanically tagging, transfer to an experi- mental 50 gallon tank, or blood sampling. Even though sex determination of catfish is difficult, an attempt was made following descriptions outlined by some commercial fish journals ( American Fish Farmer, Bureau of Sport Fish- eries and Wildlife, 1970). Blood Letting The unanesthetized bullhead was placed ven- tral side up, held gently but securely by pressing down the abdomen, just below the pectoral fins. A 0.5 or 1.0 cc tuberculin syringe fitted with a one inch 20-gauge needle, premoistened with a 3.8 percent solution of sodium citrate in Locke’s isotonic saline, was oriented with the ventral aorta and inserted gently into the heart, just beneath the pectoral girdle. The syringe plunger was gently pulled back until the desired quan- tity of blood was obtained. If sufficient blood was not obtained within 30 seconds after the initial penetration, the syringe was withdrawn and the fish was returned to the aquaria. Coagu- lation of blood, which porbably occurs in the pericardial cavity, could prevent the drawing of the blood sample through the syringe needle (Chuba, et al., 1968; Klontz, 1968; and Dupree, 1970). For repeated blood samplings of fish, blood letting from the heart is recommended (Dupree, 1970). Bleeding from the caudal fin introduces the danger of destroying the caudal vein and bacterial infection. Bleeding was always performed during the same time in the morning to avoid the complication of possible diurnal variations. It is important to note that 3.8 percent sodium citrate was found to be the only anticoagulant that did not destroy the cell morphology as seen in the blood smear preparations. Following the cardiac blood letting method outlined, fish can be bled from 0.25 ml to 2.0 ml with no mortality. Determination of Hematological Parameters Absolute blood cell counts. Erythrocyte counts were made by diluting one part blood with 200 parts of Hayem’s solution in a red blood cell diluting pipette, counting the cells in the five smaller squares in a Spencer Bright-line hemocytometer, and multiplying the total count by 10^ (Smith, et al., 1952; Hesser, 1960; and Klontz, 1968). For the white cell count Shaw’s Avian solutions were used (Hesser, 1960; and Klontz, 1968). Both solutions must be filtered prior to use. Blood was drawn up to the 0.5 mark of a red cell diluting pipette; Solution A was added to fill the bulb of the pipette approxi- mately one-half full and mixed. Next, the pipette was filled to the 101 mark with Solution B. A hemocytometer was used and the cells in the four large squares were counted and multiplied by 500. All the counts were made in duplicate for each fish and averaged. The cell counts are representative for each cu mm of whole blood. Hematocrit determination. The hematocrit was determined by the microhematocrit tech- nique (Larsen and Sneiszko, 1961a). Blood was collected in commercially prepared heparinized capillary tubes and then centrifuged at a high speed for five minutes. Each pair of tubes for each fish was examined to determine the volume of packed red blood cells. Hemoglobin concentration determination. Several procedures have been used to determine the hemoglobin concentration in catfish blood (Larsen and Snieszko, 1961b; and Larsen, 1964). The cyanmethemoglobin method has been found to be constant and the first choice for the use on catfish blood. A sample of 0.02 ml blood was mixed with 5 ml of Drabkins solution (Wintrobe, 1968). The amount of hemoglobin was then determined spectrophotometrically (at 540 mu) against a commercially prepared stand- ard solution of cyanmethemoglobin (Ortho Diagnostics, Raritan, New Jersey). The hemo- globin is reported as gm/100 ml whole blood relative to man. Using this technique, the as- sumption was made that fish hemoglobin under- goes the same reactions with potassium-ferri- cyanide solution as does mammalian hemoglobin (Larsen and Snieszko, 1961b). Absolute indices. The absolute indices were calculated for the brown bullhead catfish as follows; MCV (Mean Corpuscular Volume) in , . . Hematocrit (%) X 10 cubic microns — ,• mCH RBC (lOVcu mm) ’ (Mean Corpuscular Hemoglobin) in pico- Hemoglobin (gm %) X 10 grams = ; MCHC ^ RBC (lOVcu mm) (Mean Corpuscular Hemoglobin Concentration) in percent Weight/ Volume = Reticulocyte values. A sample of blood was drawn into a plain capillary tube. An equal volume of a one percent Brilliant Cresyl Blue solution in Locke’s isotonic saline was drawn into the tube. The mixture was expelled and thoroughly mixed on a piece of parafilm, then taken back into the capillary tube for about five minutes, mixed again and a small drop was then Weinberg, Siegel, Nigrelli, & Gordon: Brown Bullhead Catfish 73 placed on a methanol cleaned slide to be smeared. After smearing, the slide was air-dried, fixed with absolute methyl alcohol and counter- stained with Wright’s stain (Humason, 1967). One thousand erythroid cells were counted per slide under oil immersion with the aid of a reticule and reported as per cent of reticulocytes. Blood cell morphology. A drop of peripheral blood mixed with 3.8 percent sodium citrate was smeared on slides previously cleaned with 50 percent methanol, air-dried and fixed with abso- lute methanol. The blood smear preparations were then stained with Romanowsky stains; ( 1 ) Benzidine stain, Wright’s, and Giemsa; (2) Wright’s and Giemsa; and (3) May-Griinwald and Giemsa. The smears were examined with an oil immersion objective under 1000 X magnifi- cation, and measurements were made with an ocular micrometer. All procedures were carried out under sterile conditions. Results Blood Parameters In an attempt to determine the normal hema- tological picture of the brown bullhead, seven fish were mechanically tagged by caudal fin clipping, transferred to a 50-gallon tank, and bled 0.25 ml at two week intervals, thereby allowing a comparison of the blood parameters of individual fish. Similarly, 25 fish were selected at random from a stock normal population, measured, and bled 0.25 ml. The results of the blood analyses and calculations for both groups of fish are listed in Table 1, which shows the total red cell numbers (RBC), total white cell numbers (WBC), reticulocytes (percent per 100 erythroid cells), hematocrit values, hemoglobin concentrations, Mean Corpuscular Volume (MCV, cu p). Mean Corpuscular Hemoglobin (MCH, picograms, pg), and Mean Corpuscular Hemoglobin Concentration (MCHC, percent). The values are given as the mean, plus or minus one standard error of the mean; the number of fish used in each group is given in parenthesis. Blood Cell Morphology Compared to the cells of the erythrocytic series, the identification of the leucocytic series in fish is difficult, especially when attempts are made to discriminate between the granulocyte and agranulocyte. Furthermore, it is even more difficult to distinguish the young granulocytes and agranulocytes from the cells of the throm- bocytic group. The reported descriptions of the agranulo- cytes in fish are in disagreement, Yuki (1957, 1958) has attempted to resolve the discrepancy in classification of the young and transitional cells in the monocytic and granulocytic groups in rainbow trout. The difficulty in the task of correctly distin- guishing thrombocytes and small lymphocytes lies in the fact (Saunders, 1968b) that some fish species have only mature thrombocytes in their peripheral circulation, while in others both ma- ture and transitional cells can be found. The following blood cells were detected in the peripheral blood of the brown bullhead: basophilic erythrocytes, reticulocytes, mature erythrocytes, senile or senescent erythrocytes, “nuclear shadows” or basket cells, round lym- phocytes, elongated thrombocytes, round throm- bocytes, fusiform or spindle-shaped thrombo- cytes, monocytes, neutrophils, eosinophils, macrophages, and hemocytoblasts. The enucle- ate erythrocyte or erythroplastid, which arises from the pinching off of a cytoplasmic portion of an erythrocyte, was occasionally found in the peripheral smear preparations but was not con- sidered to be indicative of a blood dysfunction. No basophils were seen in any of the blood smear preparations examined. The terminology of the cellular components of the brown bullhead used in this paper coin- cides with those proposed by Jakowska ( 1956) ; Chlebeck and Phillips (1969); Yuki (1960, 1963); Srivastava (1968); and Saunders (1967). The erythrocytic series. The basophilic eryth- rocyte is slightly oval in shape, containing a centrally located, round nucleus. The size and shape of the basophilic erythrocyte varies, de- pending on the stage of polychromasia. The cy- toplasm assumes a bluish-gray color and the nuclear fine chromatin architecture takes on a light purple color with both benzidine and Romanowsky staining. The cell size measures in the range of 9p. X 6|_i to lip X 8p, and the nucleus is 3p to 4p in diameter. As with the basophilic erythrocyte, the slightly oval-shaped reticulocyte also varies in size, measuring in the range of 6p X 5p to lOp X 8p, and its centrally located round nucleus measures 3p to 4p in diameter. The fine network of reticulum is clearly seen in a homogeneous pink cytoplasm, if the blood is mixed with a one percent solution of Brilliant Cresyl Blue before staining with Wright’s stain. The circulating normal mature erythrocyte has smooth margins and is predominantly ellip- soidal to oblong in shape and contains a cen- trally located round nucleus, 2p to 5p in diame- ter. The round erythrocyte measures from 8p to lOp in diameter and the ellipsoidal to oblong cell measures llpX7p to 13pXl0p. The homogeneous cytoplasm of this mature cell takes on a pale green color and the nuclear thick chromatin-interchromatin network assumes a dark blue to violet color with benzidine and Romanowsky staining. 74 New York Zoological Society: Zoologica, Summer, 1972 Senile or senescent erythrocytes are charac- terized by a loss of the smooth intact cell mem- brane and distended cytoplasm, which gives the celt its variability in size and shape. The cell measures 12p X lOp with an eccentrically lo- cated variably-shaped nucleus, 5p, to 7p in length. In the most advanced stage of degenera- tion, the nuclear membrane is no longer intact and its inner contents extend into the distended cytoplasmic portion of the cell. The cytoplasm takes on a pale green color and the nuclear clumped chromatin appears light blue or light purple when stained with benzidine and Wright’s stains. In the disintegrated erythrocytes (“nuclear shadows” or basket cells), the cytoplasm is no longer detected, and the cell therefore assumes a pale pink color when stained with benzidine and Romanowsky stains. An increase in the number of these cells may be due to either mechanical disruption during smear preparation or an increase in the fragility of erythrocytes which may be indicative of the numbers of se- nescent erythrocytes in the circulating blood. In some of the smear preparations highly re- tractile red serum granules. Ip in diameter were found either within the cytoplasm or along the cell membrane of the erythrocyte. Highly re- tractile vacuoles larger than Ip in diameter were also observed in cells of the same smear prepara- tions. This infrequent occurrence of cytoplasmic inclusions may have been a result of the smear preparation technique. Hemocytoblast or Hemoblast. In the circula- tion, this precursor cell measures from 8p to 12p in diameter. The centrally located large nucleus, 7p to 8p in diameter, with magenta-stained chromatin filaments and nucleoli, comprises al- most the entire volume of the cell. The cyto- plasm stains a deep blue with a Romanowsky stain. The hemocytoblast cell is the only direct derivative of the mesenchyme cell and differenti- ates into the leukogenic and erythrogenic cell series. Lymphocytes. The round lymphocyte varies in size from 5p to 7p in diameter. Its round deeply basophilic nucleus with condensed chro- matin comprises the entire volume of the cell. When treated with Romanowsky stains, the deep blue-purple to violet colored nucleus appears to be surrounded by a thin rim of a light pale blue cytoplasm. The irregular cellular outlines of cytoplasmic pseudopods characteristic of these cells are indicative of the lymphocyte’s relatively rapid locomotion in the circulation. Vacuoles were not observed in these cells, however, in some instances red-colored azurophilic granules, about Ip, in diameter, approximately 20 per cell, were observed in the cytoplasm. Thrombocytes. The brown bullhead contains elongated, round, and spindle or fusiform- shaped thrombocytes, some with pointed cyto- plasmic terminal processes at one or both ends of the cell. The nucleus is centrally located and varies in outline according to the shape of the entire cell. The dimensions of the cell vary for each type: elongated cell — lOp x 4p to 14p x 5p, nucleus 6p X 3p to 9p x 6p; round cell — 3p to 4p in diameter, nucleus 3p in diameter; spindle or fusiform-shaped cell — 5p X 3p to 9p X 3p, nucleus 4p X 3p. The very deep magenta to purple-stained compact nuclear chromatin is characteristic of the round and spindle shaped thrombocyte. In the elongated thrombocyte, the nuclear fine chromatin-inter- chromatin network takes on a magenta to purple color with Romanowsky stains. The nucleus of the thrombocyte in each of the above mentioned cells is surrounded by a homogeneous very pale blue to colorless cytoplasm, indicating the ab- sence of basophilia. No granules or vacuoles were observed in the cytoplasm in any of the smear preparations, and nuclear indentations of the mature thrombocyte were not a frequent oc- currence. Based on the fine network of nuclear chromatin and the slightly deeper blue cyto- plasmic color, however, contrary to other in- vestigators (Andrew, 1965), it is our belief that, in the brown bullhead, the elongated cell is the immature thrombocyte. Granulocytes. The neutrophil is the predomi- nant granulocyte in the circulating blood of the bullhead. It possesses an abundant clear pale blue to colorless cytoplasm surrounding an ec- centrically located polymorphic magenta-stained nucleus. The mature neutrophil ranges from lOp, to 17p in length, with a nucleus measuring from 4p to lOp in length. The loosely woven thread- like chromatin pattern of the nucleus may be round, kidney-shaped, ribbon-like, or more or less segmented in shape. Eosinophils do not appear to be a frequent occurrence in the circulation of all brown bull- heads. These large granulocytes, lOp to 15p in length, appear to be round to oval in shape. The relatively small, eccentric magenta-stained nu- cleus with clumped chromatin measures from 4p to 6p in length. The characteristic highly retractile eosinophilic red-orange granules, smaller than Ip in diameter, averaging about 25 per cell, are dispersed throughout the color- less cytoplasm. Monocytes. Monocytes in the brown bullhead vary in shape and measure 7p to 14p in length. The magenta-colored eccentrically-located nu- cleus is polymorphic in nature, ranging in shape from round, kidney, or bilobed and in size from 4p to 8p in length. The abundant cytoplasm of this mature circulating cell takes on a dull gray- blue color with Romanowsky stains. In a few Weinberg, Siegel, Nigrelli, & Gordon: Brown Bullhead Catfish 15 smear preparations, azurophilic granules, smaller than l|j, in diameter, appear in the cytoplasm. The occurrence of vacuoles in the cytoplasm is also infrequent. Macrophages. The macrophage in this species of catfish is a very large highly vacuolated cell, frequently containing cellular debris. The ap- pearance of the nucleus depends on the age of the cell. The very old cell contains a small dis- tinct mass of magenta-stained clumped chroma- tin. In the younger cells, measuring 27^ to 341 in length, the variably shaped nucleus is 6j.i to lOu in length. The cytoplasm in Romanowsky- stained blood smears takes on a dull gray color. Discussion The erythrocyte count, hematocrit, and hemo- globin values of the brown bullhead reported here compare favorably with those reported by Haws and Goodnight ( 1962), Table 2. Although there were no significant differences in the hema- tocrit or hemoglobin concentration values, sta- tistical analyses indicate a significant difference in red blood cell count, white blood cell count, reticulocyte count at 17 days after bleeding the seven individual fish (P ^ 0.05). There is also a trend of an increase in the MCV which is to be expected if there is a depletion of the reticu- locyte compartment resulting in a red blood cell population consisting predominantly of mature erythrocytes. The differences observed in the reported values cannot be attributed to any change in the temperature, oxygen level, or pH of the experi- mental tank, since these values did not alter significantly during the time interval in which the seven tagged fish were housed. The hematological values obtained for the group of randomly selected 25 fish are given in Table 1. The most striking difference is observed in the total white blood cell values. This dif- ference may be accounted for by the stressed condition to which these fish were subjected when the fish were individually caught for blood sampling. In fact, the differential counts of the individual fish of this group show a definite tendency towards a decrease in lymphocyte count and an increase in thrombocyte count as the fish were subjected to stress for a longer period of time (Table 3). Consequently, the mean differential values for the group of seven fish differ significantly from those for the group of 25 fish. Otherwise, the differential counts of both groups are comparable (Table 3). Studies with the killifish conducted by Pick- ford, et al. (1971a, 1971b, 1971c) have shown a definite correlation between stress and changes in the abundance of circulating leukocytes, as shown by alternating sequences of leukopenia and leukocytosis. The typical sequence of re- covery was described as follows: leukopenia at three min, leukocytosis at 15 min, leukopenia at 30 to 60 min, leukocytosis at 2 hrs, followed by a gradual return to normal. The only other differential analysis reported for catfish blood has been for the channel cat- fish species, Ictalurus pimctatus (Dodgen and Sullivan, 1969). Their findings are comparable with those reported in Table 3, in that the pre- dominant cell found in the peripheral circula- tion is the lymphocyte. Differential counts for other species of fish have also reported the lymphocyte as the pre- vailing white blood cell form, e.g., pike (Mul- cahy, 1970), goldfish (Watson and Shech- meister, 1963; Weinreb, 1963), and killifish (Pickford, et al. 1971a). This is in contrast to studies done by Saunders with 121 species of Table 1. The Hematological Parameters of the Brown Bullhead Catfish, Ictalurus nebulosus (Le Sueur). The values are given as the mean, plus or minus one standard error of the mean. The number of fish used in each group is shown in parenthesis. RBC lO^/cu mm WBC lO^/cu mm Reties % Hct % Hb gm/ 100 ml MCV CU fx MCH Pg MCHC % Group I (7) To 1.79 ±0.067 94.7 ± 3.6 13.0 ± 0.70 24.2 ± 0.59 8.14 ±0.39 136.0 ± 3.60 45.9 ± 3.3 33.1 ± 1.9 T 17 days* 1.47 ±0.116 71.9 ± 6.4 7.13 ± 1.19 23.2 ± 0.76 7.84 ±0.31 160.8 ± 8.11 54.2 ± 2.6 33.1 ± 1.0 T 27 days** 1.44 ±0.056 65.6 ± 8.9 4.92 ± 1.00 24.7 ± 0.93 6.42 ±0.21 172.2 ± 5.88 44.7 ± 1.8 25.7 ± 0.60 Group 11 (25) 1.69 ±0.051 46.2 ± 2.8 8.79 ± 1.02 30.2 ± 1.00 8.80 ±0.24 180.2 ± 5.18 53.1 ± 1.3 29.0 ± 0.34 * 17 days after first bleeding of 0.25 ml per fish. ** 27 days after first bleeding of 0.25 ml and 10 days after second bleeding of 0.25 ml. 76 New York Zoological Society: Zoologica, Summer, 1972 marine fish of Puerto Rico (1966a), 50 species of fish from the Red Sea (1968a), several species of elasmobranchs ( 1966b), and Gardner and Yevich (1969) with cyprinodontiform species; here the thrombocyte was found to be the predominant cell in differential counts of blood smear preparations. Summary A study was made of the hematological parameters and blood cell morphology of a normal population of brown bullhead catfish presented with the intent that these would serve as baseline figures for comparison with values obtained from fish that had been subjected to various forms of hypoxia and other stresses. The changes in hematological values obtained Table 2. Comparison of present results with THOSE OF Haws, et al. in the brown BULLHEAD CATFISH. Mean (Range) Erythrocyte counts: 10® cells/cu mm Weinberg, et al. 1.69 ( 0.144 - 1.74) Haws, et al. 1.22 ( 0.75 - 1.94) Hematocrit ( % ) Weinberg, et al. 27.2 (23.0 -31.2 ) Haws, et al. 27.9 (15.0 -47.0 ) Hb in gm/100 ml Weinberg, et al. 8.8 ( 7.6 -10.0 ) Haws, et al. 6.9 { 4.0 -10.0 ) Erythrocyte length (m) Weinberg, et al. ( 9.0 -13.0 ) Haws, et al. - (11.4 -15.9 ) Erythrocyte width (p) Weinberg, et al. ( 7.0 -10.0 ) Haws, et al. - ( 7.6 -11.4 ) as a result of blood letting indicated that re- covery from an induced state of anemia did not occur before 17 days post bleeding. Further research is being conducted on the hematological parameters and blood cell mor- phology of other species of fish, and the physio- logical control of blood cell formation in dif- ferent species of fish is presently being examined. Literature Cited American Fish Farmer 1970. Determination of sex of channel catfish. 1:15. Andrew, W. 1965. Comparative hematology. Grune & Strat- ton. Baldo, B. a., and B. Boettcher 1970. ABO(H) erythrocyte specific agglutinins in the serum of the Australian freshwater catfish, Tandanus tandanus (Mitchell). Aust. J. Exp. Biol. Med. Sci., 48:241-243. Bureau of Sport Fisheries and Wildlife 1970. Report to the fish farmers. U.S. Depart- ment of the Interior. Feb. Chlebeck, Anne, and Gary L. Phillips 1969. Hematological study of two buffalo fishes, Ictiobus cyprinellits and I. bubalus (Cato- stomidae). J. Fish. Res. Bd. Can., 26: 2881-2886. Chuba, Joseph, William J. Kuhns, and Ross F. Nigrelli 1968. The use of catfish, Ictalurus nebulosiis (Le Sueur), as experimental animals for immunization with human secretor saliva and other antigenic materials. J. Immun., 101 (l):l-5. Darragh Company 1970. Cultured catfish production. Little Rock, Arkansas. Table 3. Differential Counts of the Brown Bullhead Catfish. The values are given as the mean, plus or minus one standard error of the mean. The number of fish used in each group is shown in parenthesis. lympho- cytes % thrombo- cytes % neutro- phils % mono- cytes % macro- phages % HEMOCY- toblasts % EOSINO- PHILS % Group 1(7) To 67.4±4.5 23.6±3.0 6.9±2.5 1.3±0.3 0.08±0.05 0.58±0.30 0.00 T 17 days* 67.9±3.6 22.1±4.7 6.9±3.1 1.7 + 0.9 0.15±0.10 0.90±0.36 0.00 T 27 days** 66.5±6.2 24.8±6.1 6.1±1.9 1.0±0.3 0.98±0.84 0.45±0.18 0.00 Group II (25) 30.71±2.3 58.6±2.3 7.2±1.1 1.7±0.2 0.82±0.27 0.19 + 0.05 0.18±0.11 * 17 days after first bleeding of 0.25 ml per fish. ** 27 days after first bleeding of 0.25 ml and 10 days after second bleeding of 0.25 ml. Weinberg, Siegel, Nigrelli, & Gordon: Brown Bullhead Catfish 11 Dawson, Alden B. 1935. The hemopoietic response in the catfish Ameirus nebitlosus, to chronic lead poi- soning. Biol. Bull., 68 (3): 335-346. Dodgen, Charles L., and Sue Sullivan 1 969. Hematological effects of apholate on chan- nel catfish, Ictalurus punctatiis. Proc. Soc. Exp. Biol, and Med., 131:124-126. Dupree, Harry D. 1970. Personal communication. Gardner, George R., and Paul P. Yevich 1969. Studies on the blood morphology of three estuarine Cyprinodonitform fishes. J. Fish. Res. Bd. Can., 26:433-447. Haws, T. Glenn, and Clarence J. Goodnight 1962. Some aspects of the hematology of two species of catfish in relation to their habi- tats. Physiol. Zool., 35-36:8-17. Hesser, E. F. 1960. Methods for routine fish hematology. Prog. Fish. Cult., 22 (4): 164-171. Humason, Gretchen L. 1967. Animal tissue techniques (second edition). W. H. Freeman and Company, San Fran- cisco. Jakowska, Sophie 1956. Morphologic et nomenclature des cellules du sang des teleosteens. Rev. d’Hematol., 11:519-539. Katz, Max, and L. R. Donaldson 1950. Comparison of the number of erythro- cytes in the blood of hatchery-reared and wild salmon fingerlings. Prog. Fish Cult., 12(l):27-28. Klontz, George W., and Lynwood S. Smith 1968. Methods of using fish as biological re- search subjects. In Methods of animal experimentation, William Gay, ed. Aca- demic Press, New York, pp. 323-385. Kuhns, William J., Ross F. Nigrelli, and Joseph V. Chuba 1969. Differences between populations of fresh- water catfish defined by blood group anti- gens and antibodies. J. Immn., 103 (3): 454-459. Larsen, Howard N. 1964. Comparison of various methods of hemo- globin determination on catfish blood. Prog. Fish Cult., 26 (1):11-15. Larsen, Howard N., and S. F. Snieszko 1961a. Modification of the microhematocrit tech- nique with trout blood. Trans. Amer. Fish. Soc., 90 (2): 139-142. 1961b. Comparison of various methods of deter- mination of hemoglobin in trout blood. Prog. Fish Cult., 23 (1):8-17. Mulcahy, M. F. 1970. Blood values in the pike, Esox hiciiis L. J. Fish. Biol., 2:203-209. PiCKFORD, Grace E., Anil K. Srivastava, Anna M. Slicher, and Peter K. T. Pang 1971a. The stress response in the abundance of circulating leucocytes in the killifish Fiin- duliis heteroclitns. I. The cold shock se- quence and the effects of hypophysectomy. J. of Exp. Zool., 177 (l):89-96. 1971b. II. The role of catecholamines. J. of Exp. Zool., 177 (1 ):97-108. 1971c. III. The role of the adrenal cortex and a concluding discussion of the leucocyte- stress syndrome. J. of Exp. Zool., 177 (1) :109-117. Piper, Robert G., and Robert F. Stephens 1962. A comparative study of the blood of wild and hatchery-reared lake trout. Prog. Fish Cult., 24 (2):81-84. Ralston 1970. Catfish book. Ralston Purina Company. Saunders, Dorothy Chapman 1966a. Differential blood cell counts of 121 spe- cies of marine fishes of Puerto Rico. Trans. Amer. Micros. Soc., 85 (3):427- 449. 1966b. Elasmobranch blood cells. Copeia, 1966 (2) :348-351. 1967. Neutrophils and arneth counts from some Red Sea fishes. Copeia, (3):681-683. 1968a. Differential blood cell counts of 50 spe- cies of fishes from the Red Sea. Copeia, (3) :491-498. 1968b. Variations in thrombocytes and small lymphocytes found in the circulating blood of marine fishes. Trans. Amer. Micros. Soc., 87 (l):39-43. Simmons, Arthur 1968. Technical hematology. J. B. Lippincott Company, Philadelphia. Smith, Charles, William M. Lewis, and Harold M. Kaplan 1952. A comparative morphologic and physio- logic study of fish blood. Prog. Fish Cult., 14 (4):169-172. Srivastava, Anil K. 1968. Studies on the hematology of certain freshwater teleosts. I. Erythrocytes. Anat. Anz. Bd., 123:233-249. Watson, L. L, I. L. Shechmeister, and L. L. Jackson 1963. The hematology of the goldfish, Carassius auratus. Cytologia, 28 (2): 118-130. Weinreb, Eva Lurie 1963. Studies on the fine structure of teleost blood cells. I. Peripheral blood. Anat. Rec., 147 (2):219-238. 78 New York Zoological Society: Zoologica, Summer, 1972 Wiener, A. S., J. V. Chuba, E. B. Gordon, and J. W. Kuhns 1968. Hemoglutinins in the catfish (Ictaliirus nebiilosiis) injected with saliva from hu- man secretors of various ABO blood groups. Transfusion, 8:226-234. WiNTROBE, Maxwell M. 1967. Clinical hematology. Lea & Febiger, Phila- delphia. Yuki, Ryogo 1957. Blood cell constituents in fish. I. Peroxi- dase staining of the leucocytes in rainbow trout (Salmo irideus). Bull. Fac. of Fish, 8 (l):36-44. 1958. Blood cell constituents in fish. II. Peroxi- dase staining of leucocytes in the kidney and some other organs. Bull. Fac. of Fish. Hokkaido Univ., 8 (4) :264-267. 1960. Blood cell constituents in fish. IV. On the “nuclear shadow” found in blood smear preparations. Bull. lap. Soc. Sci. Fish, 26 (5):490-495. 1963. A comment on microscopic examination of the blood cell constituents in fish. Bull. Jap. Soc. Sci. Fish, 29 (1): 1098-1 103. 4 Histochemical Analyses of the Fluid and the Solid State of the Adhesive Materials Produced by the Pre- and Postmetamorphosed Cyprids of Balanus eburneus Gould (Figures 1-6; Tables 1-10) Paul J. Cheung and Ross F. NigrellF Osborn Laboratories of Marine Sciences, New York Zoological Society, Brooklyn, New York 1 1224 Two distinct eosinophilic zones of granules found in the larval (cyprid) cement gland of Balanus eburneus Gould are referred to as medial and inner granules, according to their morphological position within the gland. No such differentiation of the secretory materials are present in the adult cement cells. Histochemical tests show that the granules, which represent the fluid state of the cement material, in the cyprid and adult, are basic protein, characterized as collagenous substances. The hardened cement produced by the cyprid and the adult that is found on the basal plate are histochemically unreactive protein masses, but the two substances are not histochemically identical. Introduction IN RECENT YEARS, studies on bamacles have been concerned with the cement, its syn- thesis, and the cement-producing complex. Thus,, the general morphology of the cement gland in the cyprid stage was reported by Bernard and Lane (1962), the histology by Walley (1969), and the ultrastructure of the gland and the attachment pads of the anten- nules by Nott ( 1970), Nott and Foster ( 1970), and more recently by Walker ( 1971 ). The histology of the cement apparatus of several species of adult barnacles was described in detail by Lacombe (1966, 1967, 1970), Lacombe and Liguori (1969), and at the ultra- structural level, by Walker (1970). Costlow ( 1959) demonstrated the presence of carbonic anhydrase in tbe sbell-forming tissue of tbe adult barnacle. Arvy and Lacombe ( 1968), Arvy, Lacombe, and Shimony (1968), and Shi- mony and Nigrelli (1971a, b) reported the pres- ence of succinic dehydrogenase, alkaline phos- photase, arylsulphatase, and polyphenoloxidase, respectively, in the cement and cement appara- tus of the adult barnacle, while Walker (1971) demonstrated polyphenoloxidase in both the gland and in the secreted cement of the cyprid stage of Balanus balanoides. On the basis of histochemical techniques, various opinions were presented as to the chemi- cal nature of the barnacle cement. Hillman and ’Supported by ONR Contract N-00014-68-C-0334. Nace (1970) and Shimoy (1971) concluded that the cement laid down by the attached cyprids was collagen, an opinion with which we concur in this report. Walker (1970, 1971) and Saroyan, et al (1970a) stated that the cement substances of the cyprid and the adult were pro- teins with phenolic groups, while Lacombe (1968) concluded that the cement in the adult barnacle consisted of acid mucopolysaccharides. These reports show the confusion in defining the chemical nature of the cement at a particular stage. The present paper deals with histochemi- cal analyses of the fluid and solid states of the adhesive materials produced by the pre- and post-metamorphosed cyprids of Balanus ebur- neus Gould. Materials and Methods Adult barnacles were collected locally or ob- tained from Biscayne Bay (Florida), attached to aluminum plates, and air shipped the same day to New York. The cyprids were raised in the laboratory and allowed to metamorphose to the barnacle stage in beakers. The glass around each barnacle was then cut in such a way that it served as a slide for microscopical examina- tions following treatment for various histo- chemical reactions. Cyprids and barnacles were fixed in 10% buffered formalin, Bouin’s fluid, Heidenhain’s Susa, Zenker’s, and Carnoy’s solutions for gen- eral histological and histochemical observations. Where necessary, 5% nitric acid was used for 79 80 New York Zoological Society : Zoologica, Summer, 1972 decalcification. After fixation and decalcifica- tion the animals were embedded in paraplast and sectioned at 6 p. The various histochemical methods employed in Tables 1-5. in this study are summarized Table 1. General Histological and Histochemical Methods Used to Demonstrate Various Components of the Cement Apparatus of the Barnacle, Balanus eburneus Gould Methods Fixatives References Purposes Hematoxylin and eosin 10% neutral formalin Thompson ( 1966) General structures; stain for acidic and basic com- ponents of the tissue Azure A and eosin B pH 3.7-4.3 10% neutral formalin Conn et al ( 1960) General structures Mallory’s phloxine methylene blue 10% neutral formalin Mallory (1938) Stain for collagen Masson’s Trichromes 10% neutral formalin Masson ( 1929) Stain for collagen and intracellular fibers Gomori’s Trichromes 10% neutral formalin Gomori ( 1950) Stain for collagen and connective tissues Phosphotungstic acid hematoxylin 10% neutral formalin and Susa Lillie (1954) Stain for collagen Mallory’s collagen stain aniline blue and orange G Zenker’s fluid Mallory ( 1936) Stain for collagen Table 2. Methods Employed to Demonstrate Carbohydrate Components and Metachromasia OF THE Barnacle Adhesive (s) Methods Fixatives References Purposes Periodic Acid Schiff (PAS) Susa and 10% neutral formalin Thompson (1966) Demonstration of adjacent glycol or amino-hydroxy groupings PAS + Acetylation 10% neutral formalin Lillie ( 1954) Blocking 1,2-glycols and 1,2-aminohydroxy groups in oxidative Schiff reaction PAS + Saponification 10% neutral formalin Lillie (1954) Deacetylation PAS + Saliva digestion 10% neutral formalin Lillie (1949) Removal of glycogen and RNA Mucicarmine 10% neutral formalin and Carnoy’s fluid AFIP (1957) Demonstration of acid muco- polysaccharide (mucin) of epithelial origin Bast’s Carmine Susa and Carnoy’s fluid Thompson (1966 ) Demonstration of glycogen Azure A 0. 1 % and 0.0 1 % pH 3.9 Susa and Carnoy’s fluid Thompson ( 1966) Demonstration of metachromasia Methenamine silver nitrate 10% neutral formalin Gomori ( 1946 ) Demonstration of glycogen and mucin Aldan blue pH 2.8 Susa and Bouin’s fluid Thompson ( 1966) Demonstration of sulfated acid mucopolysaccharide Ribonuclease + Aldan blue pH 2.8 Susa and Bouin’s fluid Thompson (1966) Hydrolysis of RNA Aldan blue pH 2.8 + sulfation Susa and Bouin’s fluid Thompson (1966) Demonstration of meta- chromasia by esterification of carbohydrates Toluidine blue 0, at 0.01% pH 2.8 + sulfation Susa and Bouin’s fluid Thompson (1966) Demonstration of meta- chromasia by esterification of carbohydrates Cheung & Nigrelli: Histochemical Analyses of Balanus eburneus Gould 81 Table 3. Methods Employed to Demonstrate Nucleic Acids in the Barnacle Adhesive(s) Methods Fixatives References Purposes Methylene blue pH 3.0 Susa’s fluid Stenram ( 1953 ) Demonstration of nucleic acids; RNA and DNA Ribonuclease + Methylene pH 3.0 Susa’s fluid Stenram ( 1953 ) Enzymatic hydrolysis of RNA Toluidine blue 0, 0.5% pH 3.0 Susa’s fluid Stenram ( 1953 ) Demonstration of nucleic acids; RNA and DNA Ribonuclease + Toluidine blue 0, 0.5%, pH 3.0 Susa’s fluid Stenram (1953 ) Enzymatic hydrolysis of RNA Nucleal Feulgen Reaction Susa’s fluid Thompson (1966) Demonstration of DNA Table 4. Methods Employed to Demonstrate Lipids and Unsaturated Fats in the Barnacle Adhesive (s) Methods Fixatives References Purposes Sudan black B Unfixed tissue and 10% neutral formalin Thompson ( 1966) Demonstration of liquid, and semi-solid fats in tissues Sudan black B + acetone Unfixed tissue and 10% neutral formalin Thompson ( 1966 ) As control Plasmal Reaction 10% neutral formalin Hayes (1949) Demonstration of unsaturated fats Luxol fast blue 10% neutral formalin Thompson ( 1966 ) Demonstration of phospholipids Table 5. Methods Employed to Demonstrate Protein and Amino Acids in the Barnacle Adhesive(s) Methods Fixatives References Purposes Tetrazotized benzidine Bouin’s Fluid Lillie ( 1957) Demonstration of proteins with j3-naphthol 10% formalin in general Ninhydrin Schiff Susa’s fluid Yasuma et al (1953) Demonstration of the sites of free a-amino acids 2,2'-dihydroxy-6, Susa’s fluid Thompson ( 1966 ) Demonstration of the 6'-dinaphthyl 10% formalin sulfhydryl group and disulfide and thioglycolic Carnoy’s fluid disulfide linkages acid (DDD) DDD without thioglycolic acid Susa’s fluid Thompson ( 1966 ) Demonstration of the sulfhydryl group DDD with benzoyl Susa’s fluid Thompson ( 1966 ) Demonstration of the chloride 10% formalin disulfide linkages Mercury Orange Susa’s fluid Carnoy’s fluid Thompson ( 1966) Demonstration of the sulfhydryl group 8-hydroxyquinoline Susa’s fluid Lillie (1957) Demonstration of amino acid arginine p-dimethylaminobenz- Susa’s fluid Adams (1957) Demonstration of indole aldehyde nitrite 10% formalin derivative (tryptophan) Diazotization with 8- 10% formalin Glenner and Demonstration of tyrosine amino- 1 -naphthol-5- Lillie (1959) and phenolic compounds sulfonic acid Millon’s Reagent 10% formalin Thompson ( 1966) Localization of tyrosyl groups in tissue sections 82 New York Zoological Society : Zoologica, Summer, 1972 Description of the Cement Apparatus^ The Cyprid Cement Gland The cement glands in the cyprid of Balanus eburneous, as in the other species of the Bala- nidae, are paired kidney-shaped structures with an average measurement of 70 X 50 X 70 mi- crons. Each gland consists of a number of secre- tory cells arranged as compartments. Each sec- retory cell (Figure 1) contains a large amount of two histochemically distinct secretory gran- ules and a peripherally located nucleus. These granules referred here as medial and inner gran- ules according to their relative position within the cement gland of the cyprid. The a- and /?-cells reported by Walker ( 1971 ) are not easily discerned with the light microscope. However, the medial and the inner granules are probably the two types of electron-dense bodies found in the a-cells. The collecting canal arises within the medullary portion of the gland and then passes into the conducting canal through the antennule to the attachment pad (Figure 2). The Adult Cement Apparatus The cement secreting cells in the adult bar- nacle are not enclosed in a compact glandular structure; instead the cells are scattered in the mantle tissue mainly along the lateral axis with the associated canals. The cement cells are large, up to 140 microns in diameter, with unusually large polymorphic nuclei containing clumps of chromatin material. The cytoplasm has evenly distributed granules and densely staining secre- tory zones. No vacuoles were noted at the se- cretory zones (Figure 3). The smallest cement cell with secretary zones measures 14 microns in diameter; the circular nucleus contains a cen- trally located nucleolus. The collecting canal makes its contact with the cement cell at secre- tory zone (Figure 4), and the conducting canal links the individual cement cells into grape-like clusters. The intracellular cement substance ac- cumulated at the secretory zones is delivered to the exterior through the canal system in the basal plate. The Basal Plate Structures The basal plate of Balanus eburneiis Gould is a complicated structure. At the center of the basal plate, the region of the initial point of attachment of the cyprid, the hardened cement consists of two spots approximately 18 microns in diameter, with the remains of the segment IV and the attachment pads of the antennules at- tached to the substrate. The initial cement of the adult barnacle, located anteriorly to the cyprid cement, also appears as two spots, approxi- mately 35 microns in diameter, with the broken portion of the conducting canal remaining at the ’ A glossary of terms used in this text is given in the appendix. center of each spot (Figure 5). Two radial canals are formed by branching along the lateral axis from the conducting canals at the center of the base with circular canals growing out from the radial canals. The adhesive substance is delivered to the edge of the growing plate through the openings in the circular canals. Fig- ure 6 is a composite schematic representation of the cement apparatus of the pre- and post- metamorphosed cyprid at the region of initial attachment. Histological and Histochemical Reactions The results of the histological and histochemi- cal reactions on various tissue and cellular com- ponents of the cement apparatus of the pre- and post-metamorphosed stages of the barnacle, Balanus ebureus Gould, are summarized in Tables 6 to 10. The medial and the inner granules of the cyprid cement secreting cells show differences in (1) degree of acidophilia; (2) permeability to orange G and hematin-lake formation; and ( 3 ) intensity in reaction with carbohydrate and amino acid detecting agents. The granules in both locations have similar histochemical reactions and stained positively with a number of collagen identifying agents. The medial and the inner granules are charac- terized by the presence of ( 1 ) tryptophan, argi- nine, and a number of a-amino acids; (2) small amount of lipid and phospholipid substance; (3) no unsaturated fats; (4) carbohydrate com- ponents of small molecular size; and (5) sulfhy- dryl group and disulfide linkage. These two types of granules probably represent stages in synthesis. The cement material at the secretory zones of the adult cement cell does not differentiate into two acidophilic zones. This fluid cement is also identified as a collagenous substance. The car- bohydrate reactions are less intense in compar- ing with the cyprid inner granules. A small amount of lipid and phospholipid substances are present. Neither unsaturated fats nor ribonucleic acid are found in the cement. No metachromasia is demonstrated. Tyrosine is found in the adult cement but absent in the cyprid of Balanus eburneus. The hardened cyprid cement, i.e. the cyprid antennular deposits at the center of the basal plate, is not stained by the acid and basic dyes and is nearly histologically and histochemically inert. However, the free adjacent glycol or amino-hydroxy groupings can still be detected with PAS. Furthermore, the cement in the hard- ened state is apparently a protein mass but no specific amino acids could be demonstrated when treated by the usual chemical agents. This observation is in agreement with the histochemi- Cheung & Nigrelli: Histochemical Analyses of Balanus ehnrneus Gould 83 cal analyses of the cyprid hardened cement by Hillman and Nace (1970). The hardened adult barnacle cement appears as rings associated with the circular canals on the basal plate, which becomes histochemically non-reactive but has the characteristic of a basic protein. It differs from the hardened cyprid cement by the following characteristics: (1) posi- tive alkaline nature; (2) PAS negative; and (3) the presence of tyrosyl groups. Figure 1. A section through the cyprid cement gland showing the cortically located secretory cells and the collecting canal at the medullary portion of the gland; the cytoplasmic granules are colorless with the tribastic stains. Col C, collecting canal; Se Ce, secretory cells; Ep, epithelial cells. (43 X). 84 New York Zoological Society: Zoologica, Summer, 1972 Figure 2. A composite schematic representation of the cyprid cement apparatus showing the relative position of the medial and the inner granules within the gland; Nu, nucleus of the secretory cell; Cyt, cytoplasm; Me G, medial granules; In G, inner granules; Col C, collecting canal; Con C, conducting canal; Ant, antennule; Att P, attachment pad. Cheung & Nigrelli: Histochemical Analyses of Balanns ebiirneus Gould 85 Figure 3. A section through an adult cement cell of Balanus eburneus Gould; the cytoplasm is grayish and the three secretory zones are pink with Mallory’s phloxine stains; Ch, chromatin materials; Col C, collecting canal; Se Z, secretory zone; Cyt, cytoplasm; Nu, nucleus. ( 100 X). n V 86 New York Zoological Society: Zoologica, Summer, 1972 Figure 4. A schematic representation of the adult barnacle cement cell associated with the collecting and the conducting canals; Se Z, secretory zone; Nu, polymorphic nucleus; Ch, chromatin materials; Col C, collecting canal; Cyt, cytoplasm; Con C, conducting canal. Cheung & Nigrelli: Histochemical Analyses of Balanus ebiirneus Gould 87 o Figure 5. The area of initial attachment on the basal plate of Balanus eburneus Gould; the cyprid cement is colorless while the adult barnacle cement and the radial canal are pink when treated with Millon’s Reagent; Ant, segment IV of the antennule; Cy Cm, cyprid ceme ntwith the antennular attachment pad; In Cm, initial cement of an adult barnacle; Ra C, radial canals branching from the center of the basal plate. (43 X). 88 New York Zoological Society: Zoologica, Summer, 1972 a On Figure 6. A composite schematic representation of the area of the initial attachment made by the barnacle showing the relationship between the cyprid cement gland and the adult cement apparatus; Cm R, cement ring; Cm App, cement apparatus; Ci C, circular canal; Y Cm Ce, young cement cell; Ad Cm, adult cement; Ra C, radial canal; Cy Cm App, cyprid cement apparatus; Cy Cm, cyprid cement. Cheung & Nigrelli: Histochemical Analyses of Balanus eburneus Gould .3 .3 a a ^ -2 S o ’o '5 a o a d M a 'O *o o> *a i-i o 'a ‘a t-4 (-1 3 U a c o T3 T3 a u U 3 ^ JD 00 c CO *3 .3 d •!> 3 o 3 o ‘a cd d cd ,d o (U TS ‘S 3 O X) -d o> -d u U jd o *n *3 .C S)"?, a CS a d H H 3 3 o X o < CO Vi S' U( Jd ^co "d ^co *c O ja c -1 2 ® X3 3 a 3 S u a 3 o CO CO cd 0 a o .3 3 f/i d o> oo a a II aniline blue and orange G Table 7. Histochemical Tests for Carbohydrates in the Barnacle Adhesives(s) 90 New York Zoological Society: Zoologica, Summer, 1972 a U c a li o .S 2 E CO •U c a a c cd .s *E *u *E t-i a E a 2 S 3 Oh 2 (J Q o 0) § o N 0> ^ 2 2^ o 2^22 *5 E *1^ 2 -2 2 o o o O O CJ c cd ’o CO 2 w E o c a c c c 2 o .IS 0> E .a o ^ o U c Q- a 'd o> T3 .a 2 ^ G 2 E J=) HP E S -2 *H cd i;' o > u & o cd cd cd CO CO + + + fe-s <- < < a CO a a a CQ O d T3 c cd O cn < X o ^ <4.1 3 cd d 2 3 ■“ AS 3 y y s 3 3 . ^ XI h-l 0) ^ 4> 3 g 3 X) -2 x> _3 3 Cd 2 2 E X) X) ^ 2i H s 3 3 . X3 X) X3 3 2 2 3 -o o> > o> > c G cd C 0) 2 "S s cd 00 ri ffi + a ^ Q < 1) 3 d a O vy ^ 3 5 £ o O o> £§ cj N V on a o >> U i/3 H3 O M QJ S c o Oi) PU C/5 o •S s Sudan black B colorless blue L. black L. black colorless colorless L. black colorless Sudan black B + Acetone colorless colorless colorless colorless colorless colorless colorless colorless Plasmal Reaction L. pink colorless L. pink colorless colorless colorless colorless colorless Luxol fast blue colorless blue colorless colorless blue blue blue blue Table 10. Histochemical Tests for Proteins and Amino Acids in the Barnacle Adhesive(s) 92 New York Zoological Society: Zoologica, Summer, 1972 CJ U T3 •C >. o o D O i •■2 g D OO "a Ui & & a s tJU u G cd u a *Sh 3 a G Oh 3 a U 0 h4 J § s o &N _ D C Q < » U c b c '5. c .S u* c G Oh 0) CiO G cd 00 G cd oo G cd 2 Oh U G Oh 00 G cd a> 00 G cd 4> OO G cd 2 xi tn 2 D 00 G 3h G 'Eh G U l-l l-i G o G O Oh G Oh X) J o> oo G cd D 00 G cd O 3 00 G cd D JS C/5 D XJ t/5 D D c/5 C/5 D 00 00 00 00 OO D c G ca G G c G G Th cd Cd cd cd cd 0 w u E o o kH u O O Ui O u O O 3 X5 X) o •J s' o T ■5 ^ U N -S « a -SR SG u O ^ DO y • ^ <0 MH ^ Oi Q 5-sulfonic acid for tyrosine Millon’s Reagent L. orange orange colorless colorless colorless colorless red orange for tyrosine Cheung & Nigrelli: Histochemical Analyses of Balanus eburneus Gould 93 Discussion Histological observations on serial sections of a newly metamorphosed barnacle show that the adult cement apparatus is not derived from the remnant of the cyprid cement gland. Therefore, these two cement producing organs have differ- ent origins and developments. Based on the mor- phological differences of the cement apparatus, it is assumed that the adhesive substances pro- duced by the pre- and post-metamorphosed cyprid are different. Therefore, it should be emphasized that our histochemical studies deal with the following; (1) the intracellular fluid material of the cyprid cement gland; (2) the “hardened” cement at the points of antennular attachment of the cyprid; (3) the intracellular and intracanicular fluid ce- ment of the post-metamorphosed barnacle; and (4) the “hardened” cement of the initial and subsequent deposits produced by the adult. Further, it should be recognized that our studies deal mainly with the surface chemistry of the so-called “hardened” cement of the cyprid and adult, which may account for some of the differences in the histochemical reactions noted by us and those reported by other investigators (Walker 1970, 1971; Hillman and Nace, 1970; and Saroyan, et al, 1970a). The histochemical analyses of the fluid and the solid state of the cement suggest that the materials are collagenous which undergo a tran- sition from a highly chemically reactive (fluid cement) to an unreactive mass upon hardening. The fluid cement of the cyprid is present in the treated material as medial and inner gran- ules characterized histochemically by differences in the numbers of free cationic groups as re- flected by differences in the degree of acidophilia of these two granules. It is our belief that these granules are the same as those reported by Walker (1971) as two types of electron-dense bodies within the a-cells of the cyprid cement gland. It further leads us to infer that the inner granules (i.e., more electron-dense bodies) are probably a polymerized state of the medial gran- ules (i.e., less electron-dense bodies) , which may represent the monomeric state of the adhesive. The above interpretations may also explain the difference in the intensity of most of the his- tochemical reactions of these two types of granules. All of the histochemical tests, as mentioned above, indicate that the fluid cement of both the pre- and post-metamorphosed cyprids is a col- lagenous substance. The presence of sulfhydryl groups and disulfide linkage indicates that the structure of the cement has a compactly coiled and/or folded configuration. The presence or absence of tyrosine in the cement material raises some questions. Our re- sults showed that this amino acid is not present in the fluid and hardened cement of the cyprid (see also Hillman and Nace, 1970), but is found in the cement of the adult (Figure 5). This is contrary to that reported by Saroyan et al (1970a) and by Walker (1970, 1971) who demonstrated tyrosine in the cement of both the adult and cyprid. The significance of the lack of the demonstrable tyrosyl group in the cyprid cement is obscure. The interpretation that the cement is an acid mucopolysaccharide, as has been suggested by several investigators, can be ruled out by the absence of metachromasia with or without prior sulfation indicating that the carbohydrate com- ponent is of small molecular size with few ani- onic and hydroxyl groups. The fluid cement undergoes considerable changes during the hardening process which, in our observations, appears to take place at the site of cement-substratum junction. The mecha- nism(s) of hardening has not been firmly estab- lished. There are some indications, however, that cement hardening agents and/or enzymes are present in the mantle tissue (Shimony and Nigrelli, 1971b). Conclusion and Summary On the basis of the selective histochemical tests employed in these studies, it is concluded that the fluid cement (intracellular and intra- canicular) in the cyprid and adult stages of the barnacles are collagenous substances. The fluid cement produced by the gland cells of the adult barnacle differs from that formed in the cyprid by the presence of tyrosine, and by a less intense color reaction for carbohydrates. The signifi- cance of these differences remains obscure at this time. The antennular deposits of the cyprid and cement particles in the basal plate produced by the adult barnacle are referred to as solid or hardened cement. The hardened cyprid cement is PAS-positive but non-reactive to acidic and basic dyes; the hardened adult cement is PAS- negative and is stained readily with acidic dyes and, as would be expected, gives a positive reac- tion for the presence of tyrosyl groups. Appendix 1. Medial Granules: stainable materials found proximal to the nucleus of the secretory cells of the cyprid cement gland. 2. Inner Granules: acidophilic granules found dis- tal to the nucleus of the secretory cells of the cyprid cement gland. 3. Secretory Zones: the acidophilic zones in the adult cement cells where the fluid cement is accumulated for secretion. 94 New York Zoological Society: Zoologica, Summer, 1972 4. Collecting Canal: the canal that penetrates the medullary portion of the cyprid cement gland or the canal that is associated with the adult cement cells at the secretory zones. 5. Conducting Canal: the canal that passes within the cyprid antennule and terminates at the at- tachment pad; in the young barnacle, it is the canal that links the collecting canals of indi- vidual cement cells and runs perpendicular to the radial and circular canals. 6. Radial Canals: the two largest canals on the basal plate extending radially along the lateral axis from the center of the basal plate where the adult cement initiates. 7. Circular Canal: the canal that branches out of the radial canal and forms rings on the basal plate. It delivers cement to the exterior through the orifices along the canals. 8. Cyprid Antennular Deposits: the hardened ce- ment which is deposited by the cyprid at the initial point of attachment. Literature Cited Adams, C. W. M. 1957. A p-dimethylaminobenzaldehyde nitrite method for the demonstration of trypto- phane and related compounds. J. Clin. Path., 10: 56-62. Armed Forces Institute of Pathology 1957. “Manual of Histologic and Special Stain- ing Technics,” Washington, D.C., 136-137. Arvy, L., and D. Lacombe 1968. Activites enzymatiques traceuses dans I’appareil cementaire des Balanidae (Crus- tacea, Cirripedia). C. R. Acad. Sci. Paris, 267: 1326-1328. Arvy, L., D. Lacombe, and T. Shimony 1968. Studies on the biology of barnacles: alka- line phosphatase activity histochemically detectable in the cement apparatus of the Balanidae (Crustacea-Cirripedia). Am. Zool., 8: 817. Bernard, F. J., and C. E. Lane 1962. Early settlement and metamorphosis of the barnacle Balanus amphitrite niveus Darwin. J. Morph., 110: 19-39. Conn, H. J., M. A. Darrow, and V. M. Emmel 1960. Staining procedures used by the Biological Stain Commission, 2nd Ed., Baltimore, The Williams and Wilkins Company. 289 pp. COSTLOW, J. D. 1959. Effect of carbonic anhydrase inhibitors on shell development and growth of Balanus improvisus Darwin. Physiol. Zool., 32: 177-184. Glenner, G. G., and R. D. Lillie 1959. Observations on the diazotization-coupling reaction for the histochemical demonstra- tion of tyrosine; metal chelating and for- mazan variants. J. Histochem and Cyto- chem., 7: 416-422. Gomori, G. 1946. A new histochemical test for glycogen and mucin. Tech. Bull. Reg. Med. Technol., 7: 177-179. 1950. A rapid one-step trichrome stain. Am. J. Clin. Path., 20: 661-664. Hayes, E. R. 1949. A rigorous redefinition of the plasmal reaction. Stain Technol., 24: 19-23. Hillman, R. E., and P. F. Nace 1970. Histochemistry of barnacle cyprid adhe- sive formation. In Adhesive in Biological System. R. S. Manly [Ed.], Academic Press, N.Y., 302 pp. Lacombe, D. 1966. Glandulas de cimento e sens canais em Balanus tintinnabulum (Cirripedia-Bala- nidae), pp. 1-39. In Publicacao Instituto de pesquisas da Marinha, Nota technica No. 32, Ministerio da Marinha, Rio de Janeiro, Brazil. 1967. Histoquimica e histofotometria das glan- dulas de cimento de Balanus tintinnabu- lum (Balanidae-Cirripedia), pp. 1-29. In Publicacao No. Oil do Instituto de Pes- quisas da Marinha, Ministerio da Marinha, Rio de Janeiro, Brazil. 1968. Histologia, histoquimica e ultra estrutura das glandulas de cimento e sens canais em Balanus tintinnabulum. pp. 1-22. In Publi- cacao No. 017 do Instituto de Pesquisas da Marinha, Ministerio da Marinha, Rio de Janeiro, Brazil. 1970. A comparative study of the cement gland In some Balanid barnacles (Cirripedia, Ba- lanidae). Biol. Bull., 139: 164-179. Lacombe, D., and V. R. Liguori 1969. Comparative histological studies of the cement apparatus of Lepas anatifera and Balanus tintinnabulum. Biol. Bull., 137(1): 170-180. Lillie, R. D. 1957. Histopathologic Technic and Practical Histochemistry. The Blakiston Co., Inc., N.Y. 560 pp. Mallory, F. B. 1936. The aniline blue collagen stain. Stain Technol., 11: 101-102. 1938. Pathological Technique. W. B. Saunders Co., Philadelphia. 430 pp. Masson, P. J. 1929. Some histological methods: Trichrome stainings and their preliminary technique. J. Tech. Methods, 12: 75-90- Cheung & Nigrelli: Histochemical Analyses of Balanus eburneus Gould 95 Nott, J. a. 1969. Settlement of barnacle larvae: Surface structure of the antennular attachment disc by scanning electron microscopy. Mar. Biol. 2: 248-251. Nott, J. A., and B. A. Foster 1969. On the structure of the antennular attach- ment organ of the cypris larva of Balanus balanoides (L. ) Phil. Trans. Roy. Soc. Lond.,B256: 115-134. Pearse, a. G. E. 1953. Histochemistry, Theoretical and Applied. Little Brown and Co., Boston. 530 pp. Saroyan, J. R., E. Lindner, C. A. Dooley, and H. R. Bleile 1970a. Barnacle cement — Key to second genera- tion anti-fouling coatings. Ind. Eng. Chem. Prod. Res. Develop., 9(2): 121-133. Saroyan, J. R., E. Lindner, and C. A. Dooley 1970b. Repair and reattachment in the Balanidae as related to their cementing mechanism. Biol. Bull., 139: 333-350. Shimony, T. 1971. Biochemical and histochemical studies of the enzyme arylsulphatase in the mantle of the barnacle Balanus eburneus Gould. Ph.D. Thesis, New York University Grad- uate School of Arts and Science. Shimony, T., and R. F. Nigrelli 1971a. Studies on arylsulphatases in the barnacle, Balanus eburneus Gou\d. I. Enzymatic and physiological properties. Marine Biology, (in press). 1971b. Phenoloxidase activity in the cement ap- paratus of the adult barnacle, Balanus eburneus Gould, and its possible function in the hardening process of the adhesive substances. Amer. Zool., (in press). Spicer, S. S., and R. D. Lillie 1961. Histochemical identification of basic pro- teins with Biebrich scarlet at alkaline pH. Stain Technol., 36: 365-370. Stenram, U. 1953. The specificity of the gallocyanin-chroma- lum stain for nucleic acids as studied by the ribonuclease technique. Exp. Cell Res., 4: 383-389. Yasuma, a., and T. Ichikawa 1953. Ninhydrin-Schiff and alloxan-Schiff stain- ing. Lab. and Clin. Med., 41: 296-299. Walker, G. 1970. The histology, histochemistry, and ultra- structure of the cement apparatus of three adult sessile barnacles, Elminius modestus, Balanus balanoides, and Balanus hameri. Mar. Biol., 7: 239-248. 1971. A study of the cement apparatus of the cypris larva of the barnacle Balanus bala- noides. Mar. Biol., 9: 205-212. Walley, L. j. 1969. Studies on the larval structure and meta- morphosis of Balanus balanoides L. Phil. Trans. Roy. Soc. Lond., B256: 237-280. 5 Blood-Group Activity in Baboon Tissues (Tables 1-4) Joseph V. Chuba, William J. Kuhns, and Ross F. Nigrelli Department of Pathology, New York University School of Medicine, New York, New York 10016, and The Osborn Laboratories of Marine Sciences, New York Zoological Society, Brooklyn, New York 11224 Stomach, salivaryrgland, pancreas, and skeletal-muscle tissues from a group-A baboon were extracted with 0.9% saline and 99% ethanol. Only saline extracts from stomach and salivary-gland tissues displayed significant blood-group (group A) activity. Both saline- and ethanol-extracted human group-Aj erythrocyte stromata, baboon stomach, and baboon salivary-gland tissues displayed similar anti-A-absorption potency in quanti- tative antibody-absorption-capacity tests. The findings are discussed from the point of view of biochemical evolution, as well as their potential importance in organ-transplanta- tion and cross-circulation procedures. Introduction The a-b-o blood-group status of virtu- ally all of the nonhuman primates has been extensively investigated by Dr. Alex- ander S. Wiener and his colleagues (Wiener and Moor-Jankowski, 1970). In addition to the elu- cidation of many fundamental questions regard- ing the phylogeny of blood groups, such studies have provided the immunohematological basis for the recently developed technique of cross- circulation therapy (Hume et al., 1969). With this supportative procedure, it has been possible to re-establish homeostatis in human patients (e.g., hepatic coma cases) by utilizing the normal functional capacity of nonhuman primate organs appropriately exchange trans- fused in advance with compatible human blood. “Consanguinity” at the level of interprimate A-B-O compatibility appears to be the only ma- jor tissue-matching prerequisite for this type of supportative therapy. In this regard, group-A and group-B baboons are readily available, and group-O baboons, although of apparently very limited frequency in nature (Wiener and Moor- Jankowski, 1969), could undoubtedly be selec- tively bred in captivity for cross-circulation pro- cedures requiring group-O compatibility. One of the unique aspects of A-B-O blood- group expression among the Old World monkeys (e.g., baboons, gelada, and rhesus monkeys), however, is the fact that virtually without excep- tion, A-B-O-active antigens are not detectable as erythrocyte agglutinogens in these primates. Thus A-B-O phenotypes cannot be established directly on the basis of hemagglutination tests (Wiener and Moor-Jankowski, 1970) . As a con- sequence, A-B-O typing of Old World monkeys, which invariably are blood-group-substance “se- cretors,” has been based largely on hemagglu- tination-inhibition tests performed with boiled samples of saliva (Candela et al, 1940; Wiener et a!., 1942). Moreover, except in the serum of gelada monkeys, the specific presence or absence of anti-A or anti-B hemagglutinins has largely followed Landsteiner’s rule (Wiener and Moor- Jankowski, 1970) , and has thus provided further confirmation for typing results obtained with individual samples of saliva. In the case of human blood-group-substance “secretors,” Beckman (1964, 1970) has ob- served that, along with their high concentrations of soluble blood-group substances in secretions and other body fluids, secretor types also display elevated serum levels of “intestinal-type” alka- line phosphatase, especially following the inges- tion of fatty meals (Langman et al., 1966). It is thus interesting to speculate that other biochemi- cal characteristics, besides merely intestinal-type alkaline phosphatase levels, may prove to be closely associated with the tissue distribution and/or ultrastructural localization of blood- group-active antigens in different species of primates. Indeed, the concept has recently been advanced (Chuba, 1971) that, in one form or another, “blood-group-like” heterosaccharides have been functioning as post-translational “in- formation” molecules vitally involved in the se- lective transport and binding of substrates throughout organic evolution. It has already been clearly established that A-B-O-active antigens are ubiquitously dis- tributed throughout both the plant and animal 97 98 New York Zoological Society: Zoologica, Summer, 1972 kingdoms (Springer, 1970; Cushing et al., 1963; Chuba et al., 1971) and are present as alcohol- extractable or water-soluble substances in vari- ous tissues besides merely the red blood cells of human beings (Wiener, 1943). Surprisingly few investigations, however, appear to have been undertaken to elucidate the morphogenesis and precise localization of A-B-O-active antigens in various primate tissues and organs (Szulman, 1966). Expanded knowledge in this area would obviously be of basic research interest from the point of view of biochemical evolution, as well as of practical clinical importance from the point of view of organ transplantation (Dausset and Rapaport, 1966, 1968) and cross-circulation therapy (Hume et al., 1969). The purpose of the present study is to explore the feasibility of investigating the distribution of A-B-O-active antigens in baboon tissues accord- ing to: (1) their water-soluble versus their alcohol-soluble properties; and (2) according to procedures adapted from the methodology de- veloped by Basch and Stetson (1962, 1963) to quantitate the tissue distribution of mouse H-2 (histocompatibility) antigens. This work was supported in part by a grant from the Scaife Family Charitable Trusts to the Osborn Laboratories of Marine Sciences. Materials and Methods Freshly autopsied tissues from a group-A baboon (Papio anubis) were provided by the Laboratory for Experimental Medicine and Sur- gery in Primates (LEMSIP) of New York Uni- versity Medical Center. Human erythrocyte stro- mata (brown preparations) were prepared as previously described (Chuba et ah, 1970) or by lysing saline-washed erythrocytes (20% suspen- sions in 0.9% saline) for several hours at 59 °C. Randomly excised portions of raw baboon tissue (stomach, salivary gland, pancreas, and skeletal muscle) were minced into small frag- ments in 3 ml of saline per gram of wet tissue for a preliminary 18-hour extraction at 4°C. The once-extracted tissues were then acetone dried at room temperature in large watch glasses, pulverized with a pestle and mortar, and weighed on a fine balance. With the procedure employed, the dry weight of the acetone-dried tissues was consistently some 81% less than the wet weight of corresponding raw tissue, except in the case of pancreatic tissue, where the dry weight was some 90% less than the wet weight. Portions of pulverized tissue were then re- extracted in saline for 20 hours at 4°C at a concentration of 100 mg of pulverized tissue per ml of saline. Ethanol extracts were obtained by extracting portions of pulverized tissue for 72 hours at 22°C at a concentration of 50 mg of pulverized tissue per ml of 99% ethanol. The saline- and ethanol-extracted substances were then tested according to the procedures described in the footnotes of Tables 1 and 2. Antibody- Table 1. Hemagglutination-Inhibition Potency of Saline-Extracted Substances FROM Baboon Tissues. l:8-titer agglutinating reagent mixed with equal volume of saline extract from Indicator Saline- reagents extracted (human) tissue Red Anti- (baboon) cells serum Minced raw tissuei Unboiled and Boiled and diluted 1: diluted 1: 4 16 64 256 4 16 64 256 Acetone-dried tissue- diluted 1 : 4 16 64 256 Stomach A, anti-A O' 0 (O) -b -b o (O) (O) -b -b o o (+) -b -b -b B anti-B -b -b -b + + + -b + + + -b + -b + -b -b + -b -b -b -b -b -b + -b + -b -b -b-b -b -b -b Salivary Al anti-A o (O) -b-b -b + -b -b (+) + -b -b -b + + + o o (O) -b -b -b gland B anti-B -b-t- H- -b + -b + + -b + -b -b ND -b -b -b -b -b -b-b-b -b -b -b Pancreas Al anti-A -b + -b -b + -b + -b + + + + + + -b + -b -b -b -b o -b + -b-b-b -b -b -b B anti-B -b -b -b + + -b + + + ND + -b -b-b -b -b-b-b -b -b -b Skeletal Al anti-A -b -b -b -b -b + + -b -b -b -b -b ND + -b -b -b -b -b -f -b-b -b -b -b muscle B anti-B + -b -b + + -b + + -b -b -b + ND -b -b -b -b -b -b -b-b-b -b + -b ND = not done. 'Each gram (wet weight) of minced raw tissue was extracted with 3 ml of 0.9% saline for 18 hrs at 4°C. Inhibition tests were performed with the tissue-free supernate (5000 x G for 5 min.). 'Each 100 mg (dry weight) of acetone-dried tissue was extracted with 1 ml of 0.9% saline for 20 hrs at 4°C. Inhibition tests were performed with the tissue-free supernate (1000 x G for 5 min.). 'Macroscopic hemagglutination is graded from + to -b-|-H-; ( + ) = trace of macroscopic agglutination; (O) = trace of microscopic agglutination; O = no detectable agglutination up to 20X magnification. ' H = hemolysis; (H) = partial hemolysis; C = clot formation. Chuba, Kuhns, & Nigrelli: Blood-Group Activity in Baboon Tissues 99 absorption-capacity tests with the extracted tis- sues were performed quantitatively according to the procedures described in the footnotes of Tables 3 and 4. All of the serological tests were performed in Kahn-type tubes as previously described (Chuba et ah, 1968; Chuba et al., 1970). Hemaggluti- nation reactions were graded after 20 to 30 minute incubation at 22°C and a light spin (cf. footnote 3, Table 1 ). Results As shown in Table 1, saline extracts from either raw or acetone-dried stomach and sali- vary-gland tissues selectively inhibited the hemagglutination of human group-Aj erythro- cytes in reactions which indicated the presence in these tissues of consequential amounts of thermostable group-A substance. The inhibition tests with pancreatic extracts, however, were equivocated by the presence of nonspecific hemolytic activity, which (not shown in any of the tables) also caused the hemolysis of homolo- gous baboon erythrocytes, even when no anti- serum was mixed with the pancreatic extracts prior to the introduction of the indicator eryth- rocytes. Skeletal-muscle extracts, on the other hand, did not display any detectable activity. As shown in Table 2, none of the ethanol- extracted substances from the baboon tissues displayed consequential blood-group activity in the hemagglutination-inhibition tests. The non- specific hemolytic activity associated with ba- boon pancreatic tissue in Table 1, however, was demonstrable in saline suspensions of both the ethanol-extract precipitate and supernate residue derived from the baboon pancreatic tissue in Table 2. Table 3 shows the quite similar anti-A- absorption potency of human group-A^ erythro- cyte stromata and the baboon stomach and salivary-gland tissues. Interestingly, neither boil- ing-water-bath treatment (for 15 minutes) nor 72-hour ethanol extraction had a notable effect on the capacity of either the human group-Aj erythrocyte stromata or the baboon stomach and salivary-gland tissues to absorb human anti-A isoagglutinins. The nonspecific hemolytic activity associated with baboon pancreatic tissue in Tables 1 and 2 was readily demonstrable in the anti-A serum supernate after absorption with either boiled or unboiled pancreatic tissues (Table 3). Hemo- lytic activity was not demonstrable in the anti-A serum supernate, however, following absorption with pancreatic tissue from which hemolytic activity had previously been extracted with ethanol in Table 2. Table 4 shows the disproportionately greater anti-B-absorption potency of human group-B erythrocyte stromata compared with the baboon tissues studied. The weak B-like activity dis- Table 2. Hemagglutination-Inhibition Potency of Ethanol-Extracted Substances’ FROM Baboon Tissues. Ethanol- extracted tissues (baboon ) Indicator reagents ( human) Red Anti- cells serum Agglutinating reagent mixed with equal volume of ethanol-derived Precipitate suspension- diluted 1: 1 4 16 64 Supernate-residue suspension^ diluted 1 : 1 4 16 64 Stomach A, anti-A + ■1 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- + 4- 4- 4- 4- 4- 4- 4- + 4- 4- B anti-B -1-4- 4- 4- + 4- 4- 4- 4- 4- 4- 4- ( + ) 4- + 4- 4- 4- + 4- 4- Salivary A. anti-A ( + ) 4- 4- 4- 4- 4- 4- 4- 4- 4- + 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- gland B anti-B + + 4- 4- 4- 4- 4- 4- + 4- 4- 4- ( + ) 4- -f 4- -f 4- + 4- + 4- Pancreas A, anti-A H’ (H) 4- 4- 4- 4- 4- 4- H H 4- 4- 4- 4- 4- 4- B anti-B ( + ) 4-4- 4- 4- 4- 4- + 4- 4- H (O) 4- 4- 4- 4- 4- 4- Skeletal A, anti-A -H + 4- 4- 4- 4- 4- 4- 4- + 4- 4- 4- 4- 4- 4- + + 4- 4- 4- 4- 4- 4- muscle B anti-B 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- + 4- 4- 4- 4- 4- 4- 4- 4- 4- 4- ’ Substances present in clear (except for pancreas) supernate obtained (1000 x G for 5 min) from acetone- dried baboon tissue extracted (50 mg dry tissue per ml 99% ethanol) for 72 hrs at 22°C. “Precipitate was collected (1000 x G for 1 min) following the additional incubation of the above 72-hr supernate for 48 hrs at — 8°C. Tests were performed with the acetone-washed precipitate finely suspended in 0.5 ml of saline for each 100 mg of dry tissue originally xtracted with ethanol. “ The supernate residue was obtained by evaporating the 48-hr supernate (decanted from the packed — 8°C precipitate above) to dryness at 22°C. Tests were performed with the supernate finely suspended in 0.5 ml of saline for each 100 mg of dry tissue originally extracted with ethanol. ’ Cf. footnotes 3 and 4, Table 1. 100 New York Zoological Society: Zoologica, Summer, 1972 played by the baboon salivary-gland tissue in Table 4, and, to a much lesser extent, by the ethanol-derived supernate residue of baboon stomach and salivary-gland tissues in Table 2, was the only evidence suggesting possible group-B activity in the baboon tissues studied. As in the case of the absorption of anti-A serum in Table 3, the anti-B serum supernate in Table 4 acquired hemolytic activity during ab- sorption with either boiled or unboiled baboon pancreatic tissue, but not during absorption with pancreatic tissue which had been previously ex- tracted with ethanol. The ethanol-extractable, apparently thermostable hemolytic activity asso- ciated with baboon pancreatic tissue in these experiments thus presents a challenging area for further study. Discussion The findings with the freshly autopsied ba- boon tissues are largely consistent with the group-A status previously established for the baboon by LEMSIP investigators on the basis of: (1) the presence of group-A and absence of group-B activity in saliva samples, and (2) the presence of anti-B and absence of anti-A agglu- tinins in serum samples studied during the life of the baboon (Dr. W. Socha, personal com- munication). The presence of anti-B agglutinins in the se- rum of the baboon during life indicates that the weak B-like activity of baboon salivary gland tissue in this post-mortem study (Table 4) was not related to B-active receptors possessing the same fine structures as those responsible for group-B activity in human tissues. If the baboon did actually possess weak B-like antigens, not- withstanding the presence of anti-human-B ag- glutinins in its serum, the situation may be some- what analogous to the presence of anti-A^ ag- glutinins in the serum of certain individuals belonging to the “weak- A” subgroups (Wiener, 1943). In any event, at this point it should be fully appreciated that blood-typing reagents employed to define extrinsic serological attri- butes or “blood factors” do not, at the same Table 3. Comparative Anti-A-Absorption Potency of Variously Extracted Human Group-A, Erythrocyte Stromata and Baboon Tissues. Human anti-A serum tested with human group Ai erythrocytes after absorption^ with Saline-extracted Ethanol- Absorbing tissue Mg/ml Unboiled tissue Boiled tissue extracted tissue conc.i Absorbed serum Absorbed serum Absorbed serum diluted 1: 12 4 8 diluted 1 : 12 4 8 diluted 1 : 12 4 8 None 0 c H--f + (O) -1- + + -M- -1- (O) •+• + + -b -b -b (O) (control) Human 40 0 o 0 o o o o 0 o o o o grp Ai 20 -1- (O) o o + (O) o o (O) o o red cell 10 4- (+) (O) o + (+) (O) 0 -1- (+) (O) o stromata 5 -h-M- + (O) 0 -I--I- + -1- (0) o -f + -f -b (O) o Baboon 20 o 0 0 o o 0 o o 0 o o o stomach 10 -1- (O) o o 0 o o o o 0 o o 5 -(- (O) (O) 0 (O) (0) 0 o o o o 0 2.5 -t- ( + ) (0) o (+) (0) o o (+) (0) 0 o Baboon 40 o o o o 0 o 0 o (O) o o o salivary 20 (+) (+) o o -f (0) 0 0 -1- -1- (0) (0) o gland 10 •f-b -1- (0) (O) -t- -f -1- (O) (O) -b + (0) (O) Baboon 40 H^ H o o H H 0 o -b-b-b -b-b o o pancreas 20 H H o o H (+) (O) (0) -b-b-b -b-b-b (+) (O) 10 H + + (O) o H ( + ) (+) (0) -b -b -b +++(+) (0) Baboon 80 ( + ) ( + ) o o -t- -1- (+) (O) 0 (O) o o o skeletal 40 + + + (+) (O) -h-f -1- (+) (O) + (+) (O) o muscle 20 +++(+) (O) -1- -t + + 3- (+) (O) -b-b-b -b -b (+) (O) ’ Absorptions were performed by mixing small aliquots of the reagent anti-serum with the number of mg of acetone-dried tissue necessary to provide the mg/ml tissue concentrations specified in column 2 above. Hemagglu- tination tests were performed with the tissue-free supernate (1000 x G for 5 min.) following 30 min. absorption at 22°C. - Cf. footnotes 3 and 4, Table 1. Chuba, Kuhns, & NigrelU: Blood-Group Activity in Baboon Tissues 101 time, establish the presence of identical “hap- tenic” structures in a virtually unlimited number of different substances capable of displaying varying degrees of “blood-group” activity (Land- steiner, 1945; Wiener, 1966). That baboon and rhesus monkey tissues, for example, may indeed possess a spectrum of B-like receptors slightly different in fine struc- ture from the B-active receptors associated with human blood-group substances is suggested by recent catfish immunization experiments (Chuba et al., 1970). In these experiments, immunization of white catfish (Ictalurus cat us) with group-B- active baboon or rhesus monkey saliva evoked a complex spectrum of serum heteroagglutinins, not all of which could readily be classified as “anti-B” when tested with human group-B and B-like fur-seal and sea-lion erythrocytes. Brown bullhead catfish (/. nebulosus) immunized with human group-B saliva, on the other hand, pro- duced heteroantibodies which, following appro- priate absorption fractionation, could be sharply distinguished as anti-B agglutinins (Chuba et al., 1968; Wiener et al., 1968; Chuba et al., 1970). Not enough catfish of each species were con- currently available for each of the foregoing experiments, however, to establish clearly the extent to which species differences in catfish im- mune responsiveness, rather than species differ- ences in the B-active antigens used in the experi- ments, may have significantly influenced the heterogeneity of antibodies produced. Interest- ing future experiments suggested by the present study would be to inject a series of catfish in parallel with both soluble and particulate anti- gens from different species of primates. During a preliminary experiment along these lines, Chuba, Kuhns, and Nigrelli (in preparation) and A. S. Wiener (personal communication) found that either boiled saliva, saline-washed erythrocytes, or brown erythrocyte stromata from the same human goup-O secretor, when in- jected separately into a series of white catfish, evoked virtually the same spectrum of anti-H and anti-Z heteroagglutinins (Wiener et al., 1968; Baldo and Boettcher, 1970; Cushing, 1970) in all of the catfish. In view of the subgroup polymorphism of animal group- A and group-B antigens (Wiener, 1943), comparative catfish immunization experi- ments with group-A and group-B antigens de- rived from different tissues of different species of primates would undoubtedly produce an even more interesting array of catfish heteroag- glutinins. Such unique immunological reagents would obviously be of basic research interest, as Table 4. Comparative Anti-B-Absorption Potency of Variously Extracted Human Group-B Erythrocyte Stromata and Baboon Tissues. Human anti-B serum tested with human group B erythrocytes after absorption^ with Saline-extracted Ethanol Absorbing tissue Mg/ml tisciip Unboiled tissue Boiled tissue extracted tissue cone.’ Absorbed serum Absorbed serum Absorbed serum diluted 1 : 12 4 8 diluted 1 ; 12 4 8 diluted 1 : 1 2 4 None 0 -1- + +2 + + + ( + ) + -F + + + + ( + ) + + + + + -F (+) (control) Human 20 o O O o o O O o o o o o grp B 10 o O O o o 0 O o o o o o red-cell 5 o O o o o o o o (O) o o o stromata 2.5 (O) (O) 0 o (O) (0) o o (+) o o o Baboon 40 (0) (O) o o + + -F + + -F + + (O) + + + + -F + (O) stomach 20 -1- -1- + + -t- -H + (+) + + -F + -F + -F-F (+) + + -F + + + -F (+) 10 -(- + + -I- + + (+) + -F + -F-F -F (+) -F-F + + + -F (+) 5 + -H- + -H -t- + (+) + + + + -F-F + + + + -F + -F-F + (+) Baboon 40 0 o o o o o o o -F ( + ) (O) (O) salivary 20 -f -t- + (O) o + (+) 0 o -F -F + (O) (0) gland 10 + + -I- + (+) (O) -F-F-F + -F (+) (O) -F + -F + + (O) (O) 5 -f + 4- -F -F + (+) + -F-F -F + (+) (O) + + + -F-F + (+) Baboon 10 H2 H H H H H (H) o + -F-F -F + -F -F + -F pancreas ’ Cf. footnote 1, Table 3. - Cf. footnotes 3 and 4, Table 1. 102 New York Zoological Society: Zoologica, Summer, 1972 well as of potential clinical usefulness in tissue- matching procedures. The negligible effect of boiling-water-bath treatment or ethanol extraction on the capacity of human erythrocyte stromata or baboon tissues to absorb human blood-group antibodies in the present study (Tables 3 and 4) is also note- worthy. It has long been an accepted dualistic concept that the blood-group activity of human erythrocytes is primarily associated with mem- brane glycolipids and that the blood-group activ- ity of secretions is primarily associated with water-soluble glycoproteins having a carbohy- drate content of some 85% (Morgan, 1970; Watkins, 1970). This dualistic concept has been challenged recently, however, with cogent ana- lytical data suggesting that membrane glyco- proteins, rather than membrane glycolipids, may actually be primarily responsible for the blood- group activity of human erythrocytes (Whitte- more et al., 1969; Poulik and Lauf, 1969; Poulik and Bron, 1970; Zahler, 1968). Our observation that ethanol extraction did not have a notable effect on the residual blood-group activity of the human erythrocyte stromata or baboon tis- sues studied (Tables 3 and 4) further supports the concept that blood-group-active macromole- cules other than alcohol-extractable glycolipids may be primarily responsible for the A-B-O blood-group activity of particulate cellular ma- terials. Moreover, recent investigations which have tended to perpetuate the dualistic concept of cellular-glycolipid versus soluble-glycoprotein blood-group antigens (e.g., Koscielak, 1963) appear to be vulnerable to the criticism that: ( 1 ) as already pointed out by Whittemore et al. ( 1969), only miniscule amounts of blood-groiip- active glycolipids — quantitatively insufficient to contribute significantly to the blood-group activ- ity of intact erythrocytes — have been extracted from erythrocyte membranes with lipid solvents; and (2) the glycolipid-oriented investigators have invariably failed to assay their “extracted” erythrocyte preparations for residual blood- group activity, such as was done by means of the quantitative antibody-absorption-capacity tests in the present study (Tables 3 and 4). The possibility thus exists that the cell-mem- brane-associated glycoproteins include a class of blood-group-active macromolecules possessing physicochemical properties quite different from those of the “water-soluble” blood-group sub- stances. This possibility is supported by the report (Rega et al., 1967) that erythrocyte- membrane glycoproteins have a carbohydrate content of only some 9%. If this proves to be generally true, then membrane-associated glyco- proteins, including those with blood-group activity, would presumably be much more read- ily coagulated and entrapped with other cellular material during various preparatory procedures (e.g., “extraction” with protein-denaturing re- agents, etc.) than the “soluble” blood-group substances possessing a carbohydrate content of some 85%. In fact, the demonstration of blood-group activity in baboon stomach and salivary-gland tissues in this investigation could possibly be largely attributable to the in vitro coagulation-entrapment of different classes of blood-group-active glycoproteins, rather than to a preponderance in these tissues of blood-group antigens intrinsically bonded in vivo with the cellular structures themselves. There is also the possibility that substantial fractions of blood- group-active membrane fragments and/or sub- cellular organelles were not removed with the more particulate tissue debris during routine centrifugation procedures (cf. footnotes. Tables 1 and 2), and thus may have contributed sig- nificantly to the “soluble” blood-group activity of the “tissue-free” preparations. In a classic series of studies. Stetson and his associates quantitated the tissue distribution of mouse H-2 antigens (Basch and Stetson, 1962, 1963), as well as their ultrastructural localiza- tion in different membrane fractions and sub- cellular organelles (Herberman and Stetson, 1965). Similar definitive studies on the tissue distribution and ultrastructural localization of primate A-B-O-active antigens will obviously be necessary before many of the fundamental questions raised by the present study can be fully elucidated. Summary 1. Stomach, salivary-gland, pancreas, and skel- etal-muscle tissues from a freshly autopsied group- A baboon (Papio anubis) were ex- tracted with 0.9% saline and 99% ethanol. 2. Only saline extracts from either raw or acetone-dried stomach and salivary-gland tissues displayed significant blood-group (group A) activity in hemagglutination- inhibition tests. 3. Both saline- and ethanol-extracted human group-Ai erythrocyte stromata, baboon stomach, and baboon salivary-gland tissues displayed quite similar anti-A-absorption potency in quantitative antibody-absorption- capacity tests. 4. In the case of pancreatic tissue, ethanol- extractable, apparently thermostable hemo- lytic activity interfered with the hemagglu- tination-inhibition and antibody-absorption- capacity tests. Chiiba, Kuhns, Nigrelli: Blood-Group Activity in Baboon Tissues 103 Literature Cited Baldo, B. a., and B. Boettcher 1970. Natural erythrocyte agglutinins in the se- rum of the Australian freshwater catfish, Tandanus tandanus (Mitchell) I. Exami- nation of the specificities of the agglutinins with emphasis on the ABH agglutinins. Immunology, 19: 569-581. Basch, R. S., and C. a. Stetson 1962. The relationship between hemagglutino- gens and histocompatibility antigens in the mouse. Ann. N. Y. Acad. Sci., 97, Art. 1: 83-94. 1963. Quantitative studies on histocompatibility antigens of the mouse. Transplantation, 1 : 469-480. Beckman, L. 1964. Associations between human serum alka- line phosphatase and blood groups. Acta Genet, 14: 286-296. 1970. Blood groups and serum phosphatase. In: Blood and Tissue Antigens, ed. D. Ami- noff. Academic Press, New York: 67-81. Candela, P. B., A. S. Wiener, and L. J. Goss 1940. New obervations on the blood group fac- tors in Simiidi and Cercopithecidae. Zoo., 25: 513-521. Chuba, J. V. 1971. Physiological “ecology” of blood groups. Exp. Med. Surg., 29: 4-10. Chuba, J. V., W. J. Kuhns, and R. F. Nigrelli 1968. The use of catfish, Ictalurus nebulosus (Le Sueur), as experimental animals for immunization with human secretor saliva and other antigenic materials. I. Immun., 101: 1-5. 1971. Forssman-like activity of different teleost and anuran erythrocytes. Immunology, 20: 611-615. Chuba, J. V., W. J. Kuhns, R. F. Nigrelli, R. A. Morris, and U. E. Friese 1970. B-like blood factor of fur seals and sea lions. Haematologia (Budapest), 4: 85-96. Cushing, J. E. 1970. Immunology of fish. In: Fish Physiology, Volume 4, ed. W. S. Hoar, and D. J. Ran- dall. Rcademic Press, New York: 488. Cushing, I. E., N. L. Calaprice, and G. Trump 1963. Blood group reactive substances in some marine invertebrates. Biol. Bull., 125: 69-80. DAUSSET, I., AND F. T. Rapaport 1966. The role of blood group antigens in human histocompatibility. Ann. N. Y. Acad. Sci., 129, Art. 1: 408-420. 1968. Blood group determinants of human his- tocompatibility. In: Human Transplanta- tion, ed. F. T. Rapaport, and J. Dausset. Grune and Staratton, New York: 383-393. Herberman, R., and C. A. Stetson, Jr. 1965. The expresion of histocompatibility anti- gens on cellular and subcellular mem- branes. J. Exp. Med., 121: 533-549. Hume, D. M., W. E. Gale, Jr., and G. M. Williams 1969. Cross circulation of patients in hepatic coma with baboon partners having human blood. Surg. Gynec. Obstet., 128: 495-517 Koscielak, j. 1963. Bloodgroup A specific glycolipids from human erythrocytes. Biochim. Biophys. Acta, 78: 313-328. Landsteiner, K. 1945. The Specificity of Serological Reactions, 2nd English edition. Republished 1962 by Dover Publications, New York: 99. Langman, M. j., E. Leuthold, E. B. Robson, J. Harris, J. E. Luffman, and H. Harris 1966. Influence of diet on the intestinal compo- nent of serum alkaline phosphatase in people of different ABO and secretor sta- tus. Nature, 212: 41-43. Morgan, W. T. J. 1970. Molecular aspects of human blood-group specificity. Ann. N. Y. Acad. Sci., 169, Art. 1: 118-130. POULIK, M. D., AND P. K. LaUF 1969. Some physico-chemical and serological properties of isolated protein components of red cell membranes. Clin. Exp. Immun., 4: 165-175. POULlK, M. D., AND C. BRON 1970. Antigenicity of red cell membrane pro- teins. In: Blood and Tissue Antigens, ed. D. Aminoff. Academic Press, New York: 343-345. Rega, a. F., R. I. Weed, C. F. Reed, G. G. Berg, AND A. RoTHSTEIN 1967. Change in the properties of human eryth- rocyte membrane protein after solubiliza- tion by butanol extraction. Biochim. Bio- phys. Acta, 147: 297-312. Springer, G. F. 1970. Importance of blood-group substances in interactions between man and microbes. Ann. N. Y. Acad. Sci., 169, Art. 1: 134- 152. Szulman, a. E. 1966. Chemistry, distribution, and function of blood group substances. Ann. Rev. Med., 17: 307-322. Watkins, W. M. 1970. Blood group substances. In: Advances in Blood Grouping, Volume III, ed A. S. 104 New York Zoological Society: Zoologica, Summer, 1972 Wiener. Grune and Stratton, New York: 550-561. Whittemore, N. B., N. C. Trabold, C. F. Reed, AND R. I. Weed 1969. Solubilized glycoprotein from human erythrocyte membranes possessing blood group A, B, and H activity. Vox Sang., 17 : 289-299. Wiener, A. S. 1943. Blood Groups and Transfusion, 3rd edi- tion. Reprinted 1962 by Hafner Publish- ing Company, New York. 1966. The law of specificity with special refer- ence to the blood groups. Exp. Med. Surg. 24: 134-147. Wiener, A. S., and J. Moor-Jankowski 1969. The A-B-O blood groups of baboons. Am. J. Phys. Anthrop., 30: 117-122. 1970. The use of apes and monkeys for the in- vestigation of the human blood groups. Ann. N. Y. Acad. Sci., 169, Art. 1: 225- 233. Wiener, A. S., P. B. Candela, and L. J. Goss 1942. Blood group factors in the blood, organs, and secretions of primates. J. Immun., 45: 229-235. Wiener, A. S., J. V. Chuba, E. B. Gordon, and W. J. Kuhns 1968. Hemagglutinins in the plasma of catfish {Ictalurus nebulosus) injected with saliva from human secretors of various A-B-O groups. Transfusion 8 : 226-234. Zahler, P. 1968. Blood group antigens in relation to chemi- cal and structural properties of the red cell membrane. Vox. Sang., 15 : 81-101. 6 Captive Breeding of Orangutans John Perry and Dana Lee Horsemen”’ National Zoological Park, Smithsonian Institution, Washington, D.C., 20009 With more than 500 orangutans in captivity, the world’s zoos have a large available breeding stock. The number of births, and the apparent birth rate, have been increasing. However, no living orangutan is the offspring of captive-born parents. The future of the species in captivity depends on successful propagation of the captive born. The orangutan is a symbolic species for wildlife conservationists and zoo directors. Although protected by law wherever it occurs, it was heavily poached for a time. It was one of the few species whose survival might be threatened by collecting for zoos. Tom and Barbara Harrisson, then in Sarawak, called world attention to the illicit traffic and organized the Orangutan Recovery Service which rescued a number of the animals. Mrs. Harrisson under- took the first experiments to determine whether and how captive orangutans might be reintro- duced in the wild. Before 1966, almost every orangutan acquired by zoos was contraband. They were so freely available that negotiating for legal capture and exportation seemed pointless. Then, at the urg- ing of the International Union for Conserva- tion of Nature and Natural Resources, the American Association of Zoological Parks and Aquariums and the International Union of Di- rectors of Zoological Gardens adopted a ban on undocumented orangutans. Other zoo federa- tions took similar action. Import controls were imposed by several nations, while others tried to regulate transit. A few orangutans were seized by authorities. Dealers found fewer cus- tomers for smuggled animals, and a number were blocked in the pipeline. Before the smuggling trade was curtailed, zoo collections were well-stocked. The first official studbook reported 460 as of December 31, 1969 (Bourne, 1.971a). The International Zoo Year- book Census counted 469 a few months later. Two years later, in 1971, the IZY total was 539, but a large part of this increase was caused by reports from additional collections. Mrs. Horsemen left the Zoo prior to pub- lication. According to Crandall (1964) , the first orang- utan brought to Europe arrived in 1640; Amer- ica saw its first in 1825. Many of the first im- ports had short lives: Frederick A. Ulmer, Jr. (1966), writes that a Dutchman, Mynheer van Goens, brought 102 Sumatran orangutans to Europe in 1927 and 1928, of which 33 came to the United States in a single shipment. Most soon died. At the Philadelphia Zoo, whose ex- perience with the species extends over almost a century, the average life of 13 animals exhibited between 1879 and 1930 was three-and-one-half years (Ulmer, 1957). Many of the early deaths were caused by mishandling from capture to delivery. Young orangutans traveling the smuggling route were often in poor condition on arrival. Zoo men gradually learned how to rehabilitate them, and zoo husbandry improved. While Crandall (1964) judged average longevity to be “still rather low” as recently as 1964, there is no reason now to believe captive life-spans are markedly lower than those of free-living orang- utans. Philadelphia’s “Guas” (Studbook #467) and “Guarina” (Studbook #468) were each about 50 years of age at the end of 1969. Until longevity improved, little propagation could be expected; but even when adult pairs were kept together, no success was achieved until 1928. In that year, there were captive births at Berlin, Nuremberg, and Philadelphia. While these babies died in infancy, the barrier had been broken. Philadelphia’s second orang- utan was born in 1930 and successfully reared (Ulmer, 1966). Until the 1960s, births occurred now and then, with a fair survival rate, but each success was notable. Then came the rapid increase in zoo collections, and an upward trend began. 105 106 New York Zoological Society: Zoologica, Summer, 1972 The IZY reported six successful births in 1964. There were 19 in 1967, 30 in 1971. The IZY census and the studbook report differ with re- spect to the 1969 total of living individuals that were captive-born: 81 in the census, 101 in the studbook. In 1971, according to IZY, there were 152 captive-born orangutans. Only 13 percent of those reported in the 1964 census were cap- tive born. There were 28 percent in 1971. The studbook provided the first data with which to measure breeding potential and results. The first edition gave information on all births occurring in 1967, 1968, and 1969, and on all orangutans reported in collections as of De- cember 31, 1969 (Bourne, 1969). Captive female orangutans are most likely to bear offspring between the ages of seven and eighteen, although Philadelphia’s “Guarina” de- livered nine babies between 1929 and 1955, a reproductive period of 26 years (Ulmer, 1966). During the three years of the studbook reports, 122 females were in the 7 to 18 age group, and 45 of them produced at least one live infant, 37 percent of those eligible. The 45 had a total of 65 successful births, including one set of twins. There is no reliable way to compare prop- agation in the wild. Barbara Harrisson (1971, private correspondence) notes that present-day orangutan habitats differ greatly in quality, with consequent effects on life spans, propagation, and infant mortality. She believes wild females seldom bear young before nine years of age. Infant mortality tends to be rather high in the wild, ranging around 40 percent. A stable pop- ulation in a protected habitat would indicate that the average female produces four to five babies, usually between the ages of nine and 30-plus (Harrisson, 1971, private correspond- ence ) . If “Guarina” were typical of captive females, the comparison would be favorable. Of her nine offspring, four were living at the end of 1969. But only two other captive females had this many surviving offspring, and only five had three surviving offspring (Bourne, 1971a). No data is available on the number of deceased off- spring. A rising percentage of captive-born orangu- tans would seem reassuring. It may not be. If, for example, accessions of wild-caught stock ceased and second-generation births fell short of the replacement rate, the percentage of cap- tive-born individuals would approach 100 per- cent as the total population approached zero. The wild-caught individuals, being older, would usually die before their progeny. The number of births has risen from year to year. It is disquieting, however, that no living orangutan is the offspring of two zoo-born par- ents''*. Before the studbook appeared, we heard rumors of second-generation and even third- generation births. In part, this may have been misunderstanding. Two correspondents used “second generation” to mean the first zoo-born generation. Since Philadelphia was the site most named in the rumors, we wrote to Fred Ulmer. His reply: In every Philadelphia orangutan birth, at least one parent was wild-caught. He added as a footnote: “We did have some breeding be- tween ‘Pinky’ and ‘Ivy’, who are brother and sister, but the matings always resulted in abor- tions” (1971, private correspondence). Has there yet been time for second-genera- tion births to occur? The number of captive- born individuals has increased slowly since 1930. A few captive-born females have given birth, and a few captive-born males have sired young. The studbook does not disclose how many captive-born pairs of breeding age have been kept together. Of the 122 females of breeding age during the 1967-1969 period, 109 were wild-caught, only 13 zoo-born. Of the 64 females then in the 9 to 1 1 age classes (the most productive years, according to studbook records), only five were zoo-born. Obviously the potential for sec- ond-generation propagation is still low. Further, many zoo men prefer to pair captive-born in- dividuals with wild-caught mates, or they do so simply because captive-born mates are not avail- able. At Philadelphia, for example, “Ivy,” born to “Guas” and “Guarina” in 1937, was mated with her father and produced a female baby in 1950. Still, there is reason for concern, and this will not be eased until a satisfying number of second and third generation births have occurred. Are there problems and challenges to be overcome by management skill? There is evidence that behaviors such as mat- ing and maternal care are, to some degree, learned by the great apes in their natural settings. While this is more pronounced among gorillas than orangutans, observations in a number of zoos indicate the latter is influenced. A number of zoo-raised male orangutans have, on reach- ing sexual maturity, responded to the sexual urge in ways more bizarre than appropriate. A number of zoo-raised females have, on giving birth, rejected their infants. Dr. Lang (1971) comments: “Orangs tend to be less affected than gorillas when deprived of social experience with their own kind in infancy, but nevertheless the A second-generation birth occurred at Rot- terdam some years ago, according to Marvin Jones. It died prior to 1967. Perry & Horseman: Captive Breeding of Orangutans 107 number of failures in maternal behavior must give rise to some concern.” When the Wild Animal Propagation Trust was organized in the United States to promote captive breeding of endangered species, the Orangutan Committee was the first named. One of the needs was to bring together unpaired males and females. Thanks to the cooperation of a number of zoos and dealers, this has been reasonably successful. At the time of the 1971 IZY census, only eight U.S.A. collections had one sex only, and seven of these had males, for which no females had been found. It has also been the experience of some zoos that males and females raised together failed to mate on becoming sexually mature. The Na- tional Zoo had such a pair. In cooperation with WAPT, our male was sent to the Cheyenne Mountain Zoo in Colorado, where he sired young; we obtained another male which promptly mated successfully with our female. WAPT has helped to arrange other transfers and deposits. Such arrangements are not easily made, how- ever. Orangutans are much prized as exhibits, and directors of city-owned zoos are often not free to dispose of valuable animals. Even when a zoo director can subordinate local interests, finding a proper match is often difficult. The IZY census is not current when published, and it does not report the ages of individuals. The studbook provides more data but it has not been issued since 1969. No service provides informa- tion on the births, deaths, accessions, and re- movals that occur from month to month. Dr. Bourne of the Yerkes Regional Primate Research Center writes ( 1971b) : “I believe the key to sustained breeding is a large group of animals, so that switching of males and females can occur to be sure you can get compatible breeding pairs.” If this is the key to sustained breeding, the present outlook is not favorable. At the time of the 1971 IZY census, the world's largest orang- utan collection was at Yerkes, with 28 indi- viduals, plus seven on loan to the nearby At- lanta Zoo. Second largest outside Indonesia, 14 individuals, was at the Cheyenne Mountain Zoo. But the average number in 128 zoos reporting the species was only four. In such small groups, there are sure to be less than the optimum num- ber of compatible adult pairs. A high proportion of young orangutans are raised in isolation after they are taken from their mothers. Many are taken shortly after birth. Some zoo men are concerned lest the hand-rais- ing of zoo-born orangutans further complicate the breeding problem. Many infants are handled like human babies, associating only with humans in the early months of life. When they are re- turned to the zoo, it is usually to a lonesome cage, since there are long odds against the zoo having a compatible cage-mate. The managers of some collections advocate hand-raising and, indeed, remove all infants as a matter of course, regardless of whether the mother seems cap- able of providing care. The Yerkes laboratory has the advantage Dr. Bourne cites in its number of possible mating combinations. Like most other collections, how- ever, it provides caging for individuals or pairs. In these circumstances, there is little opportunity for the kind of behavioral learning mentioned by Dr. Lang. Only in recent years have ideas or plans been put forward to experiment with groupings of orangutans in large enclosures. One of the authors collaborated with Barbara Harrisson in designing an experimental enclo- sure. This called for several open-frame towers, 30 to 40 feet tall, with several platforms, each of which would shade the zone below. Orangu- tans would be fed in the towers. The hope was that the towers would serve as vertical territories. This design was followed in new construction at the Surabaja and Djakarta zoos (Indonesia). The young orangutans are climbing and brachia- ting as expected, but the real test will not come until they are sexually mature. In the United States, some decisions about caging must soon be made, for there is an acute space shortage. Many juvenile orangutans are kept in quarters inadequate for adults, and new construction is lagging behind the needs of this maturing population. The National Zoo was compelled to dispose of a young pair, through WAPT, for lack of space. In this case, WAPT was able to place them advantageously, but such placements are becoming more difficult. In summary, it is far from certain that the captive orangutan population will become self- sustaining. The number of births and the birth rate'-* will probably increase for several more years, simply because more wild-caught females will be entering the period of maximum fecun- dity. In these years, births may exceed deaths, so that the captive population increases. There may be a surplus of orangutans in zoos, as avail- able quarters become crowded. '-* Birth rates are unreliable indicators in a population as small and changing as this one. Based on IZY data, captive orangutan births per 1000 population have been: 1964 22 1968 48 1965 26 1969 62 1966 54 1970 64 1967 43 Cumulative average: 48 108 New York Zoological Society : Zoologica, Summer, 1972 This increase can be misleading, however. The future of the captive population depends on births in the second and succeeding captive gen- erations. It is too soon to predict that these will not occur in sufficient numbers, but also too soon to say that they will. Predictions are likely to be unreliable unless better data is developed for analysts to ponder. It is disturbing that the studbook data is not suf- ficient. One cannot evaluate second-generation birth results without knowing how many first- generation pairs have been given the opportunity of mating. It would also be valuable to known which captive-born individuals were hand- raised. Dr. Bourne finds reason for cautious optimism in success with chimpanzees (1971b, private correspondence) : One of our chimpanzees has five great-grand- children living at the Center, and another ani- mal has two great-great grandchildren. The last two are the product of two successive captive born matings for sure . . . This is all very good, and although we cannot neces- sarily extrapolate to orangs, it is encouraging. Literature Cited Bourne, Geoferey H. 1971a. International orangutan studbook: as of December 31, 1969. Yerkes Regional Pri- mate Research Center, Emory University, Atlanta, Georgia. 1971b. Private correspondence. Crandall, Lee S. 1964. Management of wild animals in captivity. University of Chicago Press, Chicago, Il- linois. Harrisson, Barbara 1971. Private correspondence. International Zoo Yearbook 1965-1972. Volumes 5-12. Zoological Society of London. Lang, Ernst M. 1971. Experience with breeding apes in Basle Zoo. Paper presented at the International Symposium on Breeding Non-Human Pri- mates, Bern, Switzerland. Ulmer, Frederick A., Jr. 1957. Breeding of orangutans. Der Zoologische Garten, pp. 57-65. 1966. First orangutan born in rare mammal house. America’s first zoo (Philadelphia Zoological Society), September, pp. 17-20. 1971. Private correspondence. ■i! NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable), and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 3. The Hematological Parameters and Blood Cell Morphology of the Brown Bullhead Catfish, Ictalurus nebulosus (Le Sueur). By Sheila R. Weinberg, Charles D. Siegel, Ross F. Nigrelli, and Albert S. Gordon. Tables 1-3 71 4. Histochemical Analyses of the Fluid and the Solid State of the Adhesive Materials Produced by the Pre- and Postmetamorphosed Cyprids of Balanus eburneus Gould. By Paul J. Cheung and Ross F. Nigrelli. Figures 1-6; Tables 1-10 79 5. Blood-Group Activity in Baboon Tissues. By Joseph V. Chuba, William J. Kuhns, and Ross F. Nigrelli. Tables 1-4 97 6. Captive Breeding of Orangutans. By John Perry and Dana Lee Horsemen 105 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx, New York 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, New York. Published January 8, 1973 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 57 • ISSUE 3 • FALL, 1972 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Simon Dresner, Editor & Curator, Publications and Public Relations Joan Van Haasteren, Assistant Curator, Publications & Public Relations F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman; William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Chairman, Dept, of Ornithology; Joseph Bell, Curator, Ornithology; Donald F. Bruning, Associate Curator, Ornithoolgy; Hugh B. House, Curator, Mammalogy; James G. Doherty, Associate Curator, Mammalogy; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM : James A. Oliver, Director; H. Douglas Kemper, Associate Curator; Chritopher W. Coates, Director Emeritus; Louis Mowbray, Researc/i Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Senior Scientist; George D. Ruggieri, S.J., Director & Experimental Embryologist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung, Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology; Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate; George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists; Robert M. Beck, Research Fellow ^ ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants; John Budinger, Pathology; Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology; Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology; Harold S. Goldman, Radiology,' Roy Bellhorn, Paul Henkind, Alan Friedman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine tThtodoxe Kazimiroff, De/it/s/o’; Alan Belson, Resident in Pathology ® 1973 New York Zoological Society. All rights reserved. 7 Captive Propagation : A Progress Report John Perry National Zoological Park, Washington, D.C. 20009 Donald D. Bridgwater Minnesota State Zoological Gardens, St. Paul, Minnesota 55155 Dana L. Horsemen^ National Zoological Park, Washington, D.C. 20009 Of 291 mammals species and subspecies listed by lUCN as rare or endangered, 162 have been reported in zoo collections since 1962, and 73 have reproduced at least once. Only a few of these, however, have captive in terms of numbers and reproduction rates. The first zoo to propagate a species in captivity earns a mark of distinction. In recent years, there have been fewer such events than in the past. Many species once thought impossible to breed in captivity have been bred. Others that reproduced rarely now do so more often. On balance, zoos are still consumers rather than producers of wildlife. A few zoo directors have protested this statement, but available vital statistics confirm it. A typical report from a lead- ing zoo shows: Births & Hatchings Deaths Net Mammals 156 178 - 22 Birds 252 433 -181 Reptiles, Amphibians, etc. 0 318 -318 Further, births and hatchings are not evenly distributed over the species in a collection. Among the birds, for example, a few species usually account for most hatchings. A few zoos are net producers. By and large, these have specialized in their collections. Game parks, game farms, and establishments devoted to breeding waterfowl or upland game birds us- ually produce annual surpluses. That zoos might become survival centers for endangered species is not a new idea. Proposing that a national zoo be established, in 1889, Smithsonian Secretary Samuel P. Langley de- ' Mrs. Horsemen left the National Zoological Park prior to publication. populations which seem reasonably secure dared it would be “a home and a city of refuge for the vanishing races of the continent.” As more and more species approach extinction, in- terest in survival centers has increased. Citing the Przewalski horse and wisent as examples, some zoo directors assert that captive breeding will be the last hope for many species. It seems timely to consider what has been accomplished thus far. Because the preceding table is typical, we have limited this review to mammals. The lUCN Red Data Book lists 291 mammal species and subspecies as rare or endan- gered. In 1962, the International Zoo Yearbook undertook the first of its annual censuses of rare species in zoos. Since that time, 162 of the 291 species and subspecies have been represented in collections. IZY reports of births indicate that 73 of these produced offspring at least once in the ten-year period. To simplify analysis, we chose two base years, 1962 and 1965, and from the 73 species and subspecies selected those with captive popula- tions of ten or more in either year. This is a crude method of choice; a herd of eight could be a good breeding base, while two dozen widely scattered would not be. However, on reviewing the species and subspecies thus eliminated, we saw no serious omissions for purposes of this study. There were 41 mammal species and subspecies with base-year populations of ten or more. The 1971 IZY Census showed population increases for 36 of them. This is not, in itself, evidence of breeding success, since the IZY Census does not report acquisitions from the wild. Further, the number of zoos reporting to IZY increased. 109 no New York Zoological Society: Zoologica, Fall, 1972 IZY does report the numbers of captive-born individuals within each year’s totals. When this data is assembled, there are strong indications of whether a captive population is self-sustaining. (In the following tables, a blank for 1962 may mean zero response. However, some spe- cies and subspecies have since been added to the Census list.) Zoo Populations of Rare and Endangered Species: 1962-1971 Captive- 1962 1965 1971 born No. Percent Marsupialia Yellow-footed Rock Wallaby (Petrogale xanthopus) 4 52 46 42 91 Long-nosed Rat Kangaroo (Potorous tridactylus) 5 13 23 8 35 White-throated Wallaby (Macropus parma) — 19* 180 70 39 *Not reported by IZY in 1965. Data for 1966. Primates Black Lemur (Lemur macaco) 32 25 73 28* 38 Red-fronted Lemur (Lemur fulvus rufus) 3 10 43 15* 35 *Number of captive-born not reported by Tananarive. Mongoose Lemur (Lemur mongoz mongoz) 22 59 167 64 38 Red Uakari (Cacajao rubicundus) 8 32 38 4 11 Goeldi Monkey (Callimico goeldii) 10** 16 6 38 **Not reported by IZY in 1965. Data for 1967. Golden Lion Marmoset (Leontopithecus rosalia) 76 39 51 ***Not reported by IZY in 196f i. Data for 1966. Orangutan (Pongo pygmaeus) 205 349 539 152 28 Bonobo Chimpanzee (Pan paniscus) 9 22 21 4 19 Carnivora Maned Wolf (Chrysocyon brachyurus) 7 11 65 + 22 34 Spectacled Bear (Tremarctos ornatus) 13 43 85 + 16 19 Brazilian Otter (Pteronura brasiliensis) 10 10 15 3 20 min min Brown Hyena (Hyaena brunnea) 5 32 49 17 35 Asiatic Lion (Panthera leo persica) Siberian Tiger 3 37 66 + 22 + 33 (Panthera tigris altaica) (Includes Korean form) — 120 296 + 153 + 52 Sumatran Tiger (P. t. Sumatrae) 23 78 + 59 + 76 North China Leopard (Panthera pardus japonensis) _ 29 51 + 44 + 86 Snow Leopard (Panthera uncia) 22 54 98 31 32 min Perry, Bridgwater (& Horseman: Captive Propagation 111 1962 1965 1971 Captive- born No. Percent Perissodactyla Przewalski Horse (Eqiius przewalskii) 85 121 182 181 99 min min Onager (Equus heminonus onager)* 62 113 139 + 74 53 *Including animals reported as E. h. heminonus. This combination was initiated by IZY in 1966. Indian Wild Ass (E. h. khur) 3 11 11 1 9 Nubian Wild Ass (Equus asinus africanus) 7 16 17 17 100 Hartmann Mountain Zebra (Equus zebra hartmannae) 54 72 91 + 37 + 41 Baird Tapir (Tapirus bairdii) 6 11 12 + 2 17 Great Indian Rhinoceros (Rhinoceros unicornis) 26 39 45 + 16 36 Black Rhinoceros (Diceros bicornis) 119 124 128 + 28 22 Artiodactyla Pygmy Hippopotamus (Choeropsis liberiensis) 49 85 128 + 58 + 45 Vicuna (Vicugna vicugna) 72 69 57 83 Burma Brow-antlered Deer (Cervus eldi thamin) 13 11 37 5 14 Thailand Brow-antlered Deer (C. e. siamensis) _ 12* 10 9 90 * Paris Zoo herd identified as C. e. eldi in IZY 1965. Tule Elk fC. canadensis nannodes) _ 14 32 17 53 Formosan Sika (C. nippon taiouanus) _ 306 374 + 336 + 90 Pere David Deer (Elaphurus davidianiis) Anoa 130 436 550 550 100 (Anoa depressicornis) - 23 24 7 29 Wisent (Bison bonasus) 132 234 303 + 232 + 77 min Arabian Oryx (Oryx leiicoryx) 5 27 75 + 49 + 65 Scimitar-horned Oryx (Oryx tao) 18 23 141 101 72 Addax (Addax nasomaculatus) 20 63 142 116 82 Arabian Gazelle (Gazella gazella arabica) 10 44 + 19 + 43 min Ovis orientalis omitted because of apparent changes in subspecies identification. These 41 cases include a wide range of situa- tions. The Przewalski horse story is familiar. At the other extreme, the Brazilian otter is obviously insecure. In between are a number of species which show promising trends but, as yet, provide more reason for hope than confidence. Our purpose was to identify those situations where zoo propagation has been sufficient to give reasonable assurance that a species can be permanently maintained without further acquisi- tions from the wild. As a beginning, we chose two arbitrary factors: a 1971 captive population 112 New York Zoological Society: Zoologica, Fall, 1972 of 100 or more, and at least half of these captive- born. While these factors alone could not guar- antee long-term security, it is unlikely that any- thing less would. Using these two factors as a screen, only eight species or subspecies qualified: the Siberian tiger, Przewalski horse, onager, Formosan sika. Pere David deer, wisent, scimitar-horned oryx, and addax. The mongoose lemur (Lemur mon- goz mongoz) is a possible candidate for this list; of the two principal collections, one did not re- port, while the second did not report numbers of captive-born. 1. Siberian Tiger 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 36 41 49 50 51 66 71 77 Total population 104 116 149 162 191 224 248 296 Captive bred 73 66 87 109 140 161 192 253 Percent captive bred 70 57 58 67 73 72 77 85 Births (surviving) 21 28 28 43 58 59 75 Individuals per collection 3 3 3 3 4 3 3 4 *1ZY reports births for the year preceding the Census. The number of individuals per collection re- reported. The total population increased by 192, mained almost static during the years the popu- the population of captive-born individuals by lation increased by 185 percent. The number of 180. The number of wild-caught individuals in- births increased slightly more rapidly than the creased from 31 to a peak of 63 in 1969, and total population. has since declined to 43. The apparent birth rate In the years shown, 312 successful births were has increased. 2. Przewalski Horse 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 24 29 33 35 40 41 43 42 Total population 118 121 149 147 157 160 161 182 Captive bred 116 120 148 146 156 159 160 181 Percent captive bred 98 99 99 99 99 99 99 99 Births (surviving) 18 12 18 19 19 14 27 — Individuals per collection 5 4 5 4 4 4 4 4 This species is often mentioned as a prime number of zoos having the species has also in- example of survival in zoos. The total popula- creased. The average number of individuals per tion has shown a slow but steady increase, . The collection remained constant. 3. Onager* 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 26 33 38 39 32 37 43 44 Total population 89 113 135 150 118 132 145 139 + Captive bred 34 34 61 76 65 56 77 74 Percent captive bred 38 30 45 51 55 42 53 53 Births (surviving) 6 18 15 14 13 14 14 + - min Individuals per collection 3 3 4 4 4 4 3 3 *Includes animals once reported as E. h. hemionus. An apparent population decline occurred in 1968. While there were reporting inconsisten- cies, losses were also indicated, and the popula- tion total has yet to regain its 1967 peak, nor has the total of captive-born individuals. The average number per collection has re- Perry, Bridgwater & Horseman: Captive Propagation 113 mained almost static, as has the apparent birth The onager position is not yet secure, though rate. Of the 44 collections reporting in 1971, there is no immediate reason for alarm. 1 1 had only one sex. 4. Formosan Sika 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 21 33 32 29 26 26 24 30 Total population 135 306 260 327 420 414 539 374 Captive bred 113 189 209 234 233 248 361 336 Percent captive bred 84 62 80 72 55 60 67 90 Births (surviving) 49 46 65 70 38 52 68 min Individuals per collection 6 9 8 11 16 16 22 12 There appear to be problems of subspecies ported an estimated 150 in 1970. identification here. Mountain Home (Texas), a Total population in all other collections in- private game ranch, reported 105 Cervits nippon creased by 90 from 1970 to 1971. The average taiouanus in 1970, none : in 1971. However , it number per zoo declined from 15 to 12: the reported 60 C. n. mantchuricus, all captive-born, number of collections increased from 22 to 30. in 1970 and an estimated 200 in 1971. The total population shown for 1971 was further affected This subspecies appears to be in a strong posi- tion for long-term survival in cantivitv. by lack of a report from Taipeh, which had re- 5. Pere David Deer 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 43 44 45 49 51 54 60 63 Total population 410 432 436 452 485 497 525 550 Captive bred 410 432 436 452 485 497 525 550 Percent captive bred 100 100 100 100 100 100 100 100 Births (surviving) 97 87 104 120 102 27 99 Individuals per collection 10 10 10 9 10 9 9 9 The apparent decline in births for 1969 was vidual reports, 60 percent of the population was caused by a lack of report from Woburn. at Woburn. IZY now reports only totals for this species. This species appears to be ir 1 a reasonably not individual zoo data, which is available from secure position. the studbook. In 1968, last year for the indi- 6. Wisent 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 45 58 64 61 70 82 74 76 Total population 177 234 248 258 249 281 283 303 Captive bred 146 145 154 182 192 193 212 232 Percent captive bred 82 62 62 71 77 69 75 77 Births (surviving) 24 31 34 30 47 51 44 min Individuals per collection 4 4 4 4 4 3 4 4 The total population has increased, the aver- population increase has been stow, the species age per collection remaining static. Though the seems secure. 114 New York Zoological Society: Zoologica, Fall, 1972 7. Scimitar-horned Oryx 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 7 11 9 10 14 18 23 25 Total population 11 23 22 27 53 92 ca.l25 141 Captive bred 8 16 16 15 14 44 73 101 Percent captive bred 73 70 73 56 26 48 58 72 Births (surviving) 4 4 8 5 26 23 29 - Individuals per collection 2 2 2 3 4 5 6 6 The captive population of the scimitar-horned oryx has increased almost explosively since 1967, leaping from 27 to 141 individuals. From 1967 to 1968, the wild-caught population in- creased from 12 to 39, reaching a peak of 52 in 1970. Since 1968, the number of captive-bred individuals has risen from 14 to 101, and the percentage of captive-bred individuals has been rising rapidly. The average number of individ- uals per collection has also increased. If the trends continue, this species will be in a strong position for the future. 8. Addax 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 12 17 18 19 17 21 24 27 Total population 59 63 55 72 75 93 126 142 Captive bred 30 42 32 42 31 58 81 116 Percent captive bred 51 67 58 58 41 62 64 82 Births (surviving) 8 16 15 18 24 29 29 Individuals per collection 5 4 3 4 4 4 5 5 The reported wild-caught population has fluc- tuated from year to year, reaching a peak of 45 in 1970, declining to 26 in 1971. The captive- bred population has increased rapidly since * * In seven of these eight cases, captive breeding seems to have established reasonable security for the species, or nearly so. It is interesting that seven of the eight are hoofed animals, which require more zoo space than most smaller mammals. When the zoo-by-zoo data is analyzed, it appears that the collections with the largest numbers of a species tend to produce dispro- portionately large shares of the births. One rea- son for this is that the general averages are depressed by the number of collections having only one sex. In a number of cases, an increase 1968. While this species has not yet attained the total numbers of the wisent or Pere David deer, the position is becoming stronger. * in the number of collections is accompanied by an apparent decline in the average birth rate. This may be because a collection just acquiring the species may not have both sexes or it may have acquired a pair not yet of breeding age. Among the 33 other species in the initial table, a number show promising population increases. Five have total populations of more than 100. For nine others the percentage of captive-bred exceeds 50. We have chosen nine additional cases from the 33, not by formula but because of their special interest; 1. Golden Marmoset 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting ND ND 23 27 28 24 23 20 Total population ND ND 72 99 102 96 84 76 Captive bred ND ND 6 8 19 22 34 39 Percent captive bred ND ND 8 8 19 23 40 51 Births (surviving) 5 min 7 5 10 10 18 11 — Individuals per collection ND = No data available ND ND 3 4 4 4 4 4 Perry, Bridgwater & Horseman: Captive Propagation 115 The population has decreased since 1968. While the percentage of captive-bred individu- als has risen sharply, this is not in itself a hope- ful sign. Since imports of new stock have been cut off, this percentage could rise to 100 per- cent while the number in captivity approached zero. 2. Orangutan 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 96 107 120 141 120 119 117 128 Total population 278 349 389 438 434 455 469 539 Captive bred 37 44 46 55 68 81 112 152 Percent captive bred 13 13 12 13 16 18 24 28 Births (surviving) 6 9 21 19 21 28 30 Individuals per collection 3 3 3 3 4 4 4 5 The apparent population increase of 70 in Many wild-caught orangutans were acquired 1971 was largely caused by reporting incon- within a few years preceding 1967. The wild- gruities. caught population outside Indonesia reached a The percentage of captive-born individuals peak in 1967 and is now slowly declining. Thus has been rising slowly, as has the number of far, captive births have more than offset this births. The apparent birth rate has remained decline, but it will be several years more before relatively stable since 1966. the likelihood of survival in captivity can be assessed. 3. Sumatran Tiger 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 20 11 27 34 30 30 28 29 Total population 44 23 50 86 66 65 62 78 Captive bred 24 5 24 42 42 42 48 59 Percent captive bred 55 22 48 49 64 65 77 76 Births (surviving) 1 6 7 18 12 9 12 - Individuals per collection 2 2 2 3 2 2 2 3 While there have been reporting inconsisten- ber of births has not significantly increased. The cies, the population decline following the 1967 apparent birth rate over seven years has been peak seems to be real. The wild-caught total substantially below that of the Siberian tiger. has declined from a peak of 44 to 19. The num- 4. Snow Leopard 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 27 28 28 33 39 42 39 44 Total population 49 54 54 64 90 96 93 98 Captive bred 8 4 3 15 15 20 29 31 Percent captive bred 16 7 6 23 17 21 31 32 Births (surviving) 3 1 6 15 7 10 7 - Individuals per collection 2 2 2 2 2 2 2 2 The population of this species has increased chiefly through acquisitions from the wild. Only a modest increase has occurred since 1968. The number of births does not show an upward trend. The average number per collection has remained static, at two. Of the 44 collections, 11 had only one sex in 1971. 116 New York Zoological Society: Zoologica, Fall, 1972 5. Hartmann Mountain Zebra 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 21 20 22 28 28 33 34 29 Total population 80 72 78 86 81 94 84 91 + Captive bred 21 24 32 35 36 38 42 37 + Percent captive bred 26 33 41 41 44 40 50 41 Births (surviving) 9 7 14 10 9 min 9 min 10 - Individuals per collection 4 4 4 3 3 3 2 3 Total population has fluctuated only slightly during this period. The captive-bred numbers have changed only slightly since 1967. Of the 29 collections, 10 have only one sex. While 68 successful births were reported, the captive- born population increased by only 16. 6. Black Rhinoceros 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 63 66 72 68 67 72 71 65 + Total population 113 124 132 126 126 136 130 128 + Captive bred 18 16 23 21 21 24 27 28 Percent captive bred 16 13 17 17 17 18 21 22 Births (surviving) 2 6 1 3 6 7 9 — Individuals per collection 2 2 2 2 2 2 2 2 Total population fluctuated only slightly dur- ing the period. There was a modest increase in the number and percentage of zoo-born indi- viduals. While 34 successful births were re- ported, the zoo-born total increased by only 10. The average number per collection remained static. Of the 65 collections reporting in 1971, 25 had only one sex. 7. Pygmy Hippopotamus 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 31 33 38 44 43 47 46 48 Total population 80 85 99 103 + 108 124 126 128 + Captive bred 35 38 42 45 44 55 50 58 + Percent captive bred 44 45 42 44 41 44 40 45 Births (surviving) 6 12 7 3 11 5 8 - Individuals per collection 3 3 3 2 3 3 3 3 This species came close to the arbitrary selec- tion factors: 100 or more individuals, 50 per- cent or more captive-born. In the period shown, the captive population increased by 48, the captive-born total by 23. The number of births reported during the period was 52. The percentage of captive-born individuals remained remarkably static. Births averaged 7.4 per year, the actual number fluctuating from year to year. The apparent birth rate tended to decline. The average number per zoo remained static. 8. Vicuna 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 28 32 ND ND 23 20 22 22 Total population 69 72 ND ND 64 70 70 69 Captive bred 34 38 ND ND 43 53 44 57 Percent captive bred 49 53 - - 67 76 63 83 Births (surviving) 7 4 7 4 10 3 min 5 — Individuals per collection 2 2 - - 3 4 3 3 Perry, Bridgwater & Horseman: Captive Propagation 1 17 The percentage of captive-bred individuals has risen, but total population has not increased. The number of births has fluctuated from year to year, sex. Nine of the collections have only one 9. Arabian Oryx 1964 1965 1966 1967 1968 1969 1970 1971 No. zoos reporting 4 4 5 5 6 5 4 4 Total population 27 27 39 49 44 48 58 75 Captive bred 2 7 10 9 9 18 18 49 Percent captive bred 7 26 26 18 20 38 31 65 Births (surviving) 2 3 min 3 min 1 5 4 6 — Individuals per collection 7 7 8 10 7 10 15 19 The increase from 1970 to 1971 in the num- ber of captive-bred individuals is distorted by the report from Qatar, which failed to report this item in 1970, but reported 21 captive-bred ani- mals in 1971. The significant increase is in the captive-bred totals for Phoenix and Los An- geles: from 18 in 1970 to 28 in 1971. For these * On the record to date, zoos have not become a significant resource in the preservation of rare or endangered mammals. Seven species or sub- species endangered or extinct in the wild appear to have reasonably secure captive populations with potential for reintroduction. In a few other cases, favorable trends give promise of security in the near future. While these are important contributions, their number is small by compari- son with lUCN’s long and growing list. The data also indicate that zoos can become a more significant resource. The chief deficiency is managerial, not scientific. Zoos have learned, over the years, how to keep most species alive and healthy in captivity, and how to breed them. Many of these species would undoubtedly mul- tiply to satisfactory numbers if adequate breed- ing groups were brought together under proper conditions. While some species now present spe- cial problems, such as inadequate second- generation reproduction, most should be respon- sive to concerted efforts. The troublesome problem is that many spe- cies which reproduce adequately under good management do not have self-sustaining captive populations. That many zoos report only single males or single females is only part of the problem. A zoo with one of each does not necessarily have a breeding pair. The problem centers in the zoos that do the best job of propagation but, for lack of space, are compelled to dispose of offspring. Too often these offspring are sent to zoos with lesser resources and qualifications, zoos that may wish them only for exhibition. A breeding pair may produce offspring for two collections alone, the percentage of captive- bred individuals was 62 in 1970, 74 in 1971. The total captive population has increased rapidly, with a significant increase in the average number per collection. Births show an upward trend. * * several years. Then, if the male or female is lost, no replacement may be readily at hand. Further, many zoo directors report difficulty in finding takers for their surplus. They might prefer to send their animals to excellent zoos that empha- size breeding, but such discrimination may not be possible. One zoo deliberately prevented matings of an endangered species because its pens were overcrowded by the previous year’s surplus. The capacity of zoos is limited, and most still emphasize diversity in collections. A random selection of ten leading zoos shows an average number of individuals per mammal species of 3.9, the range being from 3.1 to 4.6. Since this average includes over-age individuals, non- breeders, and juveniles, there are, inevitably, many situations without breeding potential. Increased propagation of endangered species is feasible, but it may be stifled or become futile unless progeny can be accommodated in their natal zoos or in others willing and able to fur- ther propagation. Room for growing numbers must be found, either by displacing more com- mon species or by establishing rural survival centers. Literature Cited International Zoo Yearbook 1963-1972. Zoological Society of London, Lon- don, England Red Data Book 1966. Volume I. Mammalia. Noel Simon, ed. International Union for the Conservation of Nature and Natural Resources, Morges, Switzerland ‘■k 8 Status of Rare and Endangered Birds in Captivity with a General Reference to Mammals Donald D. Bridgwater Minnesota State Zoological Gardens, St. Paul, Minnesota 55155 Of the 340 bird forms reported by the lUCN as rare and endangered, 62 were reported in captivity from 1964 to 1970. Among the 62 captive forms, 30 bred once dr more in captivity but only 24 bred with frequency. Significant captive breeding success occurred primarily in the Anseriformes, Galliformes, and Psittaciformes. Only nine forms appear secure with regard to captive numbers and reproductive rate. Introduction VIRTUALLY EVERY ZOO has repeatedly stated as a major objective the captive propaga- tion of endangered species as a mecha- nism for saving such animals from extinction. This goal has served as a justification for the possession of endangered animals. The Interna- tional Union for the Conservation of Nature and Natural Resources (lUCN) has endorsed this principle as one viable alternative to extinction. This paper represents an effort to evaluate the captive status of rare and endangered birds from 1964 through 1970 currently listed in the lUCN Red Data Book. A general analysis is given for mammals during the same period. Methods Data for each form listed in the current bird and mammal Red Data Book lists were compiled from the International Zoo Yearbook’s (IZY) lists of rare and endangered animals and animals bred in captivity. Volumes 5 through 11. Areas examined for each species were : 1 ) number of exhibiting zoos; 2) number of zoos reporting births or hatchings; 3) total species number; 4) total number captive born surviving; 5) num- ber born for each year; and 6) number of zoos possessing a breeding potential. Although each year was analyzed, the tables of this paper were prepared using two-year intervals. Rather than indicate sex, a number representing reproduc- tion potential is used. It is derived for each spe- cies by counting the number of zoos indicating the possession of both sexes and/or breeding success. When this number is compared to the number of births and number of young bom, a statement of trend can be made. Birth data is always given for the year preceding the year- book volume and has been adjusted accordingly. It must be recognized that figures presented here are in many cases incomplete or inaccurate due to several reasons, including 1) irregular reporting by institutions; 2) taxonomic misiden- tification; 3) failure to provide numerical assess- ments for individual species; 4) failure of some countries to report; 5) lack of information from private breeders; 6) lack of information on ages of stock; and 7) lack of consistent survival cri- teria. Because of these factors, the data presented here are simply indicators of trends, but it is felt that they do reflect with some accuracy the cap- tive status of rare and endangered birds and mammals in major zoo facilities from 1964 through 1970. Aves General Comments. For the 27 orders of birds, seven (Struthioniformes, Rheiformes, Casuariiformes, Apterygiformes, Gaviiformes, Coliiformes, and Trogoniformes) contain no forms considered rare and endangered for the purposes of this paper (Table 1 ) . The remaining 20 orders contain 340 rare and endangered forms. Nine orders were represented in captivity by 62 forms (18.2 percent of all endangered forms). Six orders (Ciconiiformes, Anseri- formes, Galliformes, Gruiformes, Psittaciformes, and Passeriformes) contain 30 breeding forms which is 48.9 percent of all forms exhibited, but only 8.8 percent of the total number of rare and endangered species. For the 30 breeding forms represented in captivity, only 24 have bred with any consistency and quantity. Rare and endan- gered forms were represented in captivity in three additional orders (Sphenisciformes, Fal- 119 120 New York Zoological Society: Zoologica, Fall, 1972 coniformes, and Columbiformes) but no breed- Japan has so far not proven successful. The red- ing was reported. cheeked ibis or waldrapp (Table 2) is repro- Those orders containing rare and endangered ducing primarily from two large groups at Basle forms which have been reported in captivity and Innsbruck. during 1964 through 1970 are reviewed below. Anseriformes. Thirteen forms (9 sp., 4 ssp.) Sphenisciformes. The Galapagos penguin is are rare and endangered and nine were reported the single Red Book representative and was in- captivity. Eight bred frequently, while the frequently reported in captivity, only once with white-winged wood duck reported only at Slim- breeding potential. bridge in 1964, 1969, and 1970, has been less Ciconiiformes. Five forms are rare and en- than successful. Table 3 indicates status of the dangered (4 sp., 1 ssp.). The Japanese white remaining eight forms. The Hawaiian goose is a stork data is incomplete, lacking in data from classic captive-breeding success story (Fisher Chinese zoos. A captive breeding project in et al., 1969), with stock, breeding potentials. Table 1. *Ordinal Summary of Captive Status for Rare and Endangered Birds — 1964-1970 Order lUCN Forms Forms in Captivity Rarely Captive Breeding Frequent Total Sphenisciformes 1 1 0 0 0 Struthioniformes 0 0 0 0 0 Rheiformes 0 0 0 0 0 Casuariiformes 0 0 0 0 0 Apterygiformes 0 0 0 0 0 Tinamiformes 2 0 0 0 0 Gaviiformes 0 0 0 0 0 Podicipediformes 5 0 0 0 0 Procellariiformes 6 0 0 0 0 Pelecaniformes 2 0 0 0 0 Ciconiiformes 5 2 0 1 1 Anseriformes 13 9 1 8 9 Falconiformes 21 4 0 0 0 Galliformes 32 19 1 10 11 Gruiformes 21 9 2 1 3 Charadriiformes 11 0 0 0 0 Columbiformes 16 2 0 0 0 Psittaciformes 29 11 1 3 4 Cuculiformes 3 0 0 0 0 Strigiformes 9 0 0 0 0 Caprimulgiformes 2 0 0 0 0 Apodiformes 15 0 0 0 0 Coliiformes 0 0 0 0 0 Trogoniformes 0 0 0 0 0 Coraciiformes 2 0 0 0 0 Piciformes 9 0 0 0 0 Passeriformes 136 5 1 1 2 Total 340 62 6 24 30 does not include the Arabian ostrich Table 2. Analysis of Rare and Endangered Ciconiiformes Frequently Breeding in Captivity Population Number of Zoos Percent Species Year Total Breed. Poten. Births Total Births Captive Born Geronticus eremita 1964 *n.d. n.d. 3 n.d. 11+ n.d. (Red-cheeked ibis) 1966 10 4+ 3 50(28) ? 56 1968 n.d. n.d. 5 n.d. 10+ n.d. 1970 14 1+ 77(58)-h 75 * n.d. = no data Bridgwater: Status of Birds in Captivity 121 and numbers of exhibiting institutions increas- ing, and the native habitat restocked. Encourag- ing success is also true for the cereopsis, Cuban tree duck, Laysan duck, and koloa. However, in all cases small numbers of wild-caught birds continue to be taken. Data is too limited for judgment on the Mexican duck and Aleutian Canada goose. The New Zealand brown teal, although breeding at Slimbridge and Peakirk, is not increasing. It is felt that the bulk of captive waterfowl breeding success is largely attributable to special waterfowl facilities (such as Slim- bridge and Cleres) and private breeding opera- tions. The latter may be holding numbers of these species not reported to the yearbook. Falconiformes. Twenty-one forms (11 sp., 10 ssp. ) are endangered. Four were reported in captivity, none breeding in 1964 through 1970. The southern bald eagle, doubtless represented in captivity, either could not be or was not dif- ferentiated in captivity. The California condor is represented by a single Los Angeles bird. The monkey-eating eagle was exhibited in propor- tionately large numbers, as follows: 1964 17 birds 1 reported pair 1965 25 “ 0 “ 1966 18 “ 0 1967 12 “ 1 1968 11 “ ? 1969 13 “ 1 1970 14 “ 1 The Hawaiian hawk has averaged three captive specimens each year since 1’964, but infrequently as a potentially breeding pair. Table 3. Analysis of Rare and Endangered Anseriformes Frequently Breeding in Captivity Species Year Number of Zoos Breed. Total Poten, Births Population Total Births Percent Captive Born Branta sandvicensis 1964 15 14 2 127(127) 1+ 100 (Hawaiian goose) 1966 16 9 3 93 (93) 15+ 100 1968 16 13 2 153(153) 74 100 1970 18 16 173(172) 99 Cereopsis novae-holhmdiae 1964 38 33 16 166(107) 56 64 (Cereopsis) 1966 63 51 18 246(117) 52 48 1968 72 58 23 326(168) 74 52 1970 74 60 341(236) 70 Dendrocygna arborea 1964 *n.d. n.d. 3 n.d. 25 n.d. (Cuban tree duck ) 1966 23 19 6 142 (34) 22 24 1968 36 26 5 185 (75) 37 40 1970 37 26 183 (71) 39 Anas aucklandica chlorotis 1964 1 1 7 7 7 7 ( New Zealand brown teal ) 1966 4 3 1 25 (23) 23 92 1968 5 3 1 15 (10) 6 67 1970 3 1 12 (11) 92 Anas diazi 1964 n.d. 7 (Mexican duck ) 1966 n.d. 7 1964 5 4 n.d. 17 (10) n.d. 59 1970 6 5 19 (10) 53 Anas laysanensis 1964 10 10 3 107 (66) 17 62 (Laysan duck ) 1966 28 25 16 174(107) 14 61 1968 36 32 14 245(165) 64 67 1970 48 41 292(233) 80 Anas platyrhynchos wyvilliana 1964 7 6 3 45 (34) 17 76 (Hawaiian duck) 1966 11 9 4 52 (37) 14 71 1968 12 12 7 134 (39) 45 29 1970 18 18 216(168) 78 Branta canadensis leucopareia 1964 *n.d. (Aleutian Canadian goose) 1966 n.d. 1968 4 2 1 8 (4) 3 50 1970 7 4 18 (9) 50 * n.d. = no data 122 New York Zoological Society: Zoologica, tall, 1972 Galliformes. Thirty-two forms (20 sp., 12 ssp.) are reported endangered. Nineteen were exhibited in captivity, 1 1 with some reproductive success. Those species exhibited but not breeding were Asian forms which are extremely rare and virtually impossible to obtain, including Blyth’s tragopan, western tragopan, Chinese monal, Sclater’s monal, imperial pheasant, and Malay- sian peacock pheasant. Also included were the greater prairie chicken and lesser prairie chicken (Table 4). Data from private breeders is no doubt lacking. White-eared pheasant, mikado pheasant, Hume’s pheasant, and palawan peacock pheas- ant all indicate consistent captive gains with small numbers being acquired from the wild except for the white-eared pheasant. Elliot’s pheasant, Edwards’ pheasant, brown- eared pheasant, and Cabot’s tragopan are essen- tially holding their own. Table 4. Analysis of Rare and Endangered Galliformes Frequently Breeding in Captivity Species Year Number of Zoos Breed. Total Poten. Births Population Total Births Percent Captive Born Tragopan caboti 1964 13 8 1 26 (10) 3 38 (Cabot’s tragopan) 1966 1 1 1 8 (5) 7 63 1968 2 2 2 13 (10) 5 77 1970 1 1 9 (8) 89 Crossoptilon c. crossoptilon 1964 *n.d. (White-eared pheasant) 1966 12 8 1 37 (1) ? 3 1968 12 5 1 18 (3) 7 17 1970 7 6 32 (23) 72 Crossoptilon mantchuricum 1964 24 21 4 79 (62) 23 78 (Brown-eared pheasant) 1966 26 22 3 75 (29) 17 39 1968 33 24 2 91 (49) 1 54 1970 24 15 62 (41) 66 Lophura edwardsi 1964 23 15 2 52 (41) 2 79 (Edward’s pheasant) 1966 28 16 2 82 (49) 4 60 1968 23 15 6 66 (38) 29 58 1970 23 14 69 (49) 70 Lophura swinhoii 1964 65 52 19 286(203) 82 71 (Swinhoe’s pheasant) 1966 88 ? 19 343(189) 75 + 55 1968 91 7 23 339(196) 97 55 1970 *n.d. n.d. n.d. ? Syrmaticus ellioti 1964 38 31 10 138(115) 35 + 83 (Elliott’s pheasant) 1966 47 37 10 169 (98) 86 58 1968 52 39 8 216(154) 64 71 1970 36 27 114 (81) 71 Syrmaticus h. humiae 1964 9 8 2 27 (6) 20 22 (Hume’s pheasant) 1966 15 13 7 84 (42) 107 50 1968 24 20 7 133 (83) 87 62 1970 35 29 125 (99) 79 Syrmaticus mikado 1964 13 9 1 30 (12) 0 40 (Mikado pheasant) 1966 18 8 2 31 (4) 9 13 1968 25 17 2 73 (37) 40 51 1970 32 24 224(181) 81 Polyplectron emphanum 1964 *n.d. n.d. 1 n.d. 2 n.d. (Palawan peacock pheasant) 1966 20 16 3 78 (19) 7 24 1968 20 16 3 80 (19) 13 24 1970 26 14 72 (23) 32 Catreus wallichii 1964 n.d. n.d. 4 n.d. 4 + n.d. (Cheer pheasant) 1966 n.d. n.d. 6 n.d. 50 + n.d. 1968 n.d. n.d. 5 n.d. 49 n.d. 1970 21 16 99 (80) 81 * n.d. = no data Bridgwater: Status of Birds in Captivity 123 Cabot’s tragopan indicates no additions from the wild and virtually a totally captive-born group. Elliot’s pheasant, Edwards’ pheasant, and brown-eared pheasant are not thought to have been imported recently and the wild-caught numbers probably reflect failure of reporting in- stitutions to indicate origin. Swinhoe’s pheasant, although remaining stable in zoos, is being man- aged well on Taiwan, including one re-introduc- tion (six pairs) by the Ornamental Pheasant Trust. If one assumes some numbers of birds being released to or held by private breeders, it is probably self-supporting. Insufficient data makes comment on the Cheer pheasant difficult, although it is breeding well. The number of institutions exhibiting these species and/or possessing breeding potential is declining or remaining stable in all cases except, Swinhoe’s pheasant, Hume’s pheasant, mikado pheasant, and palawan peacock pheasant. Other than the breeding program at the Arizona- Sonora Desert Museum in the early 1960s, no masked-bobwhite were reported until the 1970 census (two individuals in two zoos). Gruiformes. Twenty-one forms (14 sp., 7 ssp.) are Red Book species. Nine are reported in IZY censuses, with three reported to have bred in captivity. Data on the whooping crane is famil- iar and not included here, San Antonio having the only recent zoo success. A major federal pro- gram is underway at Patuxent, Maryland. One successful hatch is reported for the Kagu. The Japanese crane (Table 5) indicates some cap- tive breeding success, although data indicates decline in all categories. Small numbers of Florida sandhill crane, Siberian crane, black-necked crane, horned coot, and Hawaiian gallinule were reported in captiv- ity with no breeding. Most distressing was the dramatic increase in hooded crane numbers without known pairings, as follows; 1964 4 birds 1 breeding potential 1965 4 “ 1 1966 23 “ 5 1967 46 “ 10 1968 73 “ 13 1969 96 “ 16 1970 85 “ 16 Columbiformes. Of the 16 endangered forms, two species of pigeon (great pigeon and Mindoro imperial pigeon) were noted in captivity in small numbers with no reported breeding. Psittaciformes. There are 29 endangered forms (10 sp., 9 ssp.). Eleven were reported in captivity, with four showing reproductive suc- cess. Table 5 shows data for the three most suc- cessful forms, all indicating upward trends in captive breeding success (hooded paradise para- kee, turquoise parakeet, splendid parakeet). Table 5. Analysis of Rare and Endangered Gruiformes, Psittaciformes and Passeriformes Frequently Breeding in Captivity Species Year Number of Zoos Breed. Total Poten. Births Population Total Births Percent Captive Born Grus japonensis 1964 20 13 1 50 (20) 1 40 (Japanese crane) 1966 22 13 2 60 (13) 1 22 1968 16 8 1 34 (19) 1 56 1970 18 8 33 (12) 36 Psephotus chrysopterygius 1964 '•''n.d. n.d. 1 n.d. ? 7 dissimilis 1966 n.d. n.d. 1 n.d. 3 9 (Hooded paradise parakeet) 1968 5 4 2 20 (14) 5 70 1970 4 3 36 (25) 69 N eophema pulchella 1964 6 4 3 18 (17) 15 + 94 (Turquoise parakeet) 1966 20 16 2 104 (50) 9 48 1968 23 20 6 117 (66) 22 56 1970 29 24 145 (85) 59 N eophema splendida 1964 6 3 1 13 (9) 5 69 (Splendid parakeet) 1966 8 5 6 46 (33) 60 72 1968 11 10 3 48 (26) 30 54 1970 14 11 101 (77) 76 Leucopsar rothschildi 1964 17 11 3 50 (5) 8 + 10 (Rothschild’s myna) 1966 33 15 3 111 (17) 6 + 15 1968 36 23 5 115 (39) 11 34 1970 43 19 114 (48) 42 * n.d. = no data 124 New York Zoological Society: Zoologica, Fall, 1972 The thick-billed parrot bred infrequently (three times). Passeriformes. One-hundred-thirty-six species and subspecies of perching birds are reported as rare and endangered. Five were reported in captivity with two reported as breeding, includ- ing an isolated hatch of grey-necked rock fowl (Picathartes) and Rothschild’s myna. Data for the latter is in Table 5. Captive reproductive success is increasing. Wild imports have slowed. Mammals Table 6 provides an ordinal summary for mammals. For the 19 orders of mammals, five (Monotremata, Dermoptera, Pholidota, Tubu- lidentata, and Hyracoidea) contain no rare and endangered forms. The remaining 14 orders con- tain 291 rare and endangered forms, of which 162 (55.6 percent) were represented in captiv- ity. Endangered whales and bats were not repre- sented. Eighty-seven forms indicated breeding in captivity (59 breeding regularly and 28 with three or less birth occurrences during the study period). This is 53.7 percent of all forms ex- hibited and 22.2 percent of all endangered forms. While by comparison with birds the overall picture looks better, a detailed analysis for mam- mals (Perry, Bridgwater, and Horseman, 1972) indicates that only three mammals (wisent, Pere David deer, and Przewalski horse) were entirely supported by captive breeding programs, while only the mongoose lemur, Siberian tiger. onager, Eormosan sika, and addax were close to self-support. The remaining forms were either being supported by wild stock only, increasing in captivity much too slowly, or diminishing. Discussion and Summary Among the 62 endangered bird forms reported in captivity during the period of the study, 32 are totally dependent upon wild acquisition. Six others have shown only one to three breeding occurrences. Status of the remaining 24 forms varies from the successful saving of the nene to the slow decline of Cabot’s tragopan without the infusion of wild stock. Only the Anseriformes, Galliformes, and to a lesser degree, the Psittaciformes indicate signifi- cant reproductive success. Generally, both the number of zoos exhibiting endangered species and the number of zoos with reproductive potentials are increasing; however, only a very few specialty institutions such as the Wildfowl Trust and the Ornamental Pheas- ant Trust and similar private institutions are responsible for the bulk of captive-bred indi- viduals. The ratio of potential breeding groups to ac- tual birth events is low and stable, while the number of zoos actually exhibiting endangered forms is increasing. An arbitrary attempt was made to identify those species where zoo propagation gives some assurance that they could be effectively main- Table 6. Ordinal Summary of Captive Status for Rare and Endangered Mammals 1964-1970 Order lUCN Forms Forms in Captivity iRarely Captive Breeding Frequent Total Monotremata 0 0 0 0 0 Marsupialia 30 4 1 3 4 Insectivora 4 1 0 0 0 Dermoptera 0 0 0 0 0 Chiroptera 2 0 0 0 0 Primates 46 32 12 8 20 Edentata 6 5 0 1 1 Pholidota 0 0 0 0 0 Lagomorpha 4 1 1 0 1 Rodentia 24 8 4 0 4 Cetacea 8 0 0 0 0 Carnivora 256 31 6 13 19 Pinnipedia 11 6 0 0 0 Tubulidentata 0 0 0 0 0 Proboscidea 1 1 0 0 0 Hyracoidea 0 0 0 0 0 Sirenia 5 5 0 0 0 Perissodactyla 17 14 1 11 12 Artiodactyla 77 44 3 23 26 Total 291 162 28 59 87 ’ Less than 3 births reported ^ Includes Bengal tiger Bridgwater: Status of Birds in Captivity 125 tained without wild infusions. The following criteria were applied to the 62 captive bird forms ; 1 ) current captive population of at least 125; 2) 50 percent of total population captive born; and 3) present at 20 institutions with breeding potential. Using these criteria, only Swinhoe’s pheasant, mikado pheasant, Hume’s pheasant, Elliott’s pheasant, nene, cereopsis, Laysan duck, Ha- waiian duck, Cuban tree duck, and turquoise parakeet qualify. These criteria are no doubt minimal, and if breeding programs are not quickly effected con- centrating upon captive survival, and if the ex- penditure of effort, time, money, and technical resources is not provided by joint effort and at the expense of an “animals for exhibit only’’ philosophy, then zoos may be in danger of fail- ing in their preservation and propagation objectives. Literature Cited Fisher, J., N. Simon, and J. Vincent 1969. Wildlife in danger. Viking Press, New York. 368 pp. International Zoo Yearbook 1964-1970a. Census of rare animals, Vols. 6-11. Zoological Society of London, England. 1964-1970b. Species of wild animals, Vols. 6-11. Zoological Society of London, England. Red Data Book 1966a. Aves, Vol. 2. International Union for Con- servation of Nature and Natural Re- sources, Morges, Switzerland. 1966b. Mammals, Vol. 1. International Union for Conservation of Nature and Natural Re- sources, Morges, Switzerland. Perry, J., D. Bridgwater, and D. L. Horsemen 1972. Captive propagation; a progress report. Zool. 57(3): 109-117. 126 New York Zoological Society: Zoologica, Fall, 1972 Plate I. Philippine Cobra, Naja naja philippinensis. 9 Venom Yields of the Philippine Cobra, Naja naja philippinensis (Plate I; Text-figure 1; Tables I-II) Enrique S. Salafranca Serum and Vaccine Laboratories, Bureau of Research and Laboratories, Department of Health, Alabang, Muntinlupa, Rizal, Republic of the Philippines. Data on venom yields of 150 cobras (Naja naja philippinensis) gave an overall average venom yield per cobra per extraction (AVY/C/E) of 0.33 ml or 70.14 mg and an overall average total lifetime venom yield per cobra ( ATLVY/C) of 2.23 ml or 527.77 mg for the fresh and corresponding dried venom, respectively. The ATLVY/C were greater at schedule every 14 days than those at schedule every 7, 21, or 28 days. The all-male groups gave bigger yields than the corresponding all- female groups. The AVY/C/E in the former increased as the intervals between extrac- tions lengthened. In the latter the reverse was observed. The solid content of the venom does not appear to be affected by either sex, time of the year, or schedule of collection, but the serial records of extractions show a tendency for this to gradually decline with time. Venom extraction at close intervals resulted in a marked decline of the solid content, the rate of reduction becoming more marked in direct proportion to frequency of extraction. The data on AVY/C/E of several groups indicate a general tendency for this to decline with time. Introduction IN VIEW of the absence of venom suppliers in the Philippines, it is essential that the Serum and Vaccine Laboratories (SVL) undertake the production of the venom to carry on its antivenin preparation program. Eor this pur- pose, a serpentarium is maintained for cobras of the species Naja naja philippinensis, and suf- ficient number of snakes are kept to ensure an adequate and uninterrupted supply of good qual- ity venom. The number of snakes is replenished with cobras caught during field collection trips undertaken every three or four months. To provide a basis for a realistic estimate of the number of cobras needed to supply an ade- quate amount of the venom to meet the require- ments of the laboratory for antivenin produc- tion, records of venom yields are necessary. The present study was conducted for this purpose as well as to enable us to work out a schedule of venom collection which will assure maximum yields with the facilities available. Experiments were designed to furnish information on the possible effects of factors like sex and schedule or frequency of extraction on venom yields. The possibility of trends in yields with time was also investigated. To the clinicians, data on venom yields may serve as aids in making an estimate of the amounts of the specific antivenin which may be required to neutralize the venom injected in a bite. Oliver (1944) states that “the quantity of venom secreted in the act of biting varies accord- ing to the species, the size, age, and the condi- tion of the snake at the time of the bite. In general, the larger the snake, the greater the quantity of venom injected, though there are many exceptions to this generality. The amount of venom injected depends on the time interval since the last bite, the venom glands usually requiring approximately two weeks to regain their maximum capacity of venom. In a normal bite, a snake does not expel its full quantity of venom, but only a small portion and is still capable of inflicting a fatal bite. Evidence indi- cates that an enraged snake injects a greater quantity of venom than one which has not been angered prior to biting. The amount of venom released during a spontaneous bite is greater than that obtained by investigators through ‘milking’ or forced expulsion of the venom.” He gives the following figures on the approxi- mate amounts of dry venom obtained at a single 127 128 New York Zoological Society: Zoologica, Fall, 1972 extraction from common poisonous snakes: North American Species Mg. Copperhead 45- 65 Water moccasin 90-150 Timber rattler 40- 90 Texas rattler 120-300 Florida rattler 240-450 South and Central American Species Tropical rattlesnake 60-150 Fer-de-lance 80-160 Bushmmaster 300-500 C. Indian Species Asiatic cobra 250-350 Russell’s viper 200-300 D. African Species Mamba 50- 80 Puff adder 70-120 E. Australian Species Tiger snake 35- 50 Death adder 60- 80 Christensen ( 1955 ), referring to the works of Grasset, Zoutendyk, and Schaafsma, and Gras- set and Schaafsma gave the following records of yields of venom of various snakes: Snake Average Yield Limit of Indi- vidual Yields Remarks N. flava 0.12 g (128 specimens) Maximum— 0.25 g D. angusticeps 0.1 g N. liaje 0.72 g ( 1 snake) Length— 7'3" An enormous yield S. haemachates 0.42 ml (0.1 g) (over 150 specimens) 0.1 ml-1.8 ml (0.033 g-0.242 g) Solid content just under 25% B. arientans 0.18 g first milking 0.07 g five days in captivity 0.1 g three weeks in captivity Maximum— 0.75 g D. typiis 0.015 g average yield from pair of glands 41% W/W solid content; (obtained direct from gland) Other data on venom yields mentioned by Christensen include: Fifteen full grown puff adders gave an aver- age yield of 0.67 ml (0.186 g). The following table gives data on venom ob- tained from different snakes in his laboratory: Species No Size Cm. Min. Yields* Max. Ave. Percentage** Min. Max. Solid Ave. 5. haemachates 10 90-120 0.11 0.72 0.35 17.5 28.3 23.9 N. flava 1 150 0.11 27.1 N. haje 1 150 0.12 33.5 C. rhombeatus 9 45- 75 0.07 0.75 0.34 19.3 28.1 23.6 B. cornuta 5 32- 35 0.008 0.087 0.045 24.5 36.4 31.1 B. caudalis 1 60 0.085 27.6 B. gabonica 1 113 1.90 26.4 * Venom from B. cornuta and B. caudalis in g, the others in ml. ** W/W for B. cornuta and B. caudalis venom W/V for others. Salafranca: Venom Yields of Philippine Cobra 129 Christensen states that the quantity, composi- tion, and toxicity of the venom of a given spe- cies may vary considerably. He mentions that age, state of health, climate, and even the “mood” of the snake may affect the character and quantity of venom. Conant (1952) states that the snake has some control over the amount of venom injected in a bite. Minton (1957) gives the average amount of venom delivered in a bite by the following venomous species: A. Elapidae Mg. North American coral snake ( Micnirus fulvius ) 3- 5 Blue krait (Biingams camUdiis) 5- 10 Tiger snake (Notechis sciitatiis) 35- 45 Indian cobra (Naja naja) 175-250 Mamba (Dendroaspis angusticeps) 75-100 B. Viperidae African puff adder (Bitis arientans) 160-200 Russell’s viper ( Vipera russelli) 150-250 C. Crotalidae Fer-de-lance (Bothrops atrox) 80-160 Bushmaster (Lachesis muta) 300-500 Western diamondback rattlesnake (Crotahis atrox) 200-300 Deoras (1966) mentions environment, sex, seasonality, and frequency of milking as pos- sible factors affecting venom yields, and in his studies showed that cobras and kraits produced more venom when kept in a farm under condi- tions simulating that of their natural habitat than when kept in separate cages in a room. The vipers however, produced more venom when kept in the room. Materials and Methods A total of 150 cobras, Naja naja philippinen- sis, 69 males and 81 females, with an overall average length of A1 .1 inches collected in the province of Camarines Sur, Luzon Island, were used to provide the data presented in this study. All were apparently healthy and freshly caught at the beginning of each experiment except as indicated. Distribution of specimens into groups for comparison was done in a completely ran- dom manner from batches comprised of speci- mens approximately of same lengths. The care and management of the snakes and the method of venom collection employed, using a beaker with rubber diaphragm, have been previously described (Salafranca, 1967). The freshly col- lected venom was spread thinly in petri dishes and placed inside a dessicator over calcium chloride. The dessicator was sealed and evacu- ated using a high vacuum pump. It was kept in the cold room (4°-10°C) until the dried crys- talline venom could be easily peeled off the glass. This required one to three or more days depending upon the thickness of the layer or the amount of the venom and the condition of the calcium chloride. The schedule of venom collections for any given experiment was ob- served until all the snakes had died. EXPERIMENT I. Venom was collected from a group of 29 cobras, 7 males and 22 females with an average length of 48 inches, every 14 days. The pooled amount was recorded for each collection schedule. The average venom yield per cobra per ex- traction (AVY/C/E) was computed by divid- ing the total amount collected by the number of snakes involved at each extraction. Following the last collection, the overall AVY/C/E and median were computed. The average total lifetime venom yields per cobra (ATLVY/C) was computed by dividing the total amount of venom collected from each schedule group by the number of snakes in- volved in the group. The percent solid was computed by dividing the weight of the dry venom by the weight of the corresponding “wet” venom. EXPERIMENT II. Thirty cobras with an average length of 48.9 inches, 15 males and 15 females, were selected from a catch of 37 speci- mens on the basis of similarities in sizes. Each sex group was randomly divided into three equal groups and each group of males paired at random with a group of females. Each of the three resulting groups was assigned to one of three schedules of venom collection, as follows; Group Schedule of Venom Collection Ila Every 7 days lib Every 14 days lie Every 28 days EXPERIMENT III. Thirty-eight cobras with an average length of 48.9 inches, 19 males and 19 females, were distributed at random into four groups and their venom collected as indicated below: Group Number and Sex Scheduel of Venom Collection Ilia 10 females Every 14 days lllb 10 males Every 14 days lllc 9 females Every 28 days llld 9 males Every 28 days EXPERIMENT IV. Fifty cobras of average length, 46.6 inches, 25 males and 25 females. 130 New York Zoological Society: Zoologica, Fall, 1972 were distributed into six groups and their venom collected according to the schedule indicated below: Group Number and Sex Scheduel of Venom Collection IVa 8 males Every 14 days IVb 8 females Every 14 days IVc 9 males Every 21 days IVd 9 females Every 21 days IVe 8 males Every 28 days IVf 8 females Every 28 days EXPERIMENT V. To determine the capacity for venom production of individual cobras, the following experiment was performed: Three male cobras were selected and subjected to repeated venom extractions at regular inter- vals during a whole working day according to the following schedule: Cobra Length Schedule of Number (inches) Venom Collection 1 45 Every 2 hours 2 43.5 Every hour 3 39 Every half-hour Cobra no. 1 and cobra no. 2 were obtained in our regular periodic hunt and had been in the laboratory serpentarium 22 days at the time of this experiment. Cobra no. 3 was a specimen caught within the laboratory premises two days before this experiment. A 50-ml beaker of known weight was assigned to each snake. The venom was extracted as de- scribed previously (Salafranca, 1967). Due to the anticipated difficulty of getting accurate vol- umetric measurements, the weight of the venom collections were determined instead. The initial amount collected was rated 100% and the sub- sequent collections as percentages of the initial collection. The data on the serial amounts of venom col- lected from certain groups in experiments one to four were plotted in a graph to determine possible trends with time. Results and Discussion The results of Experiments I to IV are sum- marized in Table I. A comparison of the ATLVY/C in three comparable groups of co- bras, Ila, Ilb, and lie of Experiment II, sub- jected to 7, 14, and 28-day schedules, respec- tively; in four comparable groups. Ilia and Illb, and IIIc and Illd, of Experiment III, subjected to 14 and 28-day schedule, respectively; and in six comparable groups, IVa and IVb, IVc and IVd, and IVe and IVf of Experiment IV, on 14, 21, and 28-day schedules, respectively, shows that, in each case, the average was greater in the group or groups on the every 14-day schedule of extraction. In Experiment II, using the dry venom data, the ATLVY/C on the 14-day schedule (lib) was 16% and 26% (or 12% and 1 1 % on the “wet” data) greater than those on 7-day and 28-day schedules (Ila and lie), respectively. In Experiment III, the all-female and all-male groups on the 14-day schedule (Ilia and Illb) averaged 35% and 37% (or 55% and 15% on the “wet” data) more venom than the corresponding all-female and all-male groups on the 28-day schedule (IIIc and Illd). The data for the all-male and all-female groups on the 14-day schedule in Experiment IV (IVa and IVb) show 15% and 9% and 18% and 25% (or 26% and 17% and 24% and 45% on the “wet” data) greater ATLVY/C than those in the corresponding all-male and all-female groups on the 21 and 28-day schedules (IVc and IVd and IVe and IVf), respectively. From the data on the dry venom yields for comparable all-male and all-female groups, it will be observed that the former consistently gave greater ATLVY/C. In Experiment III, the all-male groups on the 14 and 28-day schedules (Illb and Illd) gave 32% and 73% (or 50% and 102% on the “wet” data) more venom than the corresponding all-female groups (Ilia and IIIc), while in Experiment IV the all-male groups on 14, 21, and 28-day schedules (IVa, IVc and IVe) gave 87%, 93%, and 114% (or 102%, 100%, and 149% on the “wet” data) more than the corresponding all-female groups (IVb, IVd, and IVf). The AVY/C/E on the other hand, in com- parable groups, was observed to increase with the increase in intervals between extraction both in the mixed sex groups as well as in all-male groups. In Experiment II, we have 58.14 mg, 77.6 mg, and 101.73 mg for 7, 14, and 28-day schedules, and in Experiment IV, 72.73 mg, 75.75 mg, and 95.71 mg for the all-male groups on 14, 21, and 28-day schedules, respectively. The increase in AVY/C/E with the increase in intervals between extraction appear logical, since the gland is given correspondingly more time to recover its full capacity. With the all-female groups, however, the order is reversed; that is, the AVY/C/E decreases with increasing inter- vals. Thus we have for the all-female groups in Experiment III, 63.66 mg and 57.1 mg for the 14 and 28-day schedules, and in Experiment IV, 40.74 mg, 40.46 mg, and 35.66 mg for the 14, Salafranca: Venom Yields of Philippine Cobra 131 Table I. Summary of Results, Experiments I-IV. AVY/C/E - Average venom yield per cobra per extraction a - Average ATLVY/C - Average total lifetime venom yield per cobra m - Median 132 New York Zoological Society: Zoologica, Fall, 1972 Cobra I.D. Schedule and Record of Extractions-**- (Wet - mg) Total Amount Collected (Dry - mg) No. - 1 Sex - M Length - 45" Every 2 hours 287.25 (lOOS^) 166.84 ( 585«) 78.38 ( 21%) 94.68 ( 335?) 173.32 ( 60jg) 81.4 % No. - 2 Sex - M Length - 43.5" Every hour 200.94 (1005?) 102.25 ( 515?) 129.46 ( 645?) 144.31 ( 12%) 86.29 ( 435?) 71.60 ( 365?) 53.27 No. - 3 Sex - M Every half hour 480.72 (1005?) 122.13 Length - 39" 433.02 ( 392.47 ( 203.24 ( 163.67 ( 228.29 ( 125.47 ( 161.25 ( 152.06 ( 113.03 ( 142.14 ( 73.63 ( 98.15 ( 74.70 ( 50.25 ( 905?) 825?) 425?) 3 45?) 475?) 265?) 345?) 325?) 245?) 205?) 15^) 205?) 16^) 10^) Solid -**^<- 10^ 1% 45? -**- Amount of venom given are in the chronological order of extractions. Initial extractions are rated 1005? and subsequent extractions as percentages of the initial amount. -*^ Obtained by determining the relation of total, dry weight to total weight of corresponding wet venom. Table II. Summary of Results, Experiment V. Salafranca: Venom Yields of Philippine Cobra 133 r\j I — I o CO nO LPv -4- c; o •H -P O ct3 U (D (D a ctJ to oj U 0) > < w Text-figure 1. Graphic presentation of AVY/C/E* of indicated groups in Experiments I, II, III, and IV. Extraction Number 134 New York Zoological Society: Zoologica, Fall, 1972 21, and 28-day schedules, respectively. The ob- servation that the males give more than the females, however, is as true when we consider the data on ATLVY/C as the data on the AVY/C/E in comparable all-male and all- female groups. The data on average percent solids of the venom collected from the various groups in Experiments I, II, III, and IV do not give indi- cations that this may be affected by either sex, season, (period of the year when the experiment was undertaken), or the schedule of venom ex- traction observed. The records of the serial extraction of the majority of individual groups, however, show a tendency for this to decline gradually as the number of extractions increased. The results of Experiment V are summarized in Table II. It will be observed that the amounts obtained generally decreased in the order of the chronology of extraction. The amounts obtained from cobra no. 3, notwithstanding its smaller size and greater frequency of extraction (every half-hour), are greater. This may be explained by the fact, as pointed to above, that this speci- men at the time of the experiment was only two days in captivity and therefore more vigorous than the other two which have been in captivity 22 days at the time of this experiment. The per- cent solids of the combined extractions from each snake decreased as the frequency of extrac- tion increased, from 10% for the extractions every two hours to 4% for those at every half- hour. The percent solids of 10%, 7%, and 4% obtained for the pooled collections from snakes 1, 2, and 3, respectively, are much lower than the overall average of 22.14 (15.07% to 29.54%) percent solids from the yields in Ex- periments I to IV (Table I). Apparently, re- peated extractions at the close intervals of every two hours, hourly, and every half-hour, results in marked dilution of the venom, the dilution increasing in that order. Graphical examination of the chronological AVY/C/E of several groups included in Ex- periments I, II, III, and IV (Text-figure 1 ) indi- cates a general tendency for yields to decrease with time. In some groups, the drop in yield from the first collection to the second or third is more marked than in others. In a few, there is a rise from the first extractions to the second or third, and from then on a general but gradual tendency to diminish with occasional peaks which are generally lower than the maximum noted for that particular group. Acknowledgment The author wishes to extend his gratitude to Dr. J. S. Sumpaico, Director, Bureau of Re- search and Laboratories for encouraging inves- tigative work and for going over the final draft of the manuscript. Literature Cited Christensen, P. A. 1955. South African snake venom and anti- venoms. The South African Institute for Medical Research, P.O. Box 1038, Johan- nesburg. 4 pp. CONANT, R. 1952. Reptiles and amphibians of the north- eastern states (second edition). The Zoo- logical Society of Philadelphia. 6 pp. Deoras, P. J. 1966. Probable significance of venom yield rec- ord studies. Mem. Int. Butantan, Simp. Internac. 767 pp. Minton, Jr., S. A. 1957. Snakebite. Scientific American, 196. Oliver, J. A. 1944. Clinical tropical medicine. Harper & Brothers, New York. 870 pp. Salafranca, E. S. 1967. Longevity of the Naja naja philippinensis under stress of venom extraction. Zoo- logica, 52:3. NEWS AND NOTES Preliminary Report: Status Investigations of Morelet’s Crocodile in Mexico Field investigations in the State of Veracruz, Mexico, were carried out in early August, 1971, to examine the status of several populations of Morelet’s crocodile and to obtain basic infor- mation on their natural history and ecology. Habitat surveys, population estimates, and behavioral observations, by both day and night, were undertaken in two locations, nr. Alvarado in the mouth of the Papaloapan River and Alvarado Lagoon, and in Lake Catemaco in the Tuxtlas Range. Two field days and one night were spent in the former locality, five days and six nights at the latter. In addition to actual field observations of the crocodiles themselves a special effort was made to contact and become friendly with locals in each area who were reputed to be knowledge- able about crocodiles and/or involved in the crocodile hide industry. For convenience, the two study areas will be treated separately. Lago de Catemaco Habitat. Lake Catemaco is a freshwater lake of some 50 square miles located in the Sierra de las Tuxtlas at approximately 1,100 feet alti- tude. One major river, the Quetzaloapan, enters the lake. This forms a marsh of considerable extent where it runs into the lake on the eastern side. Several islands of several acres each dot the lake and there are several shallow embay- ments scattered along the shore line, the largest of which is the Arroyo Agrio on the north shore east of the town of Catemaco at the Coyame bottling plant. At the current time the crocodile populations are concentrated in the Arroyo Agrio, the Arroyo (Quetzaloapan River marsh), and to a lesser extent in the other embayments and on the islands. The shores of the lake still support considerable tropical vegetation though extensive clearing and other modifications of the natural cover is underway at diverse points around the lake. For the most part this habitat alteration does not involve the immediate lake shore except in a few points where human habi- tations are being constructed near the water’s edge. The lake is drained along the southwest end by a river, which is dammed where it crosses the major highway, Mex. Hwy. 180. Crocodiles currently exist immediately below the dam in a small impoundment and, by report, throughout the river. Estimated Population. Accurate census of the population by night was hindered by clear skies and a full moon. Despite this handicap croco- diles were seen frequently in the Arroyo Agrio and the Arroyo and at scattered localities throughout the lake. These observations are too scattered and erratic to support more than a crude estimate of the population’s size; a con- servative estimate might be 200 crocodiles (all sizes) in the lake. The estimates of local people ranged from several hundred to a thousand. This larger figure is rather unlikely, but cannot be refuted at this time. At the time of this visitation, the young of the year had not yet appeared and three nesting females were located. The nests are mounds of vegetation located on the shore in densely vege- tated areas and are placed from 10 to 40 feet from the water’s edge. They are highly reminis- cent of the nests of the American alligator. The lake houses at least two individuals which approach eight to nine feet in length. One of these frequents the Arroyo Agrio and the other a spit of rock backed by a small marsh south of and adjacent to the outfall of the Quetzaloa- pan River. Hunting Pressure. At the present time, hunt- ing in the lake is minimal. Seven individuals, including a five-foot specimen, are held in cap- tivity in a restaurant, “Las Olas’’, on the lake shore in Catemaco. They are well fed and cared for, and the owner of the restaurant has a posi- tive attitude toward the fauna and discourages the fishermen from molesting the wild popu- lation. All the local fishermen questioned expressed a dismay at the difficulty of marketing hides and claimed they no longer seek crocodiles though they may take one occasionally as a target of opportunity. Skins, they claim, must be sent to San Luis Potosi or Laredo for sale. 135 136 New York Zoological Society: Zoologica, Fall, 1972 One lagartero of considerable repute lives on the lake. It was not possible to determine the extent of this individual’s activities on this visit, though it was claimed he was not hunting at this time. This individual displays a commend- able ecological insight in his pursuit and makes every effort to protect crocodile nests from dis- turbance by other humans in the area. Guides and fishermen in the area were initially reluctant to take me to nests until reassured repeatedly that I only wanted to photograph the nests and would not disturb them. The general consensus in the area was that the primary drain on the lake’s population cur- rently is from visiting “sportsmen” who pay the local boatmen to take them out hunting. Most boatmen are reluctant to do this, but are suscep- tible to economic coercion. Prospectus. Prospects for the crocodile popu- lation in Lake Catemaco appear good if the pointless, random depradations of tourist sports- men can be controlled, and if development ac- tivities around the lake can be regulated to pro- vide a buffer strip of natural area immediately along the lake edge, particularly in the shallow bays and arroyos. An American ornithologist, William Schal- dach, currently lives in Catemaco and is anxious to provide support for conservation activities in the area. He has already been successful in halt- ing the slaughter of water birds along the lake by emphasizing their value as tourist attractions and hopes to extend this coverage to the croco- diles and turtles in the lake as well. His property on the lake shore provides a primitive, but con- venient, research station for visiting biologists. It would appear that Catemaco might be an excellent locale to establish a reserve for More- let’s crocodile if such an undertaking is feasible. Alvarado Alvarado is located on a large lagoon and an extensive salt-to-brackish marsh system which provides a great expanse of habitat for croco- diles of two species, Crocodylus moreletii, con- fined to the fresh-water portions, and C. acutus, in the salt and brackish areas. Several rivers empty into this lagoon, but only the Papaloapan was examined. The town of Alvarado houses a market place where many species of aquatic reptiles can be purchased for food and where crocodiles of both species have always been available. This market had been visited in spring, 1970, and crocodiles were readily available. On the current visit the situation was considerably changed. Few fishermen would admit to hunt- ing crocodiles and none of the animals were available for purchase, at least they were not displayed in the open. All fishermen claimed the topic was a “delicate” one and stated that hides were difficult to sell. Some hide traffic still exists, the chief port of exit being San Luis Potosi, but most fishermen claim that it is no longer a major activity. One night visit was made in the lagoon and up into the Papaloapan River. Five crocodiles were observed but could not be approached for positive identification due to bright moonlight. Two of the crocodiles, both four feet to six feet estimated length, were observed in the lagoon itself and were presumably C. acutus. The re- maining three specimens, in the two to three foot size range, were observed in the predominantly fresh-water situation in the Papaloapan River. My guide claimed these to be pardos, or C. more- letii, and said the specific area we were in con- tained only pardos. This identification, of course, should be considered tentative, but the habitat relationship was proper for C. moreletii. In view of the uncertainties inherent in croco- dile hunting and the unfavorable light levels encountered at this visit, I feel that the sighting of five crocodiles in five hours of boat-time is an encouraging development. Certainly the croc- odile populations in this extensive marsh sys- tem have not been completely hunted out and may prove to be healthy enough to achieve a resurgence given effective protection. Howard W. Campbell, Department of Zool- ogy, University of Florida, Gainesville, Florida 32601.1 1 Present address: Jack McCormack and Associates, 860 Waterloo Road, Devon, Pennsylvania 19333. NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences) . All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable), and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 7. Captive Propagation: A Progress Report. By John Perry, Donald D. Bridgwater, and Dana L. Horsemen 109 8. Status of Rare and Endangered Birds in Captivity with a General Reference to Mammals. By Donald D. Bridgwater 119 9. Venom Yields of the Philippine Cobra, Naja naja philippinensis. By Enrique S. Salafranca. Plate I; Text-figure 1; Tables I-II 127 News and Notes; Preliminary Report: Status Investigations of Morelet’s Crocodile in Mexico. By Howard W. Campbell 135 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx, New York 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, New York. Published March 29, 1973 ZOOLOGICA SCIENTIFIC CONTRIBUTIONS OF THE NEW YORK ZOOLOGICAL SOCIETY VOLUME 57 • ISSUE 4 • WINTER 1972 PUBLISHED BY THE SOCIETY The ZOOLOGICAL PARK, New York NEW YORK ZOOLOGICAL SOCIETY The Zoological Park, Bronx, N. Y. 10460 Laurance S. Rockefeller Chairman, Board of Trustees Robert G. Goelet President Howard Phipps, Jr. Chairman, Executive Committee Henry Clay Frick, II Vice-President George F. Baker, Jr. Vice-President Charles W. Nichols, Jr. Vice-President John Pierrepont Treasurer Augustus G. Paine Secretary ZOOLOGICA STAFF Simon Dresner, Editor & Curator, P ublications and Public Relations Joan Van Haasteren, Assistant Curator, Publications Public Relations F. Wayne King Scientific Editor AAZPA SCIENTIFIC ADVISORY COMMITTEE Ingeborg Poglayen, Louisville (Kentucky) Zoological Gardens, Chairman-, William G. Conway, New York Zoological Society; Gordon Hubbell, Crandon Park Zoo, Miami; F. Wayne King, New York Zoological Society; John Mehrtens, Columbia (South Carolina) Zoological Park; Gunter Voss, Metro Toronto Zoo, Canada William G. Conway General Director ZOOLOGICAL PARK William G. Conway, Director & Chairman, Dept, of Ornithology: Joseph Bell, Curator, Ornithology; Donald F. Bruning, Associate Curator, Ornithoolgy; Hugh B. House, Curator, Mammalogy; James G. Doherty, Associate Curator, Mammalogy; F. Wayne King, Curator, Herpetology; John L. Behler, Jr., Assistant Curator, Herpetology; Joseph A. Davis, Scientific Assistant to the Director; Walter Auffenberg, Research Associate in Herpetology; William Bridges, Curator of Publications Emeritus; Grace Davall, Curator Emeritus AQUARIUM James A. Oliver, Director; H. Douglas Kemper, Associate Curator; Chritopher W. Coates, Director Emeritus; Louis Mowbray, Research Associate, Field Biology OSBORN LABORATORIES OF MARINE SCIENCES Ross F. Nigrelli, Senior Scientist; George D. Ruggieri, S.J., Director & Experimental Embryologist; Martin F. Stempien, Jr., Assistant to the Director & Bio-Organic Chemist; Jack T. Cecil, Virologist; Paul J. Cheung, Microbiologist; Joginder S. Chib, Chemist; Kenneth Gold, Marine Ecologist; Myron Jacobs, Neuroanatomist; Klaus D. Kallman, Fish Geneticist; Kathryn S. Pokorny, Electron Microscopist; Eli D. Goldsmith, Scientific Consultant; Erwin J. Ernst, Research Associate, Estuarine & Coastal Ecology: Martin F. Schreibman, Research Associate, Fish Endocrinology CENTER FOR FIELD BIOLOGY AND CONSERVATION Donald F. Bruning, Research Associate; George Schaller, Research Zoologist & Co-ordinator; Thomas Struhsaker, Roger Payne, Research Zoologists: Robert M. Beck, Research Fellow ANIMAL HEALTH Emil P. Dolensek, Veterinarian; Consultants: John Budinger, Pathology; Ben Sheffy, Nutrition; Gary L. Rumsey, Avian Nutrition; Kendall L. Dodge, Ruminant Nutrition; Robert Byck, Pharmacology: Jacques B. Wallach, Clinical Pathology; Edward Garner, Dennis F. Craston, Ralph Stebel, Joseph Conetta, Comparative Pathology & Toxicology; Harold S. Goldman, Radiology,- Roy Bellhorn, Paul Henkind, Alan Friedman, Comparative Ophthalmology; Lucy Clausen, Parasitology; Jay Hyman, Aquatic Mammal Medicine ;ThQodor& KazimiroflF, Dentistry; Alan Belson, Resident in Pathology ® 1973 New York Zoological Society. All rights reserved. 10 Daily Activity Patterns and Effects of Environmental Conditions on the Behavior of the Yellowhead Jawfish, Opistognathus aurifrons with Notes on its Ecology. (Plates I-V; Text-figures 1-21; Tables 1-4) Patrick L. Colin Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, Florida 33149 Quantitative observations of the behavior of Opistognathus aurifrons were made at Bimini, Bahamas, for a period of 25 days using an underwater television system. Various activities of this species were described. The activity of feeding was constant during the day, but was reduced during periods of low light intensity and high current speed. The activities of “digging,” “chasing,” and “arching” varied diurnally, and “digging” and “arching” varied with the current speed. Light was a controlling factor of the uncovering of the burrow in the morning. Light and interspecific relations determined the time of closing the burrow in the evening. Various burrow-oriented activities showed a peak during dawn or dusk periods. The ranges and territories of individuals were determined. The vertical range varies diurnally. The relationship of O. aurifrons to other jawfishes and convergent species of reef fishes was examined. Introduction The yellowhead jawfish, Opistognathus aurifrons (Jordan and Thompson), (Plate I, fig. 1), occurs in the Florida Keys, the Bahamas, and the West Indies (Bohlke and Chaplin, 1968). The species (maximum stand- ard length approximately 94 mm) is found in colonies in areas of calcareous sand substrate near coral outcrops. Its known depth distribu- tion is from 3 to 50 m. Opistognathus aurifrons was originally de- scribed as Gnathypops aurifrons by Jordan and Thompson (1905) from a specimen collected at Dry Tortugas, Florida, and placed into the genus Opisthognathus when these two genera were shown to be synonymous (Meek and Hil- debrand, 1928). Briggs (1961) pointed out that the name Opistognathus was used by Cuvier (1817) and that the variation Opisthognathus was introduced by Cuvier and Valenciennes (1836) and followed in subsequent works. Longley and Hildebrand (1941) included some general information about O. aurifrons. They described the burrow as “perhaps 300-500 mm deep” and “enlarged below, the shape of the terminal chamber being largely fixed by the arrangement of the larger bits of dead coral by which it is surrounded.” One fish was observed in the midst of constructing its burrow and its behavior described. The general feeding behav- ior was described and its food characterized as planktonic. Bohlke and Thomas (1961) redescribed O. aurifrons and placed in its synonymy Gnut/iy- pops bermudezi Howell Rivero. They dealt with geographic variation in certain characters of specimens from Florida, the Bahamas, the Vir- gin Islands, and Cuba. They found Bahamian specimens generally to have black pigment spots on the chin and under the gill membrane at the isthmus, a curved dark line beneath the maxil- lary and the preopercle, and the branchiostegals edged with duskiness. Specimens from Florida lacked all of these markings, and material col- lected in the Virgin Island population was found to be intermediate to the Florida and Bahama populations in the number of branched dorsal, anal, and pectoral-fin rays, length of the lateral 137 138 New York Zoological Society: Zoologica 57(4) line, number of lower gill rakers, and the num- ber of canine teeth. An increase in the number of lower gill rakers with increasing standard length was also found. Bohlke ( 1967) reported that one specimen of a large series taken in the Florida Keys (UMML 18904) showed all the black head markings characteristic of Bahamian individuals. Bohlke and Thomas (1961) also dealt with the tear drop shape of the pupil of O. aurifrons. The long axis of the pupil is oriented horizon- tally when the fish is in a normal vertical “float- ing” or hovering position, and they thought that this, plus the position of the eyes on the head, allows binocular vision horizontally when the fish is in its normal hovering orientation. Bohlke and Thomas ( 1961 ) reported a depth distribution of from 3 to 30 m. Bohlke (1967) modified this distribution to 3 to 36 m for Bahamian specimens and to 41 m for Florida specimens. The writer has observed O. aurifrons at Long Reef, Florida, to a depth of 50 m. Speci- mens in the Florida Keys are found most often on the seaward side of the outer reefs in depths greater than 7 m. In Bimini, Bahamas, O. auri- frons is found on the seaward (west) side of the islands. Randall (1967a) examined the stomach con- tents of 16 specimens from the Virgin Islands and determined their food consisted of 85 per- cent copepods, 9.4 percent shrimp larvae, and small percentages of fish eggs, siphonophores, barnacle larvae, polychaetes, and unidentified animal remains. He also states O. aurifrons is diurnal and “covers the entrance to its burrow for the night by backing in with a large stone in its jaws.” Leong (1967) gave a general account of the breeding and territorial behavior in the yellow- head jawish. She dealt with fish from the Florida region in aquaria and described both males and females as being territorial, even during pair formation and breeding. Paired fish were al- lowed to enter one another’s territories and bur- rows; sexual dimorphism in behavior was seen for paired fish. Both fish frequented a third bur- row and the male fish led the female to this bur- row by performing a lateral display action. Spawning occurred in the burrow and the male orally incubated the eggs. The eggs could be laid down inside the burrow to allow the male to eat. A brooding male was allowed to enter the female’s burrow, but the female was not permitted to enter the brooding male’s burrow. There are differences in color pattern of Bahamian and Floridian specimens aside from the already mentioned head markings. Bohlke and Thomas (1961) provided an excellent de- scription of life colors for Florida specimens. Bahamian specimens as illustrated in Bohlke and Chaplin (1968: 489) and in Plate I, fig. 1 are paler than individuals from Florida. The yellow found on the anterior portion of the body of Florida specimens is much less intense in those from the Bahamas and is practically non-existent in specimens maintained in aquaria for a few weeks. The various dark markings on the head of Bahama individuals, with the exception of the spots on the lower jaw, are normally hidden by various bones and folds of skin. These mark- ings are clearly exposed during various intra- specific activities and are probably important in such activities. Also present on many Bahamian individuals is what may be termed an “eyebar,” a broad faint, dark band running dorsally be- tween the eyes and ventrally onto the tip of the lower jaw. Unfortunately this band is seldom visible in preserved material although it is obvi- ous in life. These important morphologic differences be- tween Bahamian and Florida populations indi- cate that consistent differences may well exist also in behavior. Therefore this study was de- voted when feasible only to the Bahamian popu- lation. Some supplementary information was obtained for the Florida population and such is noted in the text. Since it often is displayed in home and public marine aquaria, much popular literature also exists on O. aurifrons. Some of the more note- worthy references to this literature include Ray (1968), Van Doorne (1969), and Kristensen (1965). Material and Methods Behavioral observations were made through the use of the University of Miami Rosenstiel School of Marine and Atmospheric Science video-acoustic installation at Bimini, Bahamas. The underwater television system (UTV) con- sists of a closed circuit television camera and associated hydrophones (Plate II, fig. 2) situ- ated 1.5 kilometers off the west coast of North Bimini at a depth of 20 m with cables leading to a monitor room ( Plate II, fig. 3 ) at the Lerner Marine Laboratory of the American Museum of Natural History. Details of the system can be found in Myrberg et al. (1969). Daily activity of individual fish was moni- tored for 30-minute periods every two hours between 7 AM and 7 PM at the start of the study. This schedule was modified in relation to changing day length in order to retain the same relation of periods to total day length as was present at the onset of the activity measure- ments. The occurrence of specific activities was recorded by marks made upon an Esterline Angus travelling chart recorder and their fre- quency and/or duration determined from these Colin: Behavior of the Yellowhead Jaw fish 139 records. Measurements of the environmental parameters of current speed, current direction, and water temperature were taken at the same time as the behavioral measurements for pos- sible correlation. A Hydro Products model 460 current meter, with a useful range of 0.05 to 7.0 knots and an accuracy of ± 3 percent of the reading was used with a model 45 1 current speed readout module located in the monitor room at the Lerner Marine Laboratory. The rotational speed of the current meter could also be monitored visually on the UTV as a check on the readout system. The direction of the current was determined by observing the direction of motion of particles in the water on the UTV screen. A Hydro Products model 403A temperature probe, with a measurement range of 0° to 40° C and an accuracy of ±0.5° C was used with a model 401 readout module located in the moni- tor room. Identical temperature readings were obtained with those readings taken with a mer- cury bulb thermometer at the UTV site. Behavioral observations were made during both dawn and dusk periods. The videcon cam- era used in the UTV made nocturnal observa- tions impossible without the use of artificial lights. Late in the study an effort was made to de- termine vertical and horizontal distances trav- elled by the fish from the burrow by placing a grid of small markers made from 2.5 inch long carriage bolts at specified distances from the burrow. Vertical heights above the burrow were meas- ured by markers placed certain distances behind the burrow opening on the line of sight of the camera to the burrow (Text-fig. 1 ) . By knowing the distance from the camera to the burrow, from the burrow to the marker, and the viewing height of the camera off the bottom, the vertical height of the fish could be determined by the use of similar triangles as long as the fish was directly above the burrow. The position of the fish in relation to the burrow was fairly easily determined by the relative size of the fish on the screen and by its image sharpness in the limited depth of field of the camera. Definition of Behavioral Actions It is necessary to define the actions of the animal which are to be measured. Ideally, the entire behavioral repertoire of an organism should be described. For the purposes of this paper only those actions concerned in the daily activity measurements will be defined. No or- ganizational grouping of behavioral actions used in past studies has apparently fulfilled the needs of subsequent workers since it seems each study has required the modification of a previous or formation of a new hierarchy of actions. This was also true with the behavior of Opistognathus aurifrons. The behavior of the animal was di- Text-fig. 1. The method used for the determination of the height (H) of a jawfish above its burrow at the UTV site. When the fish is in line with the distance marker (C), its height above its burrow in meters B in meters equals . A -f B in meters 140 New York Zoological Society: Zoologica 57 (4) vided into six general groups; 1) locomotory movements, 2) burrow oriented, 3) feeding oriented, 4) maintenance activity, 5) inter- specific oriented, and 6) intraspecific oriented. Locomotory Movements Hover. The fish maintains itself in position above the substrate in an anterior-posterior ver- tical orientation by an alternate beating of the pectoral fins and a beating of the caudal fin, the posterior one-half of the body, the posterior part of the dorsal fin, and the anal fin. This is shown in Text-fig. 2a and 2b in lateral and ven- tral views. The dorsal, anal, and caudal fins are moderately spread while the pelvic fins may either be held tightly against the body (Text- fig. 2c) or held out at a right angle from the body (Text-fig. 2d). The movements of the pectoral and caudal fins are co-ordinated so that the thrust pro- duced by the motion of the caudal fin and the posterior portion of the body counteracts the tendency by the stroke of one of the pectoral fins to move the upper torso laterally. In this manner the cephalic portion of the body is maintained in a stable position. The pelvic fins are held out from the body most often during periods of very low current speed. The extension of the pelvic fins may well aid in balancing the animal. This is the typical fin position assumed by fish in aquaria since currents there are practically non-existent. At current speeds greater than 0.03 knot the pelvic fins are usually held posteriorly against the body. The density of seven live Florida specimens was determined utilizing a beam balance for weight determination and a graduate cylinder for volume measurement. A mean value of 1.04 g per cm3 ^^s obtained which makes O. aiirifrons slightly denser than sea water (1.02 to 1.03 g per cm^). Fish in aquaria under con- ditions of zero current speed were observed to stroke each pectoral fin at a rate of 80 to 95 strokes per minute in order to maintain a sta- tionary position in the water column. The ani- mal produces forward (upward) thrust with movements of the pectoral fin as shown in Text-fig. 3. The dorsal and ventral rays of the pectoral fin are brought forward while the medial rays of the fin lag (Text-fig. 3a) on the upstroke. On the downstroke all of the rays are brought back together (Text-fig. 3b). To maintain its hovering position above the burrow during periods of high current speed, the fish must swim at an angle to the vertical with its head oriented into the current. The angle of the anterior-posterior axis of the body with the vertical increases with the intensity of the current. At zero current speed this angle is zero degrees, at 0.15 knot the angle is 45 de- grees, at 0.20 knot the angle is 60 degrees, and at 0.25 knot the angle is 75 degrees. Currents over 0.30 knot in speed require O. aiirifrons to swim practically horizontally in the water col- umn. The beating rate of the pectoral and caudal fins is consequently increased during periods of increased current speed. Movement Forward and Rearward. Forward movement by O. aiirifrons is accomplished by simply increasing the beating rate of the caudal and pectoral fins. Rearward movements, how- ever, can be carried out in several ways. The fish can move downward (usually rearward) by decreasing the fin beating rate. The fish can also move rearward due to its density being greater than seawater and reducing resistance to rear- ward movement by folding the pectoral fins for- ward. Finally it is possible for the animal to propel itself rearward by beating the pectoral fins in a reverse manner from that used in front- ward movement. Turn (Maneuvering) in the Hover (roll, pitch, yaw). Lateral turning (yaw) from the hover is accomplished by lessening the thrust or missing completely one or more strokes by one pectoral fin and bending the body laterally (Text-fig. 4). The thrust differential produced causes the body to turn laterally toward the side of lessened thrust. Roll and pitch are accomplished in different manners. On the alternating downstrokes of the pectoral fins, either the dorsal or ventral rays can be brought back first. If this is done, the thrust produced is not on a line with the body axis. If the ventral portions of both pectoral fins are brought back first, a pitch in the ventral direction is accomplished. If the dorsal portions are brought back first, the pitch is in the dorsal direction. If the dorsal portion of one pectoral fin and the ventral portion of the other are brought back first, a roll is produced. What is henceforth referred to as a “turn” is a turning of the animal in the water column which can involve any or all of the roll, pitch, and yaw movements described. Burrow Oriented Actions Burrow oriented actions fall into two general categories: 1 ) those concerned with the burrow as a refuge from predators and a nocturnal rest- ing place, and 2) those actions concerned with maintenance of the burrow. Tailfirst Entry. The fish enters the burrow caudal-end first while backing slowly. As out- lined previously, the fish may passively retreat due to gravity and currents or actively swim rearwards. The pelvic fins are folded against the body during this and all other entries in order to clear the burrow opening. Colin: Behavior of the Yellowhead Jawfish 141 Text-fig. 2. The action of “hover”: A) lateral view; B) ventral view; C) “hover” with the pelvic fins held against the body; D ) “hover” with the pelvic fins extended. 142 New York Zoological Society: Zoologica 57 (4) A Text-fig. 3. Movements of the pectoral fins during hovering. A) The upstroke of the pectoral fin with the medial rays lagging. B) The downstroke of the pectoral fin with all rays brought back together. Colin: Behavior of the Yellowhead Jaw fish 143 Text-fig. 4. The lateral turn (yaw). Headfirst Entry. The animal turns from a normal hovering position in the water column and swims rapidly head foremost into the burrow. Tailfirst to Headfirst Entry. The fish backs toward the burrow tailend first until the caudal fin is immediately above the opening. The jaw- fish then turns 180° and enters the burrow head foremost. Exit. The fish emerges head first, using the pectoral fins for propulsion. The pelvic fins are held posteriorly against the body until the ani- mal is clear of the mouth of the burrow. The pelvic fins may then be extended. Sit. The animal maintains position in the bur- row opening with only the head (to the level of the opercular margin) exposed. Methods used for holding this position include sitting on the sides of sloping burrow entrances and wedging the body in the opening. Cover Burrow. The jawsh enters the burrow tailfirst and as it descends will bend laterally or ventrally and pick up a small stone or shell in its jaws. This stone is released covering the open- ing as the head of the fish passes into the bur- row. This action is usually performed at dusk or when predators approach. Adjust Cover. The stone or rock covering the burrow may be adjusted in its position by the fish pushing up from inside the burrow and moving the stone with its head. Uncover Burrow. The burrow is uncovered by the fish by moving forward in the burrow and pushing the covering stone out of the way with its head. The stone may then be moved by picking it up in the jaws if it is still blocking the opening. The action is usually performed at dawn or after the passing of predators. Resting. That position within the burrow, whereby the fish rests on the substrate. All fins are folded, with the possible exception of the pectoral fins. This position is noted only at night. There is no apparent opening or closing of the mouth or the opercular apparatus. Actions Related to the Maintenance of the Burrow Dig. This action can be divided into three sub-actions which are not discreet motor pat- terns, but appear to have different functions. A. Within burrow: Sand is scooped up in the mouth inside the burrow and it is then deposited outside the burrow near its margin (Plate III, fig. 4). When carrying sand, the branchial ap- paratus of the fish is expanded, the mouth closed, and the gill covers slightly open (Plate III, fig. 5). The dark line bordering the isthmus which is normally hidden is visible when the mouth is full of sand. During periods of high current speed, the jawfish will use the current to its advantage in digging from the burrow. Rather than expelling the sand on or beyond the margin of the bur- row, the fish will expell the sand vertically from its mouth without emerging fully from the bur- row opening. The current will carry the sand over the burrow margin before it can fall. 144 New York Zoological Society: Zoologica 57 (4) B. Retrieve sand: Sand is scooped into the mouth at some distance from the burrow ( Plate IV, fig. 6) and brought directly to the margin of the burrow where it is expelled. Again the dark line bordering the isthmus is visible when sand is being carried in the mouth. Movement of the caudal and pectoral fins is extremely rapid when swimming with the sand in order for the animal to stay above the bottom with this added weight. C. Remove sand: Sand is scooped up from the margin of the burrow and carried some dis- tance from the burrow where it is expelled. Again, the isthmus line is visible and the swim- ming rate rapid. Retrieve Rock. This action, like “dig,” is di- visible into three sub-units. A. Recover rock: A small stone or shell is picked up in the jaws and brought to the bur- row where it is deposited on the margin. Carry- ing rocks differs from carrying sand in that the rock is held in the jaws while sand is carried inside the mouth. Also, sand must be force- fully expelled while the rock can be released by simply opening the jaws. B. Remove rock: A small stone or rock is picked up in the jaws from the burrow margin and carried some distance away where it is deposited. C. Remove rock from within burrow: The fish picks up a rock in its jaws inside the bur- row, emerges headfirst, and drops the rock on the burrow margin (Plate III, fig. 7). Adjust Rock. This action is performed with the body in the burrow with only the head ex- posed. Rocks on, or near, the burrow margin are picked up in the jaws and positioned on the margin. Often upon placing the rock in position, the jaws of the fish are not released and the rock is moved with the head to produce a more suit- able resting place for it. Rocks may also not be removed from their original position but simply moved slightly to improve their positioning and that of surrounding rocks. Actions Associated with Feeding Thrust. The fish moves rapidly forward from a hovering position through the use of the pec- toral and caudal fins. The animal comes to a quick stop through the use of the pectoral fins (Text-fig. 5). Snap. The jawfish ingests a food particle by opening the mouth and creating a slight inrush of water by flaring out the opercular covers and spreading the branchiostegals (Text-fig. 6a and 6b). The line hidden under the maxillary is ex- posed when the mouth is opened, and the long axis of the pupil is oriented so the fish may see the food particle with binocular vision. Text-fig. 5. The action of “thrust.” Colin: Behavior of the Y ellowhead Jawfish 145 Reject. The food particle is ingested as in a snap, but is quickly expelled from the mouth by a pulling in of the gill covers and branchial apparatus. Maintenance Activity Although the jawfish possesses several appar- ent maintenance activities, these will not be de- scribed since they were not quantified diurnally. Actions Concerned with Interspecific Relationships The interspecific relationships of O. aurifrons have previously been dealt with by Colin (1971). The action, chase, was the only activity for which diurnal data are available. Chase. The act of chasing an intruding fish with jaws spread. It usually occurs within 20 cm of the burrow and the distance a fish is pur- sued varies a great deal. Swimming is carried out rapidly with the pectoral and caudal fins. Actions Concerned with Intraspecific Relationships A variety of intraspecific relationships exist among individuals of O. aurifrons. Some of these have already been mentioned by Colin (1971), while others are described below: Arch. This is the “lateral display” of Leong ( 1967). I feel the term “lateral display” is more suited for a different action involved in terri- torial defense where one fish turns laterally toward another, with the body oriented nearly vertically and spreads the mouth, branchial apparatus, and isthmus to their maximum while the head is shaken laterally. Leong (1967) did not describe such an action for O. aurifrons. The “arch” is performed by the male fish. This fish, swimming in the water column, will orient its body laterally toward a female and assume a horizontal position. The caudal and cephalic ends of the body are bent upward and the dorsal, anal, and caudal fins are spread to their maximum. Immediately after the body is bent, the branchial apparatus and the isthmus are dropped and the mouth is opened ( Plate V, fig. 8 ) . The spread of the mouth is not as great, however, as it is during the aggressive “lateral display.” The isthmus, maxillary, and branchio- stegal lines are all clearly displayed, and the arch position may be held for several seconds. Often after this action, both fish will enter one burrow for a period of several seconds. It is for this reason that the “arch” is considered as courtship behavior rather than some other type of intraspecific behavior. Brood. The male fish broods the eggs orally (Text-fig. 7) and is positioned usually directly above the burrow entrance in a hovering atti- tude. The maxillary line is clearly exposed and the branchial apparatus not greatly expanded. Text-fig. 6. The action of “snap”; A) dorso-lateral view, B) ventral view. 146 New York Zoological Society: Zodlogica 57 (4) Text-fig. 7. The action of “brood.” The mouth is usually open but it can be nearly closed while carrying the eggs. Closing the mouth when the egg mass is being carried causes a consequent expansion of the branchial appara- tus. The head is expanded somewhat laterally and the isthmus lines are apparent in a front view of the animal. Daily Activity Patterns and Behavioral Correlations In situ studies of coral reef fishes have been carried out only recently, and work dealing with in situ measurements of daily activity of coral fishes is extremely scarce. Some of the note- worthy studies are the following. Youngbluth (1968) worked with the Hawaiian, parasite- picking wrasse, Labroides pthirophagus, and de- termined feeding (cleaning) rates of these fish at two hour intervals during the day. No sig- nificant difference was found between morning and afternoon rates but feeding rates varied on different reefs. Albrecht (1969) studied the fanning of the nest by the pomacentrid, Abu- defduf saxatilis, in relation to depth and time. His observations were both diurnal and noc- turnal. Myrberg (in press) worked with the daily patterning of various sonic patterns in the pomacentrid, Pomacentrus partitus, and in- cluded observations both in the field and in labo- ratory aquaria. During the present study, two individuals of O. aurifrons were sufficiently close to the UTV camera so that detailed observations and rather precise measurements of behavior could be carried out (2m). Four other individuals were approximately 8m away and although their posi- tion in the water column could be observed, lack of detail precluded behavioral measure- ments. The jawfish, male and female, had their burrows 60 cm apart. This pair was closely observed for 25 days during the summer of 1969 and the frequency of certain actions was recorded for 30 minutes during each of seven periods throughout the day. Fifteen minutes of each period was de- Colin: Behavior of the Y ellowhead Jawfish 147 voted to the activities of each of the subjects. The onset of the periods was altered so that they maintained the same relative position in regards to the changing length of the daylight period. For example, the periods originally at 9; 00 AM on June 24 was moved to 9; 18 AM on Sept. 7 to keep the same position in total day length. The days spent observing and recording the behavior of O. aiirifrons on the UTV in- cluded June 24 to 30, July 17 to 24, August 2 to 6, August 26 to Sept. 3, and Sept. 5 to 8, 1969. Preliminary observations but no quanti- fication of behavior were carried out on June 4 to 6. For purposes of recording various activities on a 12-channel event recorder, activity was broken down into four major groups; 1) feed- ing, 2) burrow oriented, 3) interspecific, and 4) intraspecific. Whenever possible an “indica- tor action” was selected to reflect the level of a certain type of activity. This “indicator action” is often not the most direct measure of a major activity group. Sevenster (1961: 17-18) for ex- ample, correlated the number of “zigzags” of the males of Gasterosteus aculeatus with the fre- quency of the male leading the female, a purely sexual activity. He then used the more easily observed “zigzag” as a measure of the sexual activity of the male instead of the less easily observed action of leading. For O. aiirifrons the “snap” was selected as an indicator action for feeding activity since it was easily observable and reflected reasonably well the actual food intake of the animal. The actions of “thrust” and “turn” were also con- sidered possible feeding actions and their fre- quencies, along with those of “snap,” were measured during the period June 24 to 30. Regression lines were calculated from these data (Text-figs. 8 and 9) and clear correlations were noted among the occurrences of these actions. Therefore, each could be considered as elements of feeding behavior and it was neces- sary to determine only the frequency of “snaps” in succeeding behavioral measurements. Measurements were also made of the amount of time spent by a given fish in the water column above its burrow. The “Percent time in the water column” was subsequently determined, with those activities directed away from the water column, such as retrieval of sand, not being included in this percentage. The seven observation periods during the day were numbered chronologically for reasons of analysis. Text-fig. 10 illustrates feeding activity, as reflected by the mean number of “snaps” per 15 minute period for each of the seven periods of the day. The feeding rate (“snaps”) is fairly constant over the entire day considering all environmental conditions. Reaction to specific environmental factors such as current speed and light intensity will be examined later. The relationships that exist among the actions involved in burrow oriented activity are more complex than those involved in feeding. Digging from within the burrow is commonly seen, but it cannot be considered to reflect the nature of the fluctuations shown by related activities such as: remove sand, retrieve rock, remove rock, and adjust rock (Text-fig. 11 and Table 1). These, therefore, must each be considered sepa- rately. Text-fig. 1 1, illustrating mean digs per 15 minutes, shows a strong peak ( 10.4) at the fifth period (3PM — start of study) subsequent to a moderate, but fairly consistent digging rate dur- ing the first four periods (4.6 to 7.3). There is a sizable decrease in the rate after the fifth period with a rate of only 1.6 at the seventh period (7PM — start of study). Table 1 presents the mean values per 15 min- utes of three other burrow oriented activities; re- moval of sand, retrieval of sand, and adjustment of rocks. The latter two show marked increases in the final three periods of the day with their greatest values being in the last period. Removal of sand differs since its peak value is during the first period of the day with only a slight increase in the afternoon. The frequency of the activity, chase, is shown in Text-fig. 12. The increased frequency of this activity in the last three periods of the day was probably due to an increase in the swarming be- havior of various labrids at that time. Hali- choeres bivittatus and H. garnoti were very ac- tive during the final period of the day, often causing a jawfish to prematurely close its burrow for the night. The frequency of flight reactions of O. auri- frons from other species of fishes was difficult to determine objectively. Often an approaching fish could not be seen on the UTV but the jaw- fish would flee to the burrow. Conversely, the fish has been observed to move rapidly to the burrow during the approach of a large fish, then emerge several seconds later with a mouth full of sand, as noted during a “dig.” Such occur- rences in the face of two equally possible stim- ulus situations preclude the use of possible flight responses in measurements of daily activity. Other aspects of interspecific activity have been considered elsewhere (Colin, 1971). The occurrence of the probable male court- ship pattern “arch,” is shown in Text-fig. 13. Percent of total “arches” is plotted rather than a mean value since the sample size was rather small (35). Arching was most prevalent early in the morning and during late afternoon pe- riods. Aquarium observations supported this finding, most arches occurring shortly after the burrow had been uncovered in the morning. Thrusts per 15 min 148 New York Zoological Society: Zoologica 57 (4) Text-fig. 8. Correlation of the frequency of the actions of “snap” and “thrust.' Colin: Behavior of the Y ellowhead Jawfish 149 Text-fig. 9. Correlation of the frequency of the actions of “snap” and “turn.” Mean Snaps per 15 min. 150 New York Zoological Society: Zoologica 57 (4) OBSERVATION PERIOD Text-fig. 10. Diurnal patterning of the mean “snap” frequency of two specimens of Opistognathiis aurifrons for a period of 25 days at the UTV site, Bimini, Bahamas. Colin: Behavior of the Y ellowhead Jawfish 151 Z rt> a> D O rt en 3 -D 12 10 8 6 4 2 0 Text-fig. 11. Diurnal patterning of the mean “dig” frequency for two specimens of Opistognathus anri- frons for a period of 25 days at the UTV site in Bimini, Bahamas. Table I. Mean Frequency of Burrow Oriented Actions per 15 Minutes for a Period of 25 Days Period Retrieve Sand Remove Sand Adjust 1 0.21 1.68 0.31 2 0.02 0.08 0.18 3 0.59 0.02 0.25 4 0.26 0.00 0.13 5 1.10 0.51 1.48 6 3.83 0.25 1.05 7 4.24 0.05 1.81 Total Number of Periods Observed 231 ssH PERCENT OF ARCHES 152 New York Zoological Society: Zoologica 57 (4) Text-fig. 12. Diurnal patterning of the mean frequency of “chase” for two specimens of Opistognathiis aiirifrons for a period of 25 days at the UTV site in Bimini, Bahamas. 1 2 3 4 5 6 OBSERVATION PERIOD 7 Text-fig. 13. Diurnal patterning of “arch” (given as percent of occurrence) for a period of 25 days at the UTV site, Bimini, Bahamas. ssH Colin: Behavior of the Yellowhead Jawfish 153 Relationship Between Feeding and Burrow Oriented Activities The two major time-consuming activities of the field subjects were feeding and burrow oriented behavior. These two activities were in- versely related since the first involved being in the water column and the second did not. Text- fig. 14 shows the mean number of “snaps” and “digs” against the percent time in the water col- umn. As the time in the water column increases, the burrow oriented action (“dig”) decreased and the water column oriented action (“snap”) increased. However, the number of “snaps” or “digs” was not directly proportional to the per- cent time in the water column. The “snap” value at 50 percent time in the water column was only one-quarter of the value at 100 percent time in the water column, not one-half as would be expected if the feeding rate was constant over the entire period spent in the water column. Percentage values below 33 percent time in the water column were not observed in the field except when the fish remained in its burrow for known or unknown reasons for the entire 15 minutes. This is also discounting a few “aborted” periods where the jawfish was frightened into its burrow after a few minutes of normal be- havior and remained there for the remainder of the observation period. The explanation for the seeming paradox rests with both the behavior of the fish while dig- ging and the length of the observation period (15 min.). Brief periods of hovering always in- terrupted separate bouts of digging, and such periods accounted for at least 33 percent of a given observation period. There is little diurnal variation in the mean percent time spent in the water volumn as is shown in Table 2. Mean percentages from a low Table 2. Diurnal Variation in Mean Percent Time in the Water Column Observation Period Mean Percent Time in the Water Column Number of Periods 1 91.0 25 2 87.2 28 3 90.8 25 4 85.6 27 5 84.7 18 6 83.8 24 7 90.7 16 PERCENT TIME IN WATER COLUMN N = 1 2 6 11 14 28 98 TOTAL N : 160 NUMBER OF OCCURRENCES Text-fig. 14. Relationship of mean “dig” and mean “snap” frequency to percent time in the water column. 154 New York Zoological Society: Zoologica 57 (4) of 83.8 to a high of 91.0 indicate that water column oriented behavior (feeding) dominated the total daily activity. This does not, however, reflect feeding effectiveness (“snap” rate) which can be altered by environmental conditions. Effect of Current Speed Upon Various Activities Current speed affects the “snap” frequency in combination with certain other conditions. Text- fig. 15 shows the mean “snap” frequency during observation periods when the current speed was greater than 0.20 knot. There is a sizable de- crease in the mean “snap” frequency during the low light periods (periods 1, 2, and 7) in which high current speed was encountered. The "snap” frequency for periods where the current was 0.20 knot or less is shown in Text- fig. 16. In this case the “snap” rate was nearly constant throughout the day with only a slight dip at the third period. The values at low light periods during high current speed shown in Text-fig. 15 are therefore being masked in Text-fig. 10 by the greater oc- currence (85 percent of all observations) of currents 0.20 knot or less. During the winter, however, shorter days and generally less con- sistent water conditions will no doubt increase periods of low ambient light and of high current speed. It seems then that neither high current speed (0.20 knot or greater) nor low light conditions alone could produce any significant decrease in the frequency of “snaps.” In fact, the mean “snap” frequency at periods 3 and 4 is greater for currents above 0.20 knot (Text-fig. 15) than for currents of 0.20 knot or below (Text-fig. 16). During high current, low light conditions, the fish does not move far from the burrow and instead hovers with the anterior posterior axis of the body near horizontal directly above the burrow opening. I feel that O. aurifrons, being mainly a visual feeder, probably requires a cer- tain level of both light and current speed to feed effectively. Variation in only one parameter does not apparently affect the feeding rate. Digging is also somewhat related to the cur- rent speed as shown in Text-fig. 17. The mean values of “digs” in periods in which digging oc- curred was similar for currents of 0 to 0.10 knot and for currents greater than 0.10 to 0.20 knot. The mean “dig” frequency showed a significant decline (at least 95 percent separation for cur- rents greater than 0.10 to 0.20 knot and greater OBSERVATION PERIOD Text-fig. 15. Diurnal patterning of mean “snap” frequency of Opistognathiis aurifrons when current speed was greater than 0.20 knot at the UTV site in Bimini, Bahamas. Colin: Behavior of the Y ellowhead Jawfish 155 than 0.20 knot), however, for currents greater than 0.20 knot. The reasons for this decline are still unclear. The action of arching also seems to be related to the current. Text-fig. 18 plots the percentage of “arches” observed against various current conditions. Almost three-quarters of the “arches” occurred when current speed was 0 to 0.05 knot, while these currents occurred in only one- quarter of the observation periods. Current speeds of 0 to 0.10 knot were observed in 52 percent of the periods, yet over 90 percent of the arches observed occurred during this current speed regime. Additionally no arching was ob- served at current speeds over 0.20 knot. Since the arch is a complex posture supposedly di- rected at a female, the utility of performing this action under current conditions where the dis- playing fish is not quickly carried away by the current is obvious. The percent time spent by the fish in the water column possibly varied with the current and this might easily explain the results shown in Text- fig. 15 (currents speed greater than 0.20 know versus “snap” frequency) and in Text-fig. 18 (“arches” versus current speed). But Table 3 shows that the mean percent time in the water column was independent of the current speed. The parameter of current direction was re- corded with current speed data but it showed no correlation with any behavioral measurements. The only change seen in the fish was a change in the direction in which they faced in order to swim into the current. During periods of slight or no current speed, the swimming direction of the fish seemed random except when an appar- ent food item was sighted. Effect of Other Environmental Factors Upon Various Activities Temperature measurements were also made at the time behavioral data was taken, but sta- bility of temperature during the study (29° to 31° C) precluded meaningful correlations with behavior. Temperatures near the winter low of approximately 20° C might produce consider- ably different results. Yellowhead jawfish, kept in aquaria, become very inactive at tempera- tures approaching 20° C and feed very sparingly. Below 20° C the animals spend most of their time in the burrow, and at 17° C appear near their lethal lower temperature limit. Atmospheric conditions at the UTV site were also considered, qualitatively, as possibly influ- encing behavior. Numerous thunderstorms and (/) D Of •o (/) "O -f cn 3 p 1 2 3 4 5 6 7 OBSERVATION PERIOD Text-fig. 16. Diurnal patterning of mean “snap” frequency of Opistognathus awifrons when current speed was 0.20 knot or less at the UTV site in Bimini, Bahamas. 156 New York Zoological Society: Zoologica 57 (4) heavy rain, which could be heard at the site via the submerged hydrophone, seemed to have no effect on the activity of the fish. Surging of cur- rents on the bottom (depth 20 m) produced by surface waves resulted in movements of grass blades and bits of detritus. These occasionally rolled along the bottom and entered burrow openings. Such objects were quickly removed by the fish from the burrow. Turbidity measurements were not made, but increased turbidity no doubt affects the activities of jawfish since any decrease of ambient light reduced visibility. Such reduction might well cut the feeding effectiveness of the animal as well as reducing its range of movement. Sonic Activity Extended listening has been unproductive in detecting any sonic activity by O. aiirifrons. A hydrophone was positioned less than one-half meter from the fish at the UTV site for the dura- tion of the study, and several hours were spent listening with hydrophones in laboratory aquaria at various times of the day, but the results from this monitoring have been negative. Time of Covering and Uncovering THE Burrow and Nocturnal Behavior Opistognathus aiirifrons covers its burrow opening in the evening with a small rock or shell and remains within the confines of the burrow until the morning when the rock is removed. Table 4 presents the results of numerous obser- vations on the time and conditions of the open- ing and closing of the burrow. The time at which the burrow is uncovered during the morning varies from ten to 12 min- utes before to a very few minutes after sunrise. Mornings with clear skies tended to have early uncoverings and mornings with late uncoverings were usually overcast. The fish tended to rise two or three minutes after the first objects on the bottom were visible on the UTV. This time of first visibility, of course, varied with the sky conditions. Aquarium fish will uncover the bur- row any time during the night if lights are turned on, but the time required for this to happen was generally longer (as much as 10 to 15 minutes) during late night and early morning hours. When the lights were turned on at the normal uncover- ing time, the fish usually removed the cover of the burrow in less than one minute. CURRENT SPEED Text-fig. 17. Relationship of current speed to the mean number of “digs” by Opistognathus aiirifrons during periods in which digging occurred. Colin: Behavior of the Yellowhead Jawfsh 157 0-.05 >05-10 >10 -.15 >15-20 >20 CURRENT SPEED (Knots) Text-fig. 18. The relationship of percent of “arch” and percent of occurrence of various current speed regimes to the current speed. Table 3. Variation in Mean Percent Time in the Water Column with Current Speed Current Speed Mean Percent Time in the Water Column Number of Periods 0.00-0.10 Knot 88.9 82 greater than 0.10-0.20 K. 85.9 60 greater than 0.20 K. 90.5 22 Light is probably a major controlling factor in determining the uncovering time. As day length changed, the time of uncovering the bur- row was altered to match the changing sunrise time. On mornings when ambient light was low due to atmospheric conditions, this was reflected by a later rising time. A different situation existed with respect to the time of covering the burrow in the evening. Times of from 92 minutes before sunset to six minutes after sunset have been recorded. How- ever, many factors apparently entered into the determination of the closing time. The presence of other species in the area seems to have a defi- nite effect. Large numbers of the fishes Halicho- eres bivittatus, H. garnoti, and Pseitdupeneiis maculatus, browsing on the substrate near ter- ritories of O. aurifrons, often coincided with an immediate retreat to the burrow and a covering of the mouth of the burrow with a stone. If this occurred near dusk, the fish often remained in the hole for the night. Light appeared to control the absolute limits of time for closing the burrow. One-third of all closings occurred between sunset and six min- utes thereafter (maximum limit). A series of night dives on colonies of O. auri- frons by the author on the Florida reef tract in 1969 showed no evidence of any nocturnal ac- tivity. This agrees with Starck’s (1966) state- ment “at night it (O. aurifrons) has never been found, and is apparently inactive.” Fish in vari- ous aquaria (40 to over 2000 liters) also re- mained in their burrows the entire night as evi- denced by irregular frequent inspections. 158 New York Zoological Society: Zoologica 57 (4) Table 4. Time of Uncovering and Covering the Burrow by Opistognathiis aurifrons Uncovering in the morning Time in minutes before ( + ) or after ( — ) Number of Percent of sunrise occurrences occurrences greater than +10 2 13.3% + 10 to 0 10 67.0% -1 to -10 2 13.3% greater than —10 1 6.7% Covering in the evening Time in minutes before ( + ) or after (— ) Number of Percent of sunset occurrences occurrences greater than +20 9 33% +20 to +10 1 4% + 10 to 0 8 29% -1 to -10 9 33% Morning and Evening Behavior The present section deals with those events which immediately followed opening and pre- ceded closing the burrow. The activities occur- ring during these two periods of the day are ex- tremely different. Text-fig. 19 shows that after opening the bur- row in the morning, the fish quickly entered the water column with a resultant decrease in the time spent in the burrow. The time spent within the burrow opening (neither completely out or in the burrow) reaches a peak five minutes after opening the burrow, but it never occupies a sig- nificant percentage of the fish’s time. Feeding began as soon as the fish entered the water col- umn; after only four minutes the frequency of “snaps’’ was practically equal to the mean daily frequency (see Text-fig. 10). Burrow oriented activities such as digging and retrieving were non-existent in the early morning period, and interspecific activity was rarely observed. Actions which immediately preceded closing the burrow in the evening were considerably different than those following its initial opening. A striking increase in both “adjust rock’’ and “retrieve sand" (Text-fig. 20) demonstrated that individuals physically prepare their burrow for the night. The adjustment of rocks may serve to prepare the opening for its covering stone, and the retrieval of sand to hide the stones of the burrow rim or to provide for a better fit for the covering stone. One covering stone is not re- served for use day after day, but a suitable stone, often one-half of a bivalve mollusc shell, is se- lected shortly before dusk and placed near the burrow opening. Feeding declined in the final minutes (Text- fig. 20), but a low level remained up to the moment that the burrow was closed for the night. Range and Territory The concepts of range and territory are sepa- rate entities. The word, range, implies the total area into which an individual confines its pres- ence. Territory, however, is a somewhat more elusive concept bringing to mind that area which an animal “considers” its own and is willing to defend. Often the territory of an animal depends upon the type of intruder that is encountered. The burrow of O. aurifrons may logically be considered the “center” of both its range and territory with the level (or intensity) of defense decreasing rapidly with distance from that point. The territory was defended only against those fishes that are nearly the same size or smaller than O. aurifrons. Its efforts at territorial defense are minor compared to other species of reef fish, such as members of the genus Ponmcentrus (Emery, 1968a; Myrberg, in press). Small fishes would be chased from a circle 20 to 25 cm in radius, with the burrow at its center. Fishes be- yond this distance were apparently “watched” but no aggressive actions were directed toward them. The range of the yellowhead jawfish should be considered from two aspects. The first is the feeding range which in benthic feeding fishes includes horizontal movement. Since O. auri- frons is a plankton feeder, this range also in- cluded vertical movement above the bottom. Text-fig. 21 presents the modal height and also the greatest height seen in any one period during given periods throughout the day. Heights greater than one meter could not be quantita- tively measured, since there was nothing visible behind the fish for sight reference. The greatest Colin: Behavior of the Yellowhead Jawfish 159 in the Water Column in the Burrow Opening in the Burrow Text-fig. 19. Percent of two-minute periods spent in various locations by Opistognathus aitrifrons after opening the burrow in the morning at the UTV site in Bimini, Bahamas. All values are the mean of six observations. 160 New York Zoological Society: Zoologica 57 (4) Text-fig. 20. A) The occurrence of “snap” by Opistognathus aurifrons before closing the burrow for the night. Mean of ten observations. B) The occurrence of “adjust rock” by O. aurifrons before closing the burrow for the night. Mean of ten observations. C) The occurrence of “retrieve sand” by O. aurifrons before closing the burrow for the night. Mean of ten observations. Colin: Behavior of the Yellowhead Jaw fish 161 height ever reached by the animals was estimated about one-and-one-half meters above the bottom. The low heights reached during the morning and evening hours were probably the result of low light levels, and this apparent adaptation no doubt provided greater chance for avoidance of predators. The horizontal component of the vertical ranging of the animals was at most 2 m, but seldom more than 1 m from the burrow opening. The second aspect of range to be considered is that having to do with rock and sand retrieval. The number of rocks which a fish considers suitable for its burrow in a given area must, of course, be limited. If these are used in burrow construction, the fish must then extend this range to retrieve additional ones. At the UTV site the range for rock retrieval was approxi- mately 2 m. The fish would take a zigzag course outward inspecting various stones until a suit- able one was found. The return course was di- rect, straight to the burrow. The range for retrieval of sand was consider- ably smaller (generally less than V2 m) due, no doubt, to the easy availability of suitable sand. Rock stealing (i.e., taking rocks from the bur- rows of conspecifics ) is often mentioned in the popular literature. In the field, this behavior was seldom seen, but evidently occurs more often under aquarium conditions. This is probably due to crowding and lack of sufficient rocks for proper burrow construction. In a 3000 liter labo- ratory aquarium with a bottom area of nearly one-quarter square meter for each of 15 fish, rock stealing was rare since the aquarium al- lowed a reasonably natural density of fish. Comparative Relationships and Discussion OF THE Ecology of Opistognathus aurifrons The sandy areas bordering reefs possess sev- eral characteristic species of fishes. Included in this group are the sand tilefish, Malacanthus plumieri; the gobies, loglossus calliunis and I. helenae\ the bridled goby, Coryphopterus glaucofraemim; the garden eel, Nystactichthys kalis; and Opistognathus aurifrons. All are bor- rowers of one sort or another, and all are col- ored for concealment against a white sand back- ground. Two species, the gobiid /. calliurus and the heterocongrid N. kalis, are amazingly simi- -Greatest Height Reached Modal Height Reached 1 2 3 4 5 6 7 OBSERVATION PERIOD Text-fig. 21. The diurnal patterning of greatest and modal height reached by Opistognathus aurifrons above the burrow. Heights above one meter could not be quantitatively measured and were only estimated. 162 New York Zoological Society: Zoologica 57 (4) lar to O. aurifrons in the way they “make their living.” Specimens of /. calliurus known only from Florida waters hover in the water column and probably feed on floating plankton as does O. aurifrons. Randall (1967b) reports its West Indian congener, 1. helenae, to feed entirely on floating plankton. I', calliurus enters burrows head-first on the approach of danger and has been seen by the writer performing lateral dis- plays directed toward conspecifics for unknown purposes. Pairs often reside in one burrow which has a narrow vertical tunnel. Often groups of I. calliurus were found extremely close to yel- lowhead jawfish colonies on Florida reefs, but never within their boundaries. Nystactichthys kalis, attaining a length of one-half meter, does not hover as the others, but merely extends a portion of its body out of the burrow (Bdhlke, 1957). It picks small zoo- plankters out of the water column (Randall, 1967a) while remaining partially within its burrows. The position of its eyes, as well as the shape of its pupils, is similar to O. aurifrons and probably enables it to utilize binocular vision in picking plankton. A plankton-feeding existence imposes certain restrictions on the activity of a fish. A major portion of its time must be spent in feeding, due to the small size and spacing of food particles. For example, O. aurifrons spends practically 90 percent of the daylight periods feeding. Re- cent work by Emery (1968b) on the plankton within the reef ledges and caves may modify some of these generalities, since in these locali- ties tremendous amounts of zooplankton are readily available to reef fishes. Most plankton- feeding fishes visually detect their prey. Such feeding is expedited by binocular eyesight and such activity is restricted to diurnal periods. A notable exception to this rule may be the apo- gonid fishes, which apparently locate plankton visually at night (Randall, 1967a). Most other plankton-feeding fishes are inactive at night. Some of the western Atlantic congeners of the yellowhead jawfish, Opistognathus macro- gnathus, O. maxillosus, and O. whitehursti, dif- fer greatly in behavior, food habits, and colora- tion. They 1) do not hover; 2) feed primarily on benthic forms and free swimming forms liv- ing near the bottom (Randall, 1967a); and 3) are brownish, dusky, or mottled in coloration. In addition they are often found in areas of turbid water. Nothing is known of the food habits of the congeners of O. aurifrons typically found in clear water, i.e., O. lonchurus and O. gilberti. Both species have been reported not to hover as O. aurifrons (O. gilberti, Bdhlke, 1967; O. lon- churus, Bdhlke, 1967, W. A. Starck, pers. comm.), and it seems likely they are not par- ticulate plankton-feeders. Opistognathus auri- frons may well be the only plankton-feeding member of the genus Opistognathus in the west- ern Atlantic. The dear-water group of O. aurifrons, O. lonchurus, and O. gilberti apparently do not overlap in their ecologic distribution. O. gilberti is known only from the Bahamas and some areas of the Caribbean, typically on steep slopes in water 28 to 47 m in depth (Bdhlke, 1967). O. lonchurus may be continental in distribution at depths between 38 and 93 m, and apparently prefers siltier sand conditions than O. aurifrons (Starck, pers. comm.). It is not found near the rocky outcrops associated with O. aurifrons. The only area along the Florida coastline where such substrate conditions are found in combina- tion with clear water is seaward of the deep reefs, the latter terminating at a depth of about 30 m. It may well be that the distribution of O. lonchurus is determined by substrate and water conditions, not by depth. A case in point is seen at Triumph Reef, Florida; substrate con- ditions similar to those on the shallow reef are found to a depth of 50 m, and in this area O. aurifrons is abundant and O. lonchurus absent. Opistognathus macrognathus, a species found in Florida in shallow water, has also been taken occasionally at Triumph and Long Reefs to a depth of 45 m. Several specimens have been taken near to O. aurifrons colonies; in one case a specimen was collected from a burrow only 60 cm distant from a yellowhead jawfish burrow. Few reef fishes live in colonies. One clear exception to this is the yellowhead jawfish. Mem- bers of this species are reluctant to stray more than a few meters from their burrow, and this requires that the fish live close together for pur- poses of reproduction. Competition for food, which might be critical in benthic feeding fishes, is eliminated by the constant influx of plankton with the movement of water immediately above the burrowk Little is known of the larval life of O. auri- frons. Specimens of O. macrognathus reared in aquaria metamorphosed from free swimming larvae to burrow dwelling juveniles 18 days after hatching. It seems likely that O. aurifrons has a similar larval development. How long indi- viduals can remain as planktonic larvae will, of course, limit the distributional abilities through currents of this species. Spawning extends at least from spring through late autumn. Fishes brooding eggs have been observed at the following locations on the dates given; Triumph Reef, Florida, May 27; Bimini, Bahamas, May 25, June 2; Serranilla Bank, Colin: Behavior of the Y ellowhead Jawfish 163 Oct. 5. Whether a female will spawn more than once per season is unknown, but it seems likely since multiple spawning of O. macrognathus about two weeks apart have occurred in aquaria. The short larval life of jawfishes is advan- tageous since this period is the time of greatest predation. Unhatched fish are protected by the brooding parent and the burrowing, mature and immature fish have effective means of avoiding predation. This type of larval life, however, re- duces genetic interchange over long distances, and coupled with the reduction in genetic ex- change due to the non-wandering habits of the adults may result in the variability of this spe- cies over its geographic range. The occurrence of one “Bahama type” specimen in a group of Florida Keys specimens (Bohlke, 1967) may indicafe thaf larvae may rarely get across open water barriers, such as the Florida Current, or that such variants occasionally are produced. Whether this occurs with sufficient regularity to keep the populations from moving farther apart is not known. The significance of the dark head markings apparently used in behavioral displays and their absence in Florida specimens is not understood. The behavioral actions of Florida specimens are similar, if not identical, with those produced by Bahama specimens. Aquaria, containing mem- bers of both populations, might provide inter- esting insight into differences between these populations, e.g., whether the members of the populations are reproductively isolated for phys- ical or behavioral reasons (presently unknown). Various morphological adaptations of O. au- rifrons, not present in the less specialized mem- bers of Opistognathus, are interesting when viewed from a behavioral-ecologic standpoint. The yellowhead jawfish possesses recurved canine teeth on the lower jaw; yet these are not needed in the capture of food which consists of planktonic animals. It seems likely that these teeth are an aid in carrying large rocks which would otherwise easily slip out of the jaws. This therefore appears to be advantageous for life in a rubble-strewn, calcareous sand area. Similarly in this species, the large mouth is needed not for feeding but appears to be essential only for dig- ging and brooding the young. The coloration of O. aurifrons, unusual among jawfishes, is again an apparent adapta- tion to the environment. The fish blends well against a white sand background. Behavior can be thought of as a product of the environment. As previously discussed, the plankton-feeding existence imposes certain re- quirements on behavior and the burrowing ex- istence also imposes its own set of restrictions. In combination, these restrictions require that the fish have as the spatial center of its activity the burrow, yet spend the major portion of the daylight period in the water column feeding out of contact with the burrow. This results in the retreat behavior observed and accounts for the great wariness of this fish. It also makes colonial existence advantageous with the resultant effects on intraspecific behavior and genetic inter- change. An organism reflects the requirements of its environment through appropriate behav- ior; and this, in turn, is seen most readily through the various adaptations that exist for a given ecological niche. The yellowhead jawfish is a most instructive creature for demonstrating this reflection and the apparent success of its adaptations for its “chosen” niche. Acknowledgments Dr. Arthur A. Myrberg, Jr., made the full facilities of the Bimini video-acoustic installa- tion available and his help and advice were in- valuable. Dr. C. Richard Robins, Dr. Lowell P. Thomas, and Dr. Leonard J. Greenfield also gave advice and contributed to the study. Dr. Robins kindly loaned photographs and motion pictures of O. aurifrons taken by Walter R. Courtenay and himself at the Miami Sea- quarium. Stanley Walewski helped greatly in diving and underwater television operations at Bimini, and Clifford Shoemaker aided the Miami field op- erations. Robert F. Mathewson and the Ameri- can Museum of Natural History extended the facilities of the Lerner Marine Laboratory. Finally I would like to thank my fellow and former graduate students: Arnold Banner, Ste- ven Berkley, Nicholas Chitty, Martin Gomon, Samuel Ha, Robert Livingston, Glenn Ulrich, and Diana Welty for their help. This work was supported by Office of Naval Research contract N000-14-67-A-0201-0004, Arthur A. Myrberg, Jr., principal investigator, and ONR grant num- ber 552(07) to the Lerner Marine Laboratory. Contribution No. 1619 from the Rosenstiel School of Marine and Atmospheric Science, University of Miami. Literature Cited Albrecht, Helmut 1969. Behavior of four species of Atlantic dam- selfishes from Colombia, South America (Abudefduf saxatilis, A. taiiriis, Chromis midtilineata, and C. cyanea\ Pisces, Po- macentridae). Z. Tierpsych. 26: 662-676. Bohlke, James E. 1957. On the occurrence of garden eels in the western Atlantic, with a synopsis of the heterocongrinae. Proc. Acad. Nat. Sci. Phila. 109: 59-79. 164 New York Zoological Society: Zoologica 57 (4) 1967. A new sexually dimorphic jawfish {Opis- thognathus, Opisthognathidae) from the Bahamas: Notulae Naturae, No. 407: 1-12. Bohlke, James E., and Charles C. G. Chaplin 1968. Fishes of the Bahamas and adjacent tropi- cal waters. Livingston Publ. Co., Wynne- wood, Pa. 771 pp. Bohlke, James E., and Lowell P. Thomas 1961. Notes on the west Atlantic jawfishes, Opistliognathiis aurifrons, O. lonchiirus, and Gnathypops bennudezi. Bull. Mar. Sci. Gulf and Carib. 11(4): 503-516. Briggs, John C. 1961. Emendated generic names in Berg’s clas- sification of fishes. Copeia 1961(2): 161- 166. Colin, Patrick L. 1971. Interspecific relationships of the yellow- head jawfish, Opistognathiis aurifrons (Pisces, Opistognathidae). Copeia 1971(3): 469-473. Cuvier, G. L. C. F. D. 1817. Le regne animal. Deterville, Paris. 4 vols. Cuvier, G. L. C. F. D. and Achille Valenciennes 1828-1849. Histoire naturelle des poissons. Le- vrault, Paris. 22 vols. Emery, A. R. 1 968a. Comparative ecology of damselfishes (Pisces: Pomacentridae) at Alligator Reef, Florida Keys. Ph.D. Dissertation, Univ. of Miami Inst. Marine Sci. 1968b. Preliminary observations on coral reef plankton. Limnol. and Oceanog. 13(2): 293-303. Jordan, D. S., and J. C. Thompson 1905. The fish fauna of the Tortugas Archi- pelago. Bull. U.S. Bur. Fish. (1904) 24: 229-256. Kristensen, Ingvar 1965. Der maulbriitende “jack in the box” Opistliognathiis aurifrons (Jordan and Thompson). Die Aquarien und Terrarien- Zeitschrift (Datz) 18(11 ): 321-323. Leong, D. 1967. Breeding and territorial behavior in Opis- thognatlius aurifrons (Opisthognathidae). Die Naturwissenschaften 54(4): 97. Longley, W. H., and S. F. Hildebrand 1941. Systematic catalogue of the fishes of Tor- tugas, Florida. Papers of the Tortugas Lab. 34. 331 pp., 34 pis. Meek, S. E., and S. F. Hildebrand 1928. The marine fishes of Panama. Field Mus. Nat. Hist., Zool. Ser. 15 Myrberg, a. a., JR. (in press). Ethology of the bicolor damselfish, Eiipomacentrus partitas (Pisces: Poma- centridae). A comparative analysis of laboratory and field behavior. Myrberg, A. A., jr., A. Banner, and J. D. Richard 1969. Shark attraction using a video-acoustic system. Marine Biology 2(3) : 264-276 Randall, J. F. 1967a. Food habits of West Indian reef fishes. Studies in Trop. Oceanog. No. 5, Pro- ceedings of the International Conference on Tropical Oceanography: 665-847. 1967b. loglossus helenae, a new gobiid fish from the West Indies. Ichthyologica 39(3-4): 107-118. Ray, C. 1963. It steals stones and spits sand— the yellow- head jawfish. Animal Kingdom 66(2): 42-46. Sevenster, P. 1961. A causal analysis of a displacement activ- ity (fanning in Gasterosteiis aciiieatus L.). Behavior Suppl. 9: 1-156. Starck, W. a., and W. P. Davis 1966. Night habits of fishes of Alligator Reef, Florida. Ichthyologica 38(4): 313-356. Stephens, W. 1968. Southern seashores. Holiday House, New York. 181 pp. Van Doorne, R. 1969. De Goudvoorhoofd Kaakvis. De Kor 1969: 114-119. Youngbluth, M. j. 1968. Aspects of the ecology and ethology of the cleaning fish, Labroides phthirophagus Randall. Z. Tierpsychol. 25: 915-932. Colin: Behavior of the Yellowhead Jawfjsh 165 Plate I, Figure 1. A Bahamian specimen of Opistognathus aurifrons at the mouth of its burrow. 166 New York Zoological Society: Zoologica 57 (4) Plate II, Figure 2. Underwater television (UTV ) installation located at a depth of 20 m one mile off the coast of North Bimini, Bahamas. Figure 3. Monitor room for the UTV system located at the Lerner Marine Laboratory. Visible are the control console, television monitor, event recorder, and video tap»e recorder. Colin: Behavior of the Y ellowhead Jawfish 167 Plate III, Figure 4. The action of “dig” (within burrow) performed by an individual of O. aurifrons. The dark line on the isthmus, normally hidden by folds of skin, is clearly exposed. Figure 5. Lateral view of the action of “dig” (within burrow) performed by an individual of O. auri- frons. At this point the sand is being expelled from the mouth. 168 New York Zoological Society: Zoologica 57 (4) Plate IV, Figure 6. The action of “dig” (retrieve sand) performed by a specimen of O. aiirifroiis. This action is performed some distance from the burrow opening. Figure 7. The action of "retrieve rock" (remove rock from within burrow) performed by an individual of O. aiirifroiis. Colin: Behavior of the Yellowheaci Jawfish 169 Plate V, Figure 8. The action of “arch” performed by a male specimen of O. aurifrons. The male (upper fish) arches the body, spreading the fins, then opens the mouth exposing the various dark lines on the head. The lower fish is a female. 170 Zoologica: Index to Volume 57 INDEX B baboon tissues, blood-group activity in, (2) 97-104, tables 1-4 discussion, 100-102 introduction, 97-99 materials and methods, 98-99 results, 99-100 summary, 102 Balanus ebumeus Gould, histochemical analyses of the fluid and solid state of the adhesive materials produced by the pre- and postmetamorphosed cyprids of, (2) 79-95, text-figures 1-6, tables 1-10 appendix, 93-94 conclusion and summary, 93 description of the cement apparatus, 82 adult cement apparatus, 82 basal plate structures, 82 cyprid cement gland, 82 discussion, 93 histological and histochemical reactions, 82-83 introduction, 79 materials and methods, 79 birds, rare and endangered, status of in captivity, with a general reference to mammals, (3) 119-125 Aves, 119-124 Anseriformes, 120-121 Ciconiiformes, 120 Columbiformes, 123 Falconiformes, 121 Galliformes, 122-123 general comments, 119-120 Gruiformes, 123 Passeriformes, 124 Psittaciformes, 123-124 Sphenisciformes, 120 discussion and summary, 124-125 introduction, 119 mammals, 124 methods, 1 1 9 C captive propagation, a progress report, (3) 109-117 catfish, brown bullhead, Ictalurus nebulosas C LeSueur), hematological parameters and blood cell morphology of, (2) 71-78, tables 1-3 discussion, 75-76 introduction, 71 materials and methods, 71-72 aquaria and fish, 71-72 blood letting, 72-73 determination of heratological parameters, 72-73 results, 73-75 blood cell morphology, 73-75 granulocytes, 74 hemocytoblast or hemoblast, 74 lymphocytes, 74 macrophages, 75 monocytes, 74-75 thrombocytes, 74 blood parameters, 73 summary, 76 cobra, Philippine, Naja naja philippinensis, venom yields of, (3) 126-134, plate 1, text-figure 1, tables 1-2 introduction, 127-129 material and methods, 129-130 results and discussion, 130-134 crocodile, Morelet's, preliminary report, status investi- gations of in Mexico, (3) 135-136 Alvarado, 136 Lago de Catemaco, 135 estimated population, 135 habitat, 135 hunting pressure, 135-136 prospectus, 136 cyprids, see Balanus ebumeus E Eueides, see Heliconians of Brazil H Heliconians of Brazil (Lepidoptera: Nymphalidae). Part II. Introduction and general comments with a supplementary revision of the tribe, ( 1 ) 1-40, Plates 1-6, text-figures 1-12, map Appendix I: A synopsis of the Heliconians of extra- Amazonian Brazil, 19-22 A. Normally extra-Amazonian Heliconians, 19-21 B. Marginally extra-Amazonian Heliconians, 21-22 C. Hypothetically extra-Amazonian Heliconians, 22 Appendix II: A brief list of the Heliconians of Amazonian Brazil, 23-26 Appendix III: Systematic changes and remaining uncertainties, 26 cyclic annual variations in abundance, 4 explanatory note on materials and methods, 16 introduction, 1-2 marginal species in extra-Amazonian Brazil, 4-6 Siivana-group in extra-Amazonian Brazil, 15 some specific comments on the species, 6-14 Agiaulis lucina (Felder), 6-7 Eueides pavana, 7 Eueides vibilia, 7, 12-15 Philaethria dido and P, wernickei, 6 summary, 16 taxonomy, 2 zoogeography, 3-4 Part III. Ecology and biology of Heliconius malterei, a key primitive species near extinction, and comments on the evolutionary development of Heliconius and Eueides, (1) 41-69, plates 1-6, text-figures 1-4 ecology and biology of Heliconius natierei, 42-53 chrysalis, 50-53 egg, 47-48 larva, 48-50 evolution of Heliconius and Eueides species, 53-58 explanatory addition on materials and methods, 58-60 introduction, 42 summary, 60 Heliconius natierei, see Heliconians of Brazil I Ictalurus nebulosus (LeSueur), brown bullhead cat- fish, hematological parameters and blood cell morphology of, (2) 71-78, tables 1-3 discussion 75-76 introduction, 71 Zoologica: Index to Volume 57 171 materials and methods, 71-73 aquaria and fish, 71-72 blood letting, 72-73 determination of heratological parameters, 72-73 results, 73-75 blood cell morphology, 73-75 granulocytes, 74 hemocytoblast or hemoblast, 74 lymphocytes, 74 macrophages, 75 monocytes, 74-75 thrombocytes, 74 blood parameters, 73 summary, 76 I jawfish, yellowhead, Opistognaihus aurUrons, daily activity patterns and effects of environmental condi- tions on the behavior of, with notes on its ecology, (4) 137-169, plates 1-5, text-figures 1-21, tables 1-4 comparative relationships and discussion of the ecology of, 161-163 daily activity patterns and behavioral correla- tions, 146-152 definition of behavioral actions, 139-146 actions associated with feeding, 144-145 actions concerned with interspecific relation- ships, 145 actions concerned with intraspecific relation- ships, 145-146 actions related to the maintenance of the burrow, 143-144 burrow oriented actions, 140-143 locomotory movements, 140 maintenance activity, 145 effect of current speed upon various activities, 154-155 effect of other environmental factors upon various activities, 155-56 introduction, 137-138 material and methods, 138-139 morning and evening behavior, 158 range and territory, 158-161 relationship between feeding and burrow oriented activities, 153-154 sonic activity, 1 56 time of covering and uncovering the burrow and nocturnal behavior, 156-158 N Naia naja philippinensis, Philippine cobra, venon yields of, C3) 126-134, plate 1, text-figure 1, tables 1-2 introduction, 127-129 materials and methods, 129-130 results and discussion, 130-134 O Opistognaihus auritrons, yellowhead jawfish, daily activity patterns and effects of environmental condi- tions on the behavor of, with notes on its ecology, (4) 137-169, plates 1-5, text-figures 1-21, tables 1-4 comparative relationships and discussion of the ecology of, 161-163 daily activity patterns and behavioral correla- tions, 146-152 definition of behavioral actions, 139-146 actions associated with feeding, 144-145 actions concerned with interspecific relation- ships, 145 actions concerned with intraspecific relation- ships, 145-146 actions related to the maintenance of the burrow, 143-144 burrow oriented actions, 140-143 locomotory movements, 140 maintenance activity, 145 effect of current speed upon various activities, 154-155 effect of other environmental factors upon various activities, 155-156 introduction, 137-138 material and methods, 138-139 morning and evening behavior, 158 range and territory, 158-161 relationship between feeding and burrow oriented activities, 153-154 sonic activity, 156 time of covering and uncovering the burrow and nocturnal behavior, 156-158 orangutans, captive breeding of, (2) 105-108 NOTICE TO CONTRIBUTORS Manuscripts must conform with the Style Manual for Biological Journals (American In- stitute of Biological Sciences). All material must be typewritten, double-spaced, with wide mar- gins. Papers submitted on erasable bond or mimeograph bond paper will not be accepted for publication. Papers and illustrations must be submitted in duplicate (Xerox copy acceptable), and the editors reserve the right to keep one copy of the manuscript of a published paper. The senior author will be furnished first gal- ley proofs, edited manuscript, and first illus- tration proofs, which must all be returned to the editor within three weeks of the date on which it was mailed to the author. Papers for which proofs are not returned within that time will be deemed without printing or editing error and will be scheduled for publication. Changes made after that time will be charged to the author. Author should check proofs against the edited manuscript and corrections should be marked on the proofs. The edited manuscript should not be marked. Galley proofs will be sent to additional authors if requested. Original illus- trative material will be returned to the author after publication. An abstract of no more than 200 words should accompany the paper. Fifty reprints will be furnished the senior author free of charge. Additional reprints may be ordered; forms will be sent with page proofs. Contents PAGE 10. Daily Activity Patterns and Effects of Environmental Conditions on the Behavior of the Yellowhead Jawfish, Opistognathus aurifrons, with Notes on its Ecology. By Patrick L. Colin. Plates I-V; Text-figures 1-21; Tables 1-4 137 Index to Volume 57 170 ZooLOGiCA is published quarterly by the New York Zoological Society at the New York Zoological Park, Bronx, New York 10460, and manuscripts, subscriptions, order for back issues and changes of address should be sent to that address. Subscription rates: $6.00 per year; single numbers, $1.50. Second-class postage paid at Bronx, New York. Published April 25, 1973 WiV?.' 1^1 7:('* r.«ir j l, 1 0 W' St i 'i sjSTITUTION ‘^f^0f-LniliSNl_NVIN0SHiIWs‘^S3 I H VH 8 H^U B RAR \ Es'^SMITHSONIAN JNSTITUTION O “ ' O _j - _ 2 _1 2 3!Hvaan libraries Smithsonian institution NoiiniiiSNi nvinoshiii^s saiavaan jsavaaiip^ 2 r- 2 f“ , . 2 m DC^ 2 xiiVDv^ ^ rn NSTITUTION^ NOIinillSNl"NVINOSHiiy^S S3 I BVBa n~LI B RAR I Es'^SMITHSONIAN-INSTITUTION . * w 2 w 2: ..•. w z ,Nj-. X ys/,^ o If M X g g g ^>x _ 2 > 13 I ava a n^LI B R AR l Es'^SMITHSONIAN^ INSTITUTlON‘^NOIinillSNI NVIN0SHillMs‘^S3 I a Va 8 11 — (fi ^ w 2 _ _ O NSTiTUTI0N^N0liniIlSNl"^NVIN0SHillAIS^S3 I a Va 8 IT LI B RAR I ES^SMITHSONIAN^INSTiTUTIOI\ 2: _ V 2 r- 2 E CO — t/j 5 :3iavaan libraries Smithsonian institution NouniiiSNi nvinoshiiiais saiavaan 3 1 a V a a I I _ - CO 2 W ^ S 2 X o 3, >vjvAS.>i>- 2 ^ ^ ^ >' NSTITUTION ‘^N0liniIiSNl_NVIN0SHil&Ms‘^S3 I H Va a n\l B RAR 1 es‘^smithsonian_!nstitution a: - v-'CW' CO Q xMos^ z ■ o 2 — J 2 ;3iavaan~*LiBRARiEs smithsonian institution NoiiniiiSNi NviNosHims saiavaar z ni xO\£Ci$^ w rn w m xs^dc^ ni ^ rn CO _ t/> ± to - — institution NOliniliSNI NVINOSHilWS S3IHVaan LIBRARIES SMITHSONIAN INSTITUTIO^ ... CO 2 W 5 ^ 5 V> X ./ It /*z ^ ^/■.-/ Am ^ to X , t > _ - -^ to 2 to *2 to •, S3 iavaan_ LIBRARIES SMITHS0NIAN_INSTITUTI0N^N0liniliSNI_NVIN0SHillAIS^S3 I ava 8 I ■nfw wT? 1 2 /fpti 3 :NI_NVIN0SH1IWs‘^S3 I y vy a n^Ll BRAR l ES^SMITHSONIAN^INSTITUTION NOfiniliSNI_NVINOSH < cc CD ^ ^ /oJ^' -^ol o o O 2 -» 2 _J 2: — - ES SMITHSONIAN INSTITUTION NOIlfliliSNI NVINOSHillAIS S3iyvyan LIBRARIES SMITHSO Z r- z r* ^ — to coXS to — 3NI NviNosHims SBiyvyaii libraries Smithsonian institution NoiiniiiSNi nvinosh z ^ to Z .... CO 2; to 2: ^ < s .* ^ ^ > (pf :o m pi rn 'X^os'X^ ^ m to — CO ^ CO ~ to ES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHimS S3iyvyan LIBRARIES SMITHSO ^ Z > to 2 CO 2 \ w 3 ^ 8 z , / ‘^ t vh-c.. > 2 >• 2 >• z to . ^ z to z to ^ (S| > -- to ■ ■ 2 to 2 ^ iNi_NviNOSHiiws S3iyvyan libraries Smithsonian institution NoiiniiisNi__NviNosL ^ 5 to — CO ^ ^ 2 ^ I 2 I g 5 5 4 ^ z 2 _J 2 ES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHillAIS S3iyvyail LIBRARIES SMITHSC z r-_2 i~ 2 t- 2 z to H m ^ 's^ojljBsy' m m — CO ± — to *•• _ CO — ;ni NViNOSHims saiyvyan libraries Smithsonian institution NoiiniiisNi nvinosh “* CO Z to 2 ... CO 2 - I z w/ .c, »// l/J ES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSHillAIS SBIBVaan LIBRARIES SMITHSC CO Z to — fO ~ to