RETURNING MATERIALS: PV1ESI.J Place in book drop to ”BRARIES remove this checkout from .-;_—. your record. FINES wi'l] be charged if book is returned after the date stamped be10w. ‘ . SEP25 2008 HA9 15 I99§ WM“ M D at: '.3(, ECOLOGY AND BEHAVIOR OF THE BLACK-AND-NHITE CASQUED HORNBILL (WWW) IN KIBALE FOREST, UGANDA By Jan Kalina A DISSERTATION Submitted to - Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Fisheries and Wildlife 1988 5". ABSTRACT ECOLOGY AND BEHAVIOR OF THE BLACK-AND-NHITE CASOUED HORNBILL BXQANLSIES SUBQILINDRIQQS SQBQQADBAIQS) IN KIBALE FOREST, UGANDA By Jan Kalina The ecology and behavior of the black-and-white casqued hornbill (Bycanistgs subgyliggricus subguagggtgs) was studied in logged and unlogged areas of Kibale Forest, Uganda (June l981-June l984; l986 and 1987 breeding seasons). Hornbill (family Bucerotidae) breeding behavior differs from that of other birds in that the nest cavity entrance is sealed, with the female and young imprisoned inside. This was the first long-term, intensive study of an African forest hornbill species. Hornbill movements, spatial dispersion, and habitat use varied seasonally. Hornbills were present in selectively logged areas but in lower numbers than in primary forest in the core of the Reserve. Hornbill movements were closely related to their diet. Particularly during the dry seasons, hornbills traveled long distances (> 6 km) in search of fruiting trees. These birds fed on at least 67 fruit species. Fruits comprised 90% of the diet by volume, with Eigus spp. comprising 57% of the volume of all fruits. Seeds from most species of fruits were either regurgitated or Jan Kalina defecated and dispersed intact. Hornbills, therefore, mediated seed dispersal of rainforest trees. Hornbills required very large trees (> 3 m circumference; n = 45) for nesting. Trees had large (> 25 cm depth), naturally formed cavities at a minimum height of 8 m above ground. Hornbill nest densities were highest in primary forest in .the center’ of the Reserve, where the density of trees > 3 m circumference was greater and where the trees were, on average, much larger. The number of young fledged per unit area was also highest in primary forest in the core of the Reserve. Reasons for nest failures were varied, but intrusions at nests by conspecifics often caused resident birds to abandon their nests. Competition for nest sites by hornbills was high. In this study, the nest-seal appeared to function primarily to prevent intra-specific competitors from entering the nest. Three other hypotheses concerning the function of nest-sealing are discussed. Based on findings in this study, general lrecomnendations are made concerning timber-management practices which seem likely to affect the conservation of forest hornbills in Uganda and elsewhere in the tropics. Copyright by JAN KALINA 1988 Dedicated to my family and the memory of Rebekah D. Fischer. iv ACKNOWLEDGMENTS I wish to acknowledge the President’s office, Uganda National Research Council, and the Uganda Forest Department for permission to work in Kibale Forest. The New York Zoological Society, Sigma Xi, and the American Museum of Natural History’s Frank M. Chapman Memorial Fund provided financial support. Much of this thesis was written at the U.S. Forest Service’s Institute of Tropical Forestry in Puerto Rico. A return research trip to Uganda was also funded by the U.S. Forest Service. My husband, Tom Butynski, contributed to every phase of this study through his advice, encouragement, and field assistance. His companionship and knowledge of the forest helped make completion of this work possible. My research and dissertation benefited from suggestions by nw' doctoral committee members: Professors George Petrides (chairman), Don Beaver, Niles Kevern, John King, and Peter Murphy. I am thankful to them for their guidance, accessibility, and support. I am grateful to Dr. Petrides for providing the opportunity for students of diverse backgrounds and nationalities to pursue their interests in wildlife ecology and conservation tn1 an international' scale. Perhaps more important than the structured education at Michigan State University was the inspiration gained from these other graduate students. I am especially thankful for the friendships of fellow students: Stan Koster, Heidi Grether, and John and Terese Hart. In Kibale Forest, Tom Struhsaker and Lysa Leland provided logistical support and made available their observations and comments. Joseph Skorupa, Lynne Isbell, Matti Nummelin, and Isabirye Basuta also shared information from 'their own studies. Everyone at Kibale helped locate hornbill nests. Steven Yongili, John Kyalimpa, Nyakairu Godfrey, John Rwagara, and Lawrence Rusoke were particularly successful nest-finders and were enthusiastic assistants in the field. This work would not'have been possible without their cooperation. Senior Game Guard Alfred Otim and his family also offered help freely. During trips to Kampala, Oscar and Linda Rothen kindly welcomed us into their home. Anthony Katende, curator of the Makerere University Herbarium, identified many plants. Identifications of hornbill foods were made by John Njoroge (National Museum of Kenya), John Corner, and Johnathan Baranga. Alan Kemp, Steve Martindale, and Mike Underwood offered valuable coments on observations and manuscripts. Betsy Anderson drew hornbill Figure 5.1. Mark Ritchie provided information on insects. Dr. Peter G. Waterman of the University of Strathclyde analyzed fruit samples. My sister, brother, and parents were the first to teach me how to find and observe wild animals. My greatest appreciation goes to family and friends who encouraged me to continue in graduate studies and who have reliably provided moral support and warm welcomes after vi long absences. My parents, Joseph and Celeste Kalina, handled financial and other obligations at home while I was in Uganda. Margie, Bob, and Rob Blake; Joe and Kim Kalina; Rebekah D. Fischer; JoAnne Fischer; Patty Cohen; Teresa Dorn; Jill Keilblock; Joyce Manko; and Barbara Rowe also sent much-needed mail and packages to us in Uganda. vii TABLE OF CONTENTS LIST OF TABLES ....................... LIST OF FIGURES ....................... Chapter I. INTRODUCTION .................... Reasons for Study ................. Study Animals ................... Study Areas .................... Methods ...................... REFERENCES .................... II. BLACK-AND-HHITE CASOUED HORNBILL DENSITIES IN KIBALE FOREST, UGANDA: SEASONAL AND HABITAT VARIATIONS Introduction ................... Study Area .................... Methods ...................... Resident Hornbill Densities ........... Development of the Census Method ........ Line-Transect Censuses ............. Calling-Rate Index (CRI) ............ Relative Abundances Based on a Total Detection Index ..................... Results/Discussion ................ Confidence in Census Results .......... Relative Abundance of Hornbills in Different Habitats ................... Seasonal Trends in Habitat Use ......... Overall Hornbill Numbers in Kibale ....... Conclusions .................... REFERENCES .................... APPENDICES Z-A. Calling Rate Index (CRI) Data Collected During lS-Second Scans at S-Minute Inter- vals During Transects Censusing Bycanistgs Hornbills in Ngogo, Kibale Forest, Uganda (1983-84) ......... viii 43 2-8 Calling Rate Index (CRI) Data Collected During 15-Second Scans at S-Minute Inter- vals During Transects Censusing Byganistgs n ' Hornbills in K-30, Kibale Forest, Uganda (1983-84) ......... Z-C Calling Rate Index (CRI) Data Collected During lS-Second Scans at 5-Minute Inter- vals During Transects Censusing Bygagistgs subgylingriggs Hornbills in K-15, Kibale Forest, Uganda (1983—84) ......... III. DIET OF BYCANISIES SQBCILINQBIQUS, HITH EMPHASIS ON THE NESTING SEASON ............... Introduction ................... Methods ...................... Opportunistic Observations ........... Nest-Hatches .................. Nest-Trap Data ................. Nutritional Analysis .............. Fruit Morphology ................ Results and Discussion .............. Foods and Foraging ........ ‘ ....... Nest-Hatch Data ................. Nest-Hatch Size Categories ........... Nest-Trap Data ................. Nutrition .................... Conclusions .................... REFERENCES .................... APPENDICES 3-A A Systematic List of All Food Items Recorded in the Diet of Byganistgs sub- gylindrjggs ................ 3-B Description of Fruits in the Diet of Byganistgs subcylindriggs Hornbills in Uganda .................. 3-C Chemical Analysis of Some Byganistes 5gb_ cylindnigus Food Items (Analysis by P. Haterman) ................. 3-D Chemical Analysis of Some Byganjstgs sub- ;yljndrjggs Food Items (Analysis by Colo- rado State Univ.) ............. ix Page 44 7O 75 77 78 IV. BREEDING BIOLOGY OF THE BLACK-AND-HHITE CASOUED HORNBILL (BIQANISTES SUBQILINDBLQUS) IN KIBALE FOREST, UGANDA .................. 79 Introduction ................... 79 Methods ...................... 80 Results ...................... 83 Timing of Nesting ................ 83 Courtship and Nesting Behavior ......... 83 Nest-Site Characteristics ............ 89 Density of Nests ................ 90 Nesting Success ................. 90 Sex-Ratio Bias in Fledglings .......... 96 Discussion/Recommendations for Further Study . . . 98 REFERENCES .................... 101 APPENDICES 4-A Characteristics of Byganjstgs subcyljndzi- cg; Nest Sites in Kibale Forest, Uganda . . 103 4-8 Characteristics of Bycanistgs sgbcylindri- cu; Display Cavity Sites in Kibale Forest, Uganda .................. 104 4-C Dimensions of 31m subsxljpdcisus Nest Cavities in Kibale Forest, Uganda . . 105 4-0 Bycagistgs subgylindgigus Nest Occupancy and Nesting Success Over Five Years (1981- 1986) in Kibale Forest, Uganda ...... 106 V. NEST INTRUDERS, NEST DEFENSE, AND FORAGING BEHAVIOR IN THE BLACK-AND-HHITE CASOUED HORNBILL ...... 110 Introduction ................... 110 Methods ...................... 113 Results ...................... 115 Discussion .................... 121 REFERENCES .................... 124 VI. FUNCTION OF NEST-SEALING: HITH SPECIAL REFERENCE TO THE BLACK-AND-HHITE CASOUED HORNBILL (BIQANISIES SUBCYLINDRICUS) ............ 125 Introduction ................... 125 Predation Hypothesis: Nest- -Sea1ing Protects Against Predation ................ 127 Interspecific- Competition Hypothesis: Nest- Sealing Protects Against Interspecific Competition ................... 129 Intraspecifie-Competition Hypothesis: Nest- Sealing Protects Against Intraspecific Competition ................... Microclimate Hypothesis: Nest-Sealing Protects Against Adverse Heather ............. Discussion .................... Convergent Evolution Among Unrelated Bird Species .................... Natural History Studies of Hoopoes . ....... Comparative Field Studies of Hornbills ..... Summary ...................... REFERENCES .................... APPENDICES 6-A Descriptions of Bygagistgs subgyljndrjgus Behaviors, Sounds, and Vocalizations Mentioned in Appendices 6-8 and 6-C . . . . 6-B Example of Intruding Byganjstes subgylin- drjgus Supplanting Residents Early in the Nest-Cycle (Kibale Forest, Uganda) 6-C Example of Intruding Byganistes subgylin- dricus Supplanting Residents Late in the Nesting Cycle (Infanticide) ........ VII. SUMMARY AND RECOMMENDATIONS ............. REFERENCES .................... REFERENCES ......................... xi Page 132 135 139 140 140 141 143 145 150 153 156 163 168 169 Table 2.1 2.2 LIST OF TABLES Summary of Kibale Forest Study-Area Characteristics Numbers of Byganjstgs sgpgylindriggs Hornbills Detected at Ngogo During Transect Censuses in Kibale Forest, Uganda (1983-84) .......... Numbers of Byganistes subcylingriggs Hornbills Detected at K-30 During Transect Censuses in Kibale Forest, Uganda (1983-84) .......... Numbers of Bygagistes subgylindrjggs Hornbills Detected at K- 15 During Transect Censuses in Kibale Forest, Uganda (1983- 84) .............. Census-Route Characteristics, Kibale Forest, Uganda Bycagjstgs subgylindriggs Density, Biomass, and Ratio of Transients to Residents in Kibale Forest, Uganda (1983-84) .................. Bycanistes subcylindricus Nest-Hatch and Nest-Trap Locations and Collection Regimens in Kibale Forest, Uganda (1983-84) .................. Rank Order for Fruits and Invertebrates Fed at subgylingrjggs Nests During Nest- an Hatches in Kibale Forest, Uganda (1983-84) ..... Fruits and Invertebrates Fed Into Bycanjstgs Sghgxlingzigus Nests During Nest-Hatches (1983-84), by Size Category .................. subgylindriggs Food Remains Collected From Nest- -Traps, Ranked in Order of Percentage of All Remains Found in Traps ............... Rank Order by Percentage for Frequency of Presence or Absence of Particular Invertebrates in Byganistes subgyljndrjgus Nest-Traps, Kibale Forest, Uganda (1983-84) ..................... xii Page 13 20 21 22 23 33 SO 58 59 62 63 4.1 4.2 5.1 5.2 Density of Byggnistgs subgylingriggs Nest Cavities, Display Cavities, and Breeding Individuals in Kibale Forest, Uganda ............... Total Rainfall (mm) During Byganistgs Pre- -breeding and Breeding Months (July- March) in Kibale Forest, Uganda (1981- 1984) ......... Changes in Four Measurements of Foraging Behavior of Nesting Male Byganistes subcyljndrigus in Kibale Forest, Uganda (1983-84) Before, During, and After Intrusions by Conspecifics at the Nest Site ..................... Composition of Fruit Loads Brought to the Nests by Male Byganistgs subcylindriggs Hornbills in Kibale Forest, Uganda (1983-84) Before, During, and After Intruder Attacks .................. xiii Page 85 99 116 118 Figure 1.1 2.9 2.10 LIST OF FIGURES Distinguishing Characteristics of Several Individual Byganistgs subgylindniggs Hornbills in the Kibale Forest, Uganda ................... Location of Kibale Forest in H Uganda, East Africa . . Location of Study Areas in Kibale Forest Reserve, Uganda ....................... Locations of Census Route (5.34 km) at Ngogo, Kibale Forest, Uganda ................... Locations of Census Route (4.7 km) at K-30, Kibale Forest, Uganda ................... Locations of Census Route (5.28 km) at K-15, Kibale Forest, Uganda ................... Byganistg; subgylingrjcus Calling Rate Indexes on Three Study Areas in Kibale Forest, Uganda (1983-84) ..................... Relative Abundance of Bygaflistgs subcyljndrigus Hornbills in Three Study Areas During Different Stages of the Breeding Cycle (1983-84; Kibale Forest, Uganda) .................. Percentages of Bygagistes subgyliggrjgus Hornbills Detected at Various Distances During Censuses in Three Study Areas of Kibale Forest, Uganda (1983-84) ..................... Relative Abundance of Bycanistes sybgyljngriggs in Kibale Forest, Uganda, Showing Seasonal and Among- Habitat Variation ................. Byganistes subgylindrjggs Hornbills Recorded in K-15 During August 2, 1983, Census ........... Byggnjstgs subcyljndrjgus Hornbills Recorded in K-15 During February 10, 1983, Census .......... xiv Page 12 17 18 19 27 29 30 32 37 38 Relationship Between Percentage Annual Rainfall, Fruit Abundance (Measured by the I'Surplus Fruit Index”), and the Period Over Hhich Byganistgs subgylingrjggs Start Nesting in Kibale Forest, Uganda (January 1982-1984) ............. Courtship Feeding by Bygagistgs sybgyljndrigus . . . . Locations of Misses subsxundnisus Nests (n - 30) and Display Cavities (n - 10) at Ngogo (Unlogged Forest) in Kibale Forest, Uganda (1981-1987) . . . . Locations of Bygaflistgs subcyljndrjgus Nests (n - 7) and Display Cavities (n - 4) at Kanyawara (K-30 Unlogged Forest) in Kibale Forest, Uganda (1981- 1987) ....................... Locations of Byganistes subcyligdrigus Nests (n - 4) at Kanyawara (K-l4 Light-Moderately Logged Forest) in Kibale Forest, Uganda (1981-1987) ........ Locations of Byganistgs subcyljngrjgus Nests (n - 3) at Kanyawara (K-lS Heavily Logged Forest) in Kibale Forest, Uganda (1981-1987) ......... Difference Between Casque Development on Fledgling Male and Fledgling Female Byganistes subgylin- SLDLCJLS OOOOOOOOOOOOOOOOOOOOOOO During the Nesting Period, the Male Byggnistes subgylindrjggs Is Solely Responsible for Feeding Himself, His Mate, and His Offspring in the Nest. At the Same Time, His Foraging Behavior Hill Be Influenced by Risks of Predation and Demands of Protecting the Young ................ Compass Direction by Bxganistgs subcylingrigus Nest Entrances in Kibale Forest, Uganda (n - 44) Hypothetical Relationship Between Byganjstes sub- gyljndrigus Nest Success and Number of Trees Removed Per Unit Area ............... XV Page 84 86 91 92 93 94 97 111 138 CHAPTER I INTRODUCTION W Hornbills (family Bucerotidae), with their unique nest-sealing behavior and ornate casques, have interested scientists and laymen alike. Although approximately 10 species have been the subject of detailed study (Kemp 1976; Kemp & Kemp 1980; Leighton 1982), little is known about most of the 53 species in this group of birds (Kemp 1979). Practical constraints probably account for the lack of data, since most of the species inhabit remote rainforests of Africa and SE Asia (Kemp 1979, 1988). Today, as rainforest destruction escalates (Sommer 1976; Bolin 1977; IUCN 1980; Myers 1984), it is timely to focus attention on hornbills which depend on that habitat to survive. Hornbills mediate seed-dispersal for rainforest trees (Leighton 1982; Kalina & Butynski in prep) and, therefore, facilitate regeneration of these threatened ecosystems. Previous studies of the behavior and ecology of' hornbills have focused on species inhabiting either African savannahs (Kemp 1976; Kemp & Kemp 1978) or SE Asian rainforests (Leighton 1982). Little is known about the 12 hornbill species, including the black-and-white casqued hornbill (Bycanistgs subgylingrjcus subguadratus), which occur' in African rainforests (Kemp 1979). This is the first long-term, intensive study of an African forest hornbill. Objectives of this research are to: (1) determine. hornbill densities, movements, spatial dispersion, and habitat use in areas of disturbed and undisturbed forest; (2) describe food habits; (3) investigate hornbill reproductive ecology and reasons for nesting failure; (4) develop suitable methods for collecting this information; and (5) make recommendations concerning timber-management practices which are pertinent to the conservation of forest hornbills in Uganda and elsewhere in the tropics. It is planned that results from this study will be relevant to the formulation of scientifically based policies for the protection of wildlife and the utilization of tropical rainforest by humans. Study Anjmals Black-and-white casqued hornbills (B. subcylindricus) are found in forests of Cameroon, N and E Zaire, S Sudan, Uganda, H Kenya, NH Tanzania, and N Angola. They are large (1.3 kg; Kemp 1979), black birds with gleaming white secondary feathers, abdomens, rumps, and outer tail feathers. During their distinctive glide-flapping flight, B. subgylindricus wings produce a loud whooshing sound which can be heard 100 m away. They are extremely vocal birds and produce a great variety of sounds, the loudest of which can be heard up to 2 km away. Males are larger than females, with deeper voices and. more developed casques. B. suchlindrigus are easily habituated and make excellent research subjects. They can also be recognized individually' by the unique patterns of white on the bicolored casque. The angle of the forward projection of the casque, the number of vertical and horizontal ridges, and variations in size or shape of the casque also help in the identification of individual birds (Fig. 1.1). The age of a hornbill can be determined by noting the color and development of the casque and by Observing the birds’ behavior and interactions with other individuals. Hornbills less than one year old have brown feathers on the forehead and make squeaky vocalizations. Subadults travel in flocks of approximately 5 to 12 individuals. Adults are most often seen in pairs. u as A Data were collected in moist, evergreen rainforest in Kibale Forest Reserve (560 kmz; elevation 1590-1110 m), H Uganda (O°13'-0° 41'N and 30°19'-30°32'E; Fig. 1.2). Kibale is one of the few forest patches left in Uganda, a country with less than 3% of its land surface now covered in closed forest, of which an estimated 2% (110 kmz) is being lost each year (Hamilton 1984; Struhsaker 1987). Kibale has been subjected to logging in its northern third (Struhsaker 1972) and has been the focus of much comparative ecological research in logged and unlogged areas for the past 15 years (Struhsaker 1975; Ghiglieri 1984; Skorupa 1988; Butynski in prep). The Reserve includes 560 km2 of land, but because of historical, edaphic, and altitudinal factors, only about 60% to 70% of any given area is covered by high forest (Hing 81 Buss 1970). Hith the northern third selectively logged (Struhsaker 1972) and much of the southern third deforested by agricultural encroachment Fig. l.l.--Distinguishing characteristics of several individual Byganjstes subcylindrjggs hornbills in the Kibale Forest, Uganda. - (Al-’— , \ 111141111011 1 p J (V ( nunuun;,' TANZANIA Fig. 1.2.--Location of Kibale Forest in H Uganda, East Africa. Inset: Location of Ngogo and Kanyawara (K-30, K-l4, K-15) study areas in Kibale Forest Reserve, H Uganda. (Adapted from Ghiglieri 1984 and Skorupa 1988.) (Struhsaker 1981), only an estimated 185 km2 remains as primary forest (Kingston 1967; Collar & Stuart 1985). Four study areas in Kibale were used for this research, three at Kanyawara in the northwest (K-30, K-14, K-15) and one in the center of the Reserve (Ngogo) (Fig. 1.2 inset). Ngogo is part of a 60 km2 nature reserve and is protected from all human use other than that of nonintrusive scientific research (Kingston 1967; Struhsaker 1972). Kanyawara is divided into adjacent timber compartments, which were subjected to various levels of selective timber harvesting in 1968-69 (see Skorupa 1988 for detailed description of management history of Kanyawara). Since 1970, all legal exploitation of timber has stopped at Kanyawara. Although some poaching of animals occurs in the Reserve, hornbills are not hunted. Two hornbill species occur sympatrically with B. subgyljngrjggs in Kibale Forest (Hilliams 1967). The crowned hornbill (IQCLNS alboterminatgs) (0.3 kg; Kemp 1979) was commonly observed, primarily along forest edge. The African pied hornbill (Loggus fasciatus) (0.3 kg; Kemp 1979) was not seen but possibly can be found locally in other parts of Kibale Forest. Methods A series of separate papers has been prepared, each of which relates to a different aspect of B. subgylindrjcus behavior and ecology. Methodology for data collection pertaining to each subject is presented at the beginning of each chapter. REFERENCES Bolin, B. 1977. Changes of land biota and their importance for the carbon cycle. Science 196:613-615. Butynski, T. M. in prep. Comparative ecology of blue monkeys (Cercopithecus mitis) in high and low density subpopulations. Collar, N. J. and S. N. Stuart. 1985. r n ir fri and Related Islands. IUCN, Cambridge, UK. Ghiglieri, M. P. 1984. T e im n ' 1 F re . Columbia Univ. Press, N.Y. Hamilton, A. C. 1984. Deforestation in Uganda. Oxford Univ. Press. Nairobi, Kenya. IUCN. 1980. Save the rainforests. IUCN Bgllgtin: 11(5). Gland, Switzerland. Kalina, J. and T. M. Butynski in prep. Seed dispersal ecology of Trichilia splendida in Kibale, Forest, Uganda. Kemp, A. C. 1976. A study of the ecology. behaviour and systemat- ics of Tockus hornbills (Aves: Bucerotidae). Transv. Mus. Mom. 29. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. Living Birg 17:105-136. _ Kemp, A. C. 1988. Bucerotidae--In Urban, E., S. Keith and H. Fry (eds.), The Birds ofsAfrjca Vol. 3. Academic Press, London. Kemp, A. C. and M. I. Kemp. 1978. nggrygs and Sagittarius: Two modes of terrestrial predation. Proc. Svmp. Afr. Predatory Birds: 13-16. Kemp, A. C. and M. I. Kemp. 1980. The biology of the Southern ground hornbill, Bucorvgs Igadbeatgri (Vigors), (Aves: Bucerotidae). Ann. Iransvaal Mus. 32:65-100. Kingston, 8. 1967. Horking Plan for Kibale and Itwara Central fg£g§1_fig§gnvg§. Uganda Forests Department, Entebbe, Uganda. Leighton, M. 1982. Fruit resources and patterns of feeding, spacing and grouping among sympatric Bornean hornbills. Ph.D. dissertation, Univ. Calif., Davis. Myers, N. 1984. The Primar ur Tr i r and r Future. H. H. Norton, N.Y. Skorupa, J. P. 1988. The effects of selective timber harvesting on rainforest primates in Kibale Forest, Uganda. Ph.D. disserta- tion, Univ. Calif., Davis. Sommer, A. 1976. Attempt at an assessment of the world’s tropical moist forests. Unasylve 28:5-24. Struhsaker, T. T. 1972. Rainforest conservation in Africa. Erimetes 13:103-109. Struhsaker, T. T. 1975. 1he_3eg_§eleBe§_flenkey. Univ. of Chicago Press, Chicago, Illinois. Struhsaker, T. T. 1981. Forest and primate conservation in East Africa. Afr. J. Ecgl. 19:99-114. Struhsaker, T. T. 1987. Forestry issues and conservation in Uganda. ngl. genserv. 39:209-235. Hilliams, J. G. 1967. A Field Guide to the National Parks ef E, Afrjee. Collins, London. Hing, L. D. and I. O. Buss. 1970. Elephants and forests. Hildlife Monographs: No. 19. The Hildlife Society, Hashington, D.C. CHAPTER II BLACK-AND-HHITE CASOUED HORNBILL DENSITIES IN KIBALE FOREST, UGANDA: SEASONAL AND HABITAT VARIATIONS Introduction Today, as rainforest destruction escalates (Sommer 1976; Bolin 1977; IUCN 1980; Myers 1984), it is timely to focus attention on hornbills (family Bucerotidae) which depend on that habitat to survive. The majority of hornbill species require large, hollow trees for nesting and vast areas of tr0pical forest for foraging (Kemp 1979). These large birds disperse tree seeds, thereby facilitating regeneration of the ecosystems they inhabit (Leighton 1982; Kalina & Butynski in prep). As selective timber harvesting continues to convert primary forests to agricultural land or to secondary forest, conservationists have begun to examine the capacity of disturbed forest to support wildlife (Johns 1983, 1985; Skorupa 1988). Results from surveys of SE Asian forests indicate that most hornbill species studied can persist in selectively logged forest, provided the area is large enough (Kemp & Kemp 1975; Hilson & Johns 1982; Johns 1987) and has sufficient numbers of large trees (Kemp & Kemp 1975). On the other hand, hornbills in primary forest have abandoned territories after even neighboring areas were disturbed lO (Leighton in Kemp 1985). Little is known about the effects of logging on African forest hornbill species since, prior to this research, they have not been the focus of detailed study. Because so many hornbill species are nomadic, or seasonally so, censuses designed to compare populations in' different habitats should be conducted at least several times of year. Descriptions of seasonal movements are needed to improve the interpretation of census data. Also needed are quantitative data concerning breeding activities and reproductive success in various habitats. Seasonal fluctuations of black-and-white casqued hornbill (Bycanistes subcylindrieus eubguadratus) densities in logged and unlogged areas were investigated in Kibale Forest, Uganda. Results of population censuses and breeding activity in various habitats are the focus of this report. This paper is the first in a series of reports comparing the socioecology of black-and-white casqued hornbills (hereafter referred to simply as "hornbills”) in logged and unlogged forest. Recomendations are made concerning timber- management practices which seem likely to affect the conservation of forest hornbills wherever they occur in the tropics. Stgdy Aree Data were collected in the moist, evergreen rainforest of the Kibale Forest Reserve (560 kmz; elevation 1590-1110 m), western Uganda (0°13'-O°41' N and 30°l9'-30°32' E). Kibale is one of the few forest patches left in Uganda, a country with less than 3% of its land surface now covered in closed forest and an estimated 2% of 11 in closed forest and an estimated 2% of that remnant (110 kmz) being lost each year (Hamilton 1984; Struhsaker 1987). Kibale has been subjected to logging in its northern third (Struhsaker 1972) and has been the focus of comparative ecological research in logged and unlogged areas for the past 15 years (Struhsaker 1975; Ghiglieri 1984; Skorupa 1988; Butynski in prep). The Reserve includes 560 km2 of land, but because of historical, edaphic, and altitudinal factors, only about 60-70% of any given area is covered by primary forest (Hing 81 Buss 1970). Hith the northern third selectively logged (Struhsaker 1972) and much of the southern third deforested by agricultural encroachment (Struhsaker 1981), an estimated 185 km2 remains as undisturbed forest (Kingston 1967; Collar & Stuart 1985). Four study areas in Kibale were used for this research, three at Kanyawara in the northwest (K-30, K-14, K-15) and one in the center of the Reserve (Ngogo) (Fig. 2.1). Ngogo is part of a 60 km2 nature reserve and is protected from all human use other than that of nonintrusive scientific research (Kingston 1967; Struhsaker 1972). Kanyawara is divided into adjacent compartments, which were subjected to various levels of selective timber harvesting in 1968- 1969 (Table 2.1; see Skorupa 1988 for description of management history of Kanyawara). Since 1970, all legal exploitation of timber has stopped at Kanyawara. Illegal timber harvesting has continued at a low level outside of the study areas. Although some poaching of animals occurs in the Reserve, hornbills are not hunted. 12 ‘ 01115 K14 Kanyawara .1130 NATURE RESERVE KIBALE FOREST RESERVE PD ‘ D Fig. 2.1.--Location of study areas in Kibale Forest Reserve, Uganda. (Adapted from Skorupa 1988.) 13 .meP mansoxm .m own .Awmp .m> ompv oaomz ems» saw» eon mxmc have» mgos Mxmpv mm .m>m co um>pmums mgmzmxzmx .omomz coca mgmzmzcmx um gmzmvg “umv so m was Aewm_1m~mpv menu m sow ppmwcmuc Fmaccm cums-unmgg cm mxmcaazm .z..hm pcmsugmaeou e.~ a>mmz emnsrh m.m ~.p Amp omm— mpnx mumgmnoe acmEpngEoo P.m nugmpg gwnswh m.m p.~ Amp comp epux unwspgmanu o noncsumwuca gmnsvh o.m m.p nmp comp omux m>gmmmg o uaaezpmwuca agape: o.om 8.0 map ommp omomz nm>osmm mcpmmod msumum ANExV m~wm ANExV masuv AEV mmg< a: mewpm pmwu . acmEpEmaeoo wNVm mog< Ppmmcvmm newssou .02 .u mo mmgoma acmewmmcmz amused auaum szcc< cam: wczuwup< auaum .muwumwcmuomgmso amen-avspm amused mpmnvx mo mumsssmuu.F.~ epoch 14 A system of trails l m wide is maintained in each study area. The grid system includes c. 140 km of trails mostly 50 m apart at Kanyawara and c. 140 km of trails mostly 100- m apart at Ngogo. Since only stems < 4 cm dbh are cut for these paths, there seems to be little effect on the wildlife. Yet these trails are highly beneficial to research, permitting easy movement through the forest without disturbing animals. They also served as spatial references during censuses. Ngogo and Kanyawara study areas have been sites of comprehensive habitat analysis. Detailed descriptions of habitat and primate communities in these areas have been completed by Ghiglieri (1984) and Butynski (in prep) for Ngogo and by Struhsaker (1975), Skorupa (1988), and Butynski (in prep) for Kanyawara. Ngogo was found to have a higher tree density and .basal cover plus a greater tree-species richness and diversity than Kanyawara (Butynski in prep). Butynski (in prep) compared his results with those of Skorupa (1985) and concluded that the habitat and primate communities of the undisturbed Ngogo and the unlogged K-30 were more similar than that of the unlogged K-30 and logged areas (K-l4, K-15) at Kanyawara. Study-area descriptions are summarized in Table 2.1. Methods i t rn 1 iti In June 1981, I began a long-term study of black-and-white. casqued hornbills in Kibale Forest. Intensive research ended in June 1984, but nest monitoring continued every year (except 1985) 15 through 1987. This research has focused on hornbill behavior and on reproductive and feeding ecology. From the start, every effort was made to locate all hornbill nests in the four study areas of Ngogo, K-30, K-l4, and K-15. By the 1983-84 breeding season, all nesting territories in these study areas had been found. ”Resident hornbills" are defined here as those with nesting territories. In this paper, resident hornbill densities for 1983-84 are presented for comparison with hornbill density estimates from 1983—84 line- transect census results. Beveloumeut uf the Census Methog Although nests provided information about the densities of breeding birds, it was impossible to determine seasonal changes in abundance and habitat use of hornbills without making regular population censuses“ 'The objectives of these censuses were to describe seasonal changes in abundances of hornbills within and between logged and unlogged habitats and to estimate overall numbers and biomass of hornbills in Kibale Forest. In developing a technique suitable for censusing hornbills, it was first necessary to determine if calling rates changed seasonally and whether hornbills were easier to detect in logged forest than in unlogged forest. From April 1982 to February 1983, I walked 107.4 km of censuses (37.4 km at Ngogo, 70.0 km at Kanyawara) before developing a method that seemed adequate to address potential census problems. Data collected during the April 1982-February 1983 censuses are not 16 reported here, but experience and information gained in that preliminary work were essential for accuracy in the later data collection. During this period I (l) learned to detect, identify, and record the hornbill’s sex, age, and type of call quickly; and (2) developed skills for recording distances accurately in different forest types. i - n n Due to time constraints, population censuses were limited to Ngogo (unlogged), K-3O (unlogged), and K-15 (heavily logged) study areas. Hornbill abundance in these study areas was measured using a modified version of line-transect techniques used for sampling bird communities (Emlen 1984) and for estimating primate densities in Kibale (Struhsaker 1975; Ghiglieri 1984; Skorupa 1988; Butynski in prep). As a means of comparison with prior studies in Kibale, my census routes were nearly identical to those used previously by these other researchers. I conducted replicate censuses along three_routes, one each in Ngogo, K-30, and K-15 study areas (Figs. 2.2, 2.3, 2.4). These were conducted each month from February 1983 through May 1984 (Tables 2.2, 2.3, 2.4). Every effort was made to conduct one census per month along each route. Sometimes, however, two census were conducted in one month to make up for previous months when no census was run. Biannually, I conducted a series of five replicate censuses on consecutive days, or nearly so, in each study area. These "census-series" were undertaken immediately after the hornbill CENSUS ROUTE [5.34 km) a Fig. 2.2.--Locations of census route (5.34 km) at Ngogo, Kibale Forest, Uganda. Shaded areas are grassland, is trail grid, ----- is elephant trail, and ==-== is motor- able track. (Adapted from Ghiglieri 1984.) v .s/ It 4"“. 4‘ \wgy....“‘\ \$~MWN , 060 \ ., $0, .. ’ f. ’ \ 4 O ossoooowoosoc 00 a. s. I .. m 3.41%.???»....“.».“m§.....we.., 1 .QooooosOSOIoo s o 4.400.403 349?. II! x . 89¢ a. » Svevevfiuws. x14 Al". .§nwnsv‘v .I.‘v.v “Wm Monk 4% A». m. a OI ubll m Census route (4.70 km) is trail grid. -Locations of census route (4.7 km) at K-30, Kibale Forest, Uganda. Fig. 2.3.- l9 K-15 l 1 \ N 1NBRD’ 1 1 1 41 Q‘ss‘fi o 100 200 zoom CENSUS ROUTE (5.28 km) Fig. 2.4.--Locations of census route (5.28 km) at K-15, Kibale Forest, Uganda. is trail grid. Table 2. 2. --Numbers of Byeeu_stes transect censuses in Kibale Forest, Uganda (1983- 84). Route length - 5. 34 km. Means are + standard error. 20 suueylinguieus hornbills detected at Ngogo during Total No. No. Detected Total No. No. Detected No. per km2 Detected per km Detected' per km Min. Density Census Strip Hidth (m): (400) (400) (50) (50) (Based on 50 m) Date (1983) Feb. 11 21 3.9 20 3.8 74.0 Feb. 24 10 1.9 5 0.9 18.5 Feb. 25 2 0.4 -, 2 0.4 _ 7.4 Feb. 26 7 1.3 X-1.0+0.7 6 1.1 X-O.7+O.5 22.2 Feb. 27 7 1.3 " 6 1.1 ' 22.2 Feb. 28 1 0.2 O 0 0 Mar. 25 l 0.2 O O 0 Apr. 30 O O 0 O 0 May 25 9 1.7 2 0.4 7.4 June 26 20 3.8 9 1.7 33.3 July 22 34 6.4 9 1.7 33.3 Aug.16 16 3.0 5 0.9 18.? Aug. 17 42 7.9 __ 23 4.3 _ 85. - Aug. 19 24 4.5 x-4.s+2.1 9 1.1 x-z.3+1.s 33.3 3'49-0129-0 Aug. 21 30 5.6 ’ 18 3.4 ' 66.6 Aug. 23 15 2.8 6 1.1 22.2 Sept. 30 44 8.2 23 4.3 85.1 Nov. 9 14 2.6 7 1.3 25.9 Nov. 30 22 4.1 9 1.7 33.3 Dec. 31 43 8.1 7 1.3 25.9 (1984) Jan. 25 26 4.9 7 1.3 25.9 Feb. 19 5 0.9 3 0.6 11.1 Mar. 24 8 1.5 7 1.3 25.9 Apr. 30 7 1.3 3 0.6 11.1 May 14 10 1.9 2 0.4 7.4 Total 78.3 188.0 35.2 695.6 7+ SE 3.1:2.5 1.5:5.7 1.411.: 27.8+24.8 Table 2.3.--Numbers of 21 ' hornbills detected at K-3O during transect WW censuses in Kibale Forest, Uganda (1983-84). are 1 standard error. Route length - 4.7 km. Means Total No. No. Detected Total No. No. Detected No. per km2 Detected per km Detected per km Min. Density Census Strip Hidth (m): (400) (400) (50) (Based on 50 m) 1211: (1983) Mar. 5 0 0 O 0 0 Mar. 7 O 0 O 0 . 0 Mar. 10 5 1.1 X-O.6+1.0 2 0.4 Y-O.21-0.2 8.3 Mar.11 10 2.1 ' 2 0.4 8.3 Mar. 12 0 0 0 0 0 Mar. 30 5 1.1 2 0.4 8.3 May 3. 9 1.9 2 0.4 8.3 May 31 O O O O 0 July 3 22 4.7 8 1.7 33.4 July 26 12 2.6 2 0.4 8.3 July 27 22 4.7 7 1.5 29.2 __ July 28 24 5.1 I-a.3+1.6 a 1.7 li-o.9+o.7 33.4 xa16.0+12.7 July 29 13 2.8 ' 5 1.1 ' 20.9 T Aug. 1 6 1.3 O 0 0 Sept. 12 8 1.7 4 0.9 16.7 Oct. 3 14 3.0 7 1.5 29.2 Nov. 3 11 2.3 4 0.9 16.7 Dec. 10 20 4.3 9 1.9 37.5 (1984) Feb. 11 19 4.0 5 1.1 20.9 Feb. 27 5 1.1 0 0 0 Mar. 11 0 0 0 0 0 Apr. 2 2 0.4 0 0 0 May 2 10 2.1 4 0.9 16.7 Total 46.2 71.0 15.1 296.1 X'+ SE 2.0:1.7 3.1:3.1 0.7:0.7 12.9112.8 22 Table 2. 4. --Numbers of Byeenistes suheylingrieus hornbills detected at K-15 during transect censuses in Kibale Forest, Uganda (1983- 84). Route length - 5.28 km. Means are + standard error. Total No. No. Detected Total No. No. Detected No. per km2 Detected per km Detected per km Min. Density Census Strip Hidth (m): (400) (400) (50) (50) (Based on 50 m) Date (1983) Mar. 2 Mar. 3 Mar. 6 Mar. 8 Mar. 9 Apr. 2 May 7 June 1 July 7 Aug. Aug. Aug. Aug. Aug. Sept. 11 Oct. 5 Nov. 5 Dec. 11 (1984) Feb. 10 45 Feb. 22 29 Mar. 14 14 19.3 Apr. 3 0 0 May 3 3 2 7.7 Total 37.1 77.0 14.6 296.5 Y+ SE 1.15:2.0 3.4:s.o 0.611.0' 12.91192 0 N '0 {-03:02 Y-o.o:o.1 d O woomo 6.." NNO O o o o O ‘4‘.“ GNND 404004 mmwwmuuadmomeJ-NNG 230.8104 i-o.3:o.2 114.2353 d “$.50,” o —l NNDNUNOhM‘OO-‘Nd-‘U w-fl-‘NO-‘OO-‘Odd-‘OOOOO maamouNNdo-a-auoooo-o :HOO-‘OOOOOOOOOQOOOO .UIUIPO-‘NNUOMN—‘OOOOOO ”—0 OO u-a—o-a 80.9 50.1 21 13 5 CON!!!” 0! N090! OO-‘N. . O O I -. 009° 23 nesting season (Feb./Mar. 1983) and then again immediately before the next nesting season (August 1983). Biannual census-series were conducted in order to obtain larger samples for comparing changes in hornbill abundance during different seasons. 'Daily censuses also helped identify variations in hornbill densities on a daily rather than monthly basis. The total number of censuses conducted per study area varied from 23 to 25 (Tables 2.2, 2.3, 2.4). This sample size coincides with previous analyses recommending that 20-30 repetitions would be optimal for censusing primates in Kibale (Struhsaker 1981). Route lengths ranged from 4.7 km to 5.3 km (Table 2.5). The total number of km censused per route ranged from 108.1 to 133.5. Including all study areas, a total of 363.0 km of censuses were walked (Table 2.5). Table 2.5.--Census-route characteristics, Kibale Forest, Uganda. Total Study Area Route Type Route Census Length (km) Route (km) Ngogo Trail grid 5.34 133.50 K-3O Trail grid 4.70 108.10 K-15 Trail grid and 5.28 21.44 logging road Total for all study areas 363'04 24 Two censuses (Ngogo and K-30) were conducted entirely along the 1 m wide trails which were part of the study-area grid system. The K-lS census route was located along trails, with the final 1 km along an old logging road (Figs. 2.2, 2.3, 2.4; Table 2.5). Vegetation on this road was periodically cut, but it was my opinion that visibility there was no better than along the usual trails. Logging roads have been incorporated into census routes by other researchers in Kibale, and according to analysis by Skorupa (1988), for five of seven primate species studied there was no bias to results. The trail system in Kibale was extremely beneficial during censuses. Hith undergrowth removed, I walked quietly at a steady pace (c. 1 km/h) and therefore avoided startling animals. I con- ducted all censuses alone and timed my progress along the route by noting the time I reached each trail intersection. Trails also provided the spatial references necessary for precise estimates of the locations of hornbills detected. Routes were designed as rectangular circuits, returning to the starting point. Censuses were started between 0730 and 0800, weather permitting. Censuses were postponed for up to l h or canceled on rainy days. Two separate data-collection techniques were used simultaneously during censuses, one for determining cue production rates (for the “calling-rate index” [CR1]) and one for determining the relative abundance of hornbills. 25 Callin -R e Inde R Data for* calculating the calling-rate index: were collected during regular periods when I stopped along the census route to listen and record the number of individual hornbills heard calling. Throughout the length of each census, I listened for 15 sec. at 5 min. intervals, recorded the number of different individual hornbills heard calling, the number and types of calls, and whether or not the hornbill(s) heard was a hornbill not previously recorded that day during the 15 sec. listening scans. I decided whether or not the hornbill had already been recorded on the census based on its location, direction of travel, and sometimes visual identification. The calling-rate index (CRI) was then calculated for each census route by dividing the mean number of new individuals heard calling/scan by the mean number of all hornbills heard calling/scan - 1. Relative Abundances Based on a Total Detection Index Hhile traversing the census route, I plotted the location of each hornbill seen or heard directly on prepared maps and also onto a tally sheet. The combination of map and tally-sheet records facilitated the cross-checking and summary of information gathered at the end of the day’s census. Hornbills were represented as dots on the map, and arrows were drawn to indicate their direction of travel. Plots of hornbill locations were extremely useful for (1) identifying and keeping track of individuals during a census, (2) tracing changes in spatial distribution of hornbills over time and 26 between censuses, and (3) confirming the attachment of particular birds to particular parts of the study area. For each hornbill detected, I recorded the time, my location, estimated "perpendicular distance” from the census route to the hornbill, mode of detection, and direction of travel of the hornbill. This information was used to describe the relative abundance of hornbills. Additional information was gathered during censuses concerning age, sex, activities, and behaviors of hornbills observed and will be presented in other reports. Perpendicular-distance data were both visual and auditory estimates. 1 consider precision of these estimates to be good, based on experience gained before these censuses during the longitudinal studies and practice censuses mentioned earlier. During practice sessions and when possible during censuses, I actually paced out distances to hornbills after making original estimates. Estimates and paced distances were usually within 5%. Results Disc ssion Cunfigence in Census Results In all study areas censused, auditory cue detection represented by the calling-rate index (CRI) varied about as much on a daily basis as seasonally or among study areas (Fig. 2.5, Appendices 2-A, 2-B, 2-C). Data recorded during line-transect censuses evidently — reflected actual changes in hornbill numbers rather than a change in detectability of birds at different times of year. This is suggested by the absence of a relationship between the CRI (Fig. 27 538: m E. Sum—«3:3 an... Swan—:2:- cage: 32.9299 «(use 54%. 5:2. 2.5— 52:33 1 984 1983 Fig. 2.5.--Bycanistes subcyljuduicus calling rate indexes on three study areas 1n K1bale Forest, Uganda (1983-84) 28 2.5) and the relative abundance of hornbills (Fig. 2.6; Tables 2.2, 2.3, 2.4) censused. In fact, the CRI can be low when the relative abundance of hornbills is high. For example, during October- December 1983, the CRIs were low, while the number of hornbills contacted was high. Several authors (Kemp & Kemp 1975; Struhsaker 1975; Dates 1977; Marsh & Hilson 1981) have suggested that animals are generally more visible in logged than in unlogged forest. Results of censuses may, therefore, be biased in favor of logged areas when comparing relative abundance of hornbills between logged and unlogged fbrest types. Results of line-transect counts for hornbills in Kibale showed, however, that overall detection (auditory and visual combined) was roughly comparable among study areas (Fig. 2.7). There was one noticeable difference in cue detection between habitats. A higher percentage of detections at perpendicular distances greater than 250 m was made in the logged than in unlogged study areas (Fig. 2.7). This is probably because there were fewer hornbills to census in K-15 during most months, so I had more time to listen and look for individuals that were very far away. Since hornbills are such conspicuous, noisy birds, I am confident that few individuals went undetected within the 25 m perpendicular' distance to ‘the census trail (50 m strip width). Hornbill densities as determined from these data are likely close to albeit slightly less than the absolute densities because of 29 7L 5- _ E 5 5 GB 4— E a. 5 E 3' "‘ 13 2- ° . 2 1— E u 11 E El 5' 5 4- e a a V u E E s 2- g H ‘ 4'5 13 1- i ID a 3 o E 3' “I? O =. 5- ; E 1:: 4' 8 s 3- .3. a 2- g 2 4;, 4555 o .é‘ *Q Fig. 2.6.--Relative abundance of Bycanistes subcylindricus hornbills in three study areas during different stages of the breed- ing cycle (1983-84; Kibale Forest, Uganda). Census lengths are 5.34 km at Ngogo, 4.7 km at K-30, and 5.28 km at K-15. Data are based on the total detection index. 50 4O 30 20 IO 50 40 SO 20 IO Detentions (“/o] 50 4O 30 20 IO 3O lDGGED N45 uuuuuuu 0000000000000000 """""""""" ........................... ........... In. ‘ ............... ....................................... n = 236 UNLOGGED K-SO UNLOGGED NGOGO ............................ nnnnnnnnnnnnnn nnnnnnnnnnnnnn no. I ................... nnnnnnnnnn ooooooooooo ................ a o .......... oooooo ooooooooooooooo oooooooooooooooooooooooooooo ..... ......................................................................................... 50 150 Perpendicular Distance (m) Fig. 2.7.--Percentages of Bygeuistes subeyljndrjeus hornbills detected at various distances during censuses in three study areas of Kibale Forest, Uganda (1983-84). 31 individuals I may have missed. Most detections based on sight and sound (92% for Ngogo, 83% for K-30, 78% for K-15) were made, however, within a perpendicular distance of 200 m of the trail (400 m strip width) (Fig. 2.7). A comparison between relative abundance of hornbills using data from the 50 m strip width and 400 m strip width revealed similar trends (Fig. 2.8; Tables 2.2, 2.3, 2.4). Since the 400 m strip appeared to be representative of actual trends, I used this perpendicular strip width for most analyses involving relative hornbill densities since this allowed me to utilize more of the data. I believe that these data reflect actual differences in population density among study areas since patterns of abundance were consistent whether I used data based on a 50 m, 200 m, or 400 m census strip width. Relative Abundance of Hornbills in Different Habitats From detailed studies of nesting hornbills (Table 2.6), I estimated that in the 1983-84 breeding season there was a resident hornbill density of 11.2/km2 at Ngogo, 10.5/km2 at K-30, 5.5/km2 in the lightly to moderately logged K-14 study area, and 3.3/ka in heavily logged K-15. Independently, these results showed that undisturbed forest in Kibale provides better habitat for hornbills than heavily logged forest” Habitat use by. hornbills is more complicated, however, since these birds travel long distances between areas and habitats and become nomadic when not nesting. A clearer' picture of’ hornbill habitat use and spatial dispersion 32 .Aem-mmmpv awgbm msmcau au_3 5 ace 8 55.5 3:35am 3.529. some “5338. 398 3.88 .33..." 328 on; E cm a cmcuvz oucmucanm m>mumpmc cums “communes mmmcmimmmuwa .co mwcm> unavamgimcosm can pmcommmm mcwzogm .mucmm: .ummgou wpmnwx :? mauwgc:VPama:m mmumwcmu m Co aucmucanm m>wpmpm¢11.w.m .mwu KdeNfiN m—.m.m.m m—.m.m.m H = «2:32 ._._< . . . «ma—559 .34 .535: .25an :39: 5.2::— X. La; N .D m ind m mm w 16.” a w a L . m _..5 “w. / H. mm 1:.m 33 .mmmamcmo me .o: u cu .mmwgoawgcmu mcpcmmgn saw: mppwncco; we .oz co woman summcmu acmuwmwmu .Amuap asaxv P_.aceo;\a comp Lo #3528; came so gamma amascwmn .A¢.~ .m.~ .~.~ mapaaev puamcaeb cream a om no woman agrmcmom Amm 1 av PHN.F ¢.m N.K Am.e v ¢.m Am.er m.NP USMWOd apw>cmz e_-¥ -- -- -- A..K v m.m -- -- uaamop .uos\a;mvs Amm 1 av an.o m.m o.o_ Am.m_v m.o_ A~.opv m.NF om-¥ commopca uAmN 1 EV F"¢.m o.mm ~.me Ao.e_v ~.~_ AN.omV m.- omomz commopca . am 1a . a 1» mpcmcwmmm on u Em gwmw a Hm cwuw a 52 so; omax so; «Ex m Ex so; mucmmmcmgh mummwmcmgh m Pracgo: Aux mmmsomm mucmuwmmm emu Amxv mp wacco: mwc< xuzum we ovumm .oz cam: .oz.cwmz acmuwmmm Lo .oz mmmeowm .oz :mmz .Aew-mmmpv macaw: .bmacou apan_¥ cw macmummmc op mucwmmcmcp mo ovum; new .mmmsomn .mpmmcmu magmgc:WPAun:m mmummcmuamiu.m.m epoch 34 became evident upon analysis of line-transect census results from different study areas. Based on a 400 m strip width, the number of hornbills per km of census was higher at Ngogo than at K-30. The heavily logged forest (K-lS) had fewest hornbills during most months (Tables 2.2, 2.3, 2.4; Fig. 2.8). Hhen minimum densities of hornbills were calculated using the 50 m wide strip, unlogged Ngogo supported more than twice as many individuals (27.8/km2) than either unlogged K-30 (12.9/km2) or logged K-15 (12.9/km2) (Tables 2.2, 2.3, 2.4, 2.6; Fig. 2.3). Ngogo, located near the center of the Forest Reserve, is not adjacent to logged areas. The K-30 study area, although itself undisturbed, is contiguous with logged forest and is only 2.5 km from the heavily logged K-15 study area. Hornbills move freely between unlogged and logged forest at Kanyawara. They probably did not travel the 11 km between Ngogo and Kanyawara often. It is not known whether the lower density of hornbills at Kanyawara was a result of the effects of logging or natural differences in vegetation type, or both. Seasonal use of logged versus unlogged habitats at Kanyawara does indicate, however, that although hornbills persisted in logged areas, they preferred not to breed there or could not breed there because of fewer nest sites. Seasunal Trends insHebitet_use Seasonal fluctuations in hornbill densities in different study areas were dramatic. So also was the pattern of habitat use in relation to the breeding cycle (Figs. 2.6 and 2.8). During "pre and 35 early breeding" months (July, Aug., Sept.), hornbills arrived at nest sites and claimed their territories. Once all nest sites were occupied by breeding pairs, hornbills which were unsuccessful at obtaining nest sites continued to traverse the area searching for nest cavities. These nonbreeding birds were "transients.” The ratio of transients to breeding pairs in a particular habitat may be an indication of the preference of that habitat for breeding. From July through Sept. 1983-84, the ratio of 3.4:1 for Ngogo (Table 2.6) indicated that unlogged Ngogo was the study area most preferred by hornbills for breeding. Kanyawara study areas had ratios of 0.5:1 (K-30) and 1.2:1 (K-15), which were considerably less (Table 2.6). An estimate of' the density' of resident hornbills in each area indicated that unlogged habitats were several-fold higher than logged (Table 2.6). Although unlogged forest attracted more hornbills for breeding, a better indicator of habitat quality for breeding was to measure the nesting success of hornbills among habitats. Nesting success, which was consistently highest at Ngogo, is discussed in detail in Chapter III. Although logged habitat (K-lS) was not preferred for nesting, it did temporarily attract many hornbills during some other times of the year, as shown by the influx of birds in Feb. 1984 (Fig. 2.6). Large numbers of adult, subadult, and fledgling hornbills flocked to K-15 immediately following the breeding season in Feb. 1984 to forage in mass-fruiting Ejeus geuei and Eieus uetalensjs trees. Hornbills flocked and became nomadic during the nonbreeding season, causing the reduction in numbers of hornbills detected on 36 censuses in all study areas during the months of March through May (Fig. 2.6). Hornbill pairs, which were somewhat homogeneously spaced at nest sites during prebreeding (July-Aug.) and breeding months (Sept.-Feb.), congregated at super-abundant fruit resources during the dry, nonbreeding season. Since fruiting trees were patchily distributed, hornbills became scarce in some parts of the forest and common in others. A good example of the seasonal change in spatial distribution is illustrated in Figs. 2.9 and 2.10. During an August 1983 census in K-15, only three hornbill pairs were detected. These birds were active around potential nest sites. During a Feb. 1984 census in K-15, 45 hornbills (including adults, subadults, and fledglings) were observed flying around and foraging in fruiting trees. QueuelltHornbill Numbers in Kibele Using the mean overall density of 12.9 hornbills per km2 at Kanyawara (Table 2.6), and approximately 336 km2 (60% of 560 kmz; Hing 8 Buss 1970) of forested land (disturbed or undisturbed) in Kibale, there are an estimated 4,300 black-and-white casqued hornbills in Kibale Forest Reserve (Table 2.6). At a mean body weight of 1300 g (Kemp 1979), the estimated hornbill biomass is 5,630 kg. 0 i 11 Data from monthly line-transect censuses in Kibale show that hornbill movements, spatial dispersion, and habitat-use vary 37 K-15 100m Fig. 2.9.--Bycanistes subcylindricus hornbills recorded in K-15 during August 2, 1983, census. Dots indicate where individuals were encountered. All were active near potential nest sites. is trail grid. 38 100m Fig. 2.10.--Bycanistes subeylinguicus hornbills recorded in K-15 during February 10, 1984, census. Dots indicate where individuals were encountered. They were concentrated at mass-fruiting fig trees (Fjeus deuei at NB/GS; Eieus neteleusjs at 2E/lS). is trail grid. ' 39 seasonally. Because of this temporal variability in hornbill numbers, surveys meant to estimate hornbill densities should be repeated in the same area several times per year or should be based on nesting birds. Results from this first detailed survey of an African forest hornbill population are similar to findings from surveys of hornbills in logged versus unlogged forests of SE Asia (Kemp 8 Kemp 1975; Hilson 81 Johns 1982; Johns 1987). Like hornbills in Asia, black-and-white casqued hornbills were present in selectively logged areas, but in lower numbers than in primary forest in the core of the Reserve. Although long-term ecological studies are necessary to determine what factors limit hornbill numbers (Chapters II, III, and IV), census results alone indicate that primary forest is the most suitable habitat for these birds. Hith an estimated mean density of 12.9 birds/km2 at Kanyawara and 27.8 birds/km2 at Ngogo, black-and-white casqued hornbills in Kibale are at greater densities than any other forest hornbill species on record (i.e., Kemp & Kemp 1975; see Johns 1987 for 8 species). The ‘total population estimate of 4,300 hornbills in Kibale Reserve is, however, lower than the minimum viable population size of 10,000 recommended by Johns (1987) from his studies on eight hornbill species in Malaysia. Hornbill densities seem to be determined by the extent of logging in and around forest tracts. Therefore, they might be an indicator species to measure the level of disturbance in a tropical forest. The two-fold difference in hornbill densities in logged 40 versus unlogged areas of Kibale Forest suggests that this is true. Additional studies are needed, however, to determine habitat types and minimum forest patch size necessary for preventing local extinctions of hornbill populations. 41 REFERENCES Butynski, T. M. in prep. Comparative ecology of blue monkeys (Ceueopitheeus mitts) in high and low density subpopulations. Bolin, B. 1977. Changes of land biota and their importance for the carbon cycle. Beieuee 196:613-615. Collar, N. J. and S. N. Stuart. 1985. Threatened Bjuds uf Afrjea and Relateu Islauus. IUCN. Cambridge, U.K. Emlen, J. T. 1984. An observer-specific, full-season, strip-map method for censusing songbird communities. The Auk 101:730- 740. Ghiglieri, M. P. 1984. The chimpanzees uf Kibale Eorest. Columbia . Univ. Press. N.Y. Hamilton, A. C. 1984. Deforestation in Dgauda. Oxford Univ. Press. Nairobi, Kenya. IUCN. 1980. Save the rainforests. IDDN Bulletin 11(5). Gland, Switzerland. Johns, A. D. 1983. Ecological effects of selective logging in a Hest Malaysian rain forest. Ph.D. dissertation, Cambridge Univ. Cambridge, U.K. Johns, A. D. 1985. Selective logging and wildlife conservation in tropical rainforest: Problems and recommendations. Bjul, Donserv. 31:355-375. Johns, A. D. 1987. The use of primary and selectively logged rain- forest by Malaysian hornbills (Bucerotidae) and implications for their conservation. Bjol, Dunserv. 40:179-190. Kalina, J. and T. M. Butynski in prep. Seed dispersal ecology of Injehilja selengiga in Kibale Forest, Uganda. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. Ljviug Bird 17:105-136. Kemp, A. C. 1985. IDBP Bornujl! Sues. Ga, Baum. 5. Kemp, A. C. and M. I. Kemp. 1975. Report on a study of hornbills in Sarawak, with comments on their conservation. Kuala Lumpur, Horld Hildlife Fund Malaysia. (unpublished report.) 42 Kingston, B. 1967. H k n P n or b wara r Eurest Reseuues. Uganda Forests Department, Entebbe, Uganda. Leighton, M. 1982. Fruit resources and patterns of feeding, spac- ing, and grouping among sympatric Bornean hornbills. Ph.D. dissertation. Univ. Calif., Davis. Marsh, C. H. and H. L. Hilson. 1981. A survey of primates in peninsular Malaysian forest. Final report for the Malaysian primate research programme. Univ. Kebangsaan, Malaysia and Cambridge Univ., Cambridge, U.K. ‘ Myers. N. 1984. WWW Eutuue. H. H. Norton, N.Y. Oates, J. F. 1977. The guereza and man: How man has affected the distribution and abundance of Dolubus guereza and other black colobus monkeys. pp. 419-467 in Prince Rainer and G. H. Bourne (eds.), Eujmate Donservatjgn. Academic Press. London and N.Y. Skorupa, J. P. 1985. Responses of rainforest primates to selective logging in Kibale Forest, Uganda: A summary report. Skorupa, J. P. 1988. The effects of selective timber harvesting on rainforest primates in Kibale Forest, Uganda. Ph.D. disserta- tion, Univ. Calif., Davis. Sommer, A. 1976. Attempt at an assessment of the world’s tropical moist forests. Dnasylua 28:5-24. Struhsaker, T. T. 1972. Rainforest conservation in Africa. Erimates 13:103-109. Struhsaker, T. T. 1975. h e bus n . Univ. of Chicago Press. Chicago, Illinois. Struhsaker, T. T. 1981. Forest and primate conservation in East Africa. Afr, J, Eeul. 19:99-114. Struhsaker, T. T. 1987. Forestry issues and conservation in Uganda. Bjul, Buusery. 39:209-235. Hilson, H. L. and A. D. Johns. 1982. Diversity and abundance of selected animal species in undisturbed forest, selectively logged forest and plantations in East Kalimantan, Indonesia. B191. Bunserx. 24:205-218. Hing, L. D. and I. O. Buss. 1970. Elephants and forests. Bilglife Buuuguauhs: No. 19. The Hildlife Society, Hashington, D.C. 43 APPENDIX 2-A CALLING RATE INDEX (CRI) DATA COLLECTED DURING lS-SECOND SCANS AT 5-MINUTE INTERVALS DURING TRANSECTS CENSUSING HORNBILLS IN NGOGO, KIBALE FOREST, UGANDA (1983-84) Mean No. Mean No. New Census Date Hornbills Individuals CRI No. Recorded/Scan Recorded/Scan (1983) 1 Feb. 11 0.3 0.2 0.3 2 Feb. 24 0.4 0.3 0.3 3 Feb. 25 O O O 4 Feb. 26 0.1 0.1 0.5 5 Feb. 27 0.1 0.1 0.2 6 Feb. 28 O O O 7 Mar. 25 0 O O 8 Apr. 30 0 0 O 9 May 25 0.2 0.2 0.2 10 June 26 0.6 0.4 0.4 11 July 22 0.6 0.4 0.3 12 Aug. 16 0.3 0.3 0.2 13 Aug. 17 0.7 0.6 0.1 14 Aug. 19 0.6 0.3 0.5 15 Aug. 21 0.6 ? ? 16 Aug. 23 0.2 0.2 0 17 Sept. 30 0.6 0.5 0.1 18 Nov. 9 0.1 0.1 O 19 Nov. 30 0.2 0.2 O 20 Dec. 31 0.6 0.5 0.2 (1984) 21 Jan. 25 0.5 0.3 0.3 22 Feb. 19 O O O 23 Mar. 24 0 l 0.1 0 2 24 Apr. 30 0 O O 25 May 24 0.1 0.1 0 Total 3.6 'i_+_ SE 0.23.0.2 44 APPENDIX 2-B CALLING RATE INDEX (CRI) DATA COLLECTED DURING lS—SECOND SCANS AT 5-MINUTE INTERVALS DURING TRANSECTS CENSUSING T BDBBILINDBLBDB HORNBILLS IN K-30, KIBALE FOREST, UGANDA (1983-84) Mean No. Mean No. New Census Date Hornbills Individuals CRI No. Recorded/Scan Recorded/Scan (1983) .1 Mar. 5 0 0 0 2 Mar. 7 O 0 0 3 Mar. 10 0.1 0.1 0.6 4 Mar. 11 0.1 0.1 0.5 5 Mar. 12 0 O 0 6 Mar. 30 0 O 0 7 May 3 0.1 0.1 0.4 8 May 31 0 O 0 9 July 3 0.6 0.4 0.3 10 July 26 0.2 0.2 0.3 11 July 27 0.5 0.4 0.2 12 July 28 0.4 0.4 0.1 13 July 29 0.3 0.2 0.3 14 Aug. 1 0.1 0.1 O 15 Sept. 12 0.3 0.2 0.4 16 Oct. 3 0.1 0.1 0.2 17 Nov. 3 0.1 0.1 0.1 18 Dec. 10 0.2 0.2 0 (1984) 19 Feb. 11 0.4 0.2 0.5 20 Feb. 27 0.1 0.1 0 21 Mar. 11 0 O O 22 Apr. 2 0 0 0 23 May 2 0.2 0.1 0.3 Total 4 1 1T: SE 0.2102 45 APPENDIX 2-C CALLING RATE INDEX (CRI) DATA COLLECTED DURING lS-SECOND SCANS AT S-MINUTE INTERVALS DURING TRANSECTS CENSUSING BDBBXLLNDBLBBB HORNBILLS IN K-15, KIBALE FOREST, UGANDA (1983-84) Mean No. Mean No. New Census Date Hornbills Individuals CRI No. Recorded/Scan Recorded/Scan (1983) 1 Mar. 2 O 0 O 2 Mar. 3 0 O O 3 Mar. 6 O O O 4 Mar. 8 O 0 O 5 Mar. 9 O 0 0 6 Apr. 2 0.1 0.1 O 7 May 7 0.2 0.2 O 8 June 1 0.1 0.1 O 9 July 7 0.2 0.2 O 10 Aug. 2 0.1 0.1 O 11 Aug. 3 0.1 0.1 0 12 Aug. 4 0 0 O 13 Aug. 5 0.3 0.1 0.5 14 Aug. 8 0.2 0.1 0.3 15 Sept. 11 0.1 0.1 0.3 16 Oct. 5 0.2 0.2 0.2 17 Nov. 5 0.1 0.1 O 18 Dec. 11 0.2 0.2 0.1 (1984) 19 Feb. 10 0.5 0.4a 0.33 20 Feb. 22 0.4 0.4 O 21 Mar. 14 0.4 0.3 O 2 22 Apr. 3 O 0 0 23 May 3 0 O 0 Total 1.8 X 1 SE 0.110.1 aApproximate value. CHAPTER III DIET 0F mm W. WITH ENPHASIS ON THE NESTING SEASON 1113mm Previous studies on the food habits of hornbills, family Bucerotidae, have focused on species inhabiting an African savannah (Kemp 1976a; Kemp 81 Kemp 1978) and SE Asian rainforest (Leighton 1982). There has been no detailed account of the diet for any of the 12 hornbill species which occur in African rainforests (Kemp 1979). This paper describes the diet of the black-and-white casqued hornbill (Byeanistes subcylindrieus) in Kibale Forest, Uganda. Knowledge of the diet is a prerequisite for understanding the relationship between hornbills and ‘their‘ environment. “The availability of food has a strong influence. on the timing of hornbill breeding (Kemp 1976b; Chapter IV), reproductive success (Chapters IV & V), ranging patterns, and social organization (Leighton 1986). Since hornbills mediate seed-dispersal for rainforest trees (Leighton 1982; Kalina in prep; Kalina 8 Butynski (b) in prep), particular attention to dietary needs of these birds and their influences on the regeneration of this fragile and. threatened ecosystem is needed. 46 47 Methods In 1981, I began a long-term study of the black-and-white casqued hornbill in Kibale Forest, western Uganda. The study area consists of both primary forest and selectively logged tropical rain forest (Chapter I). Data used in this paper were collected during over 4000 h of field work in Kibale Forest from 1981-1984 and during yearly working-visits since 1986. Methods used to identify foods in the diet of B. suDeylindrieus included: 1. opportunistic observations, inspections of the ground below nests,. ”nest-watches,” wa "nest-trap” collections. Dpuortunistic Observations Opportunistic observations on food habits were recorded while I was following hornbills, watching fruiting trees, or conducting censuses, tree phenology, and other work in the forest. Some food items were first found during sporadic inspections of the ground below active nests. NeSL-Hatshfl Since 1981, 45 B. Wen nests have been found. All were located in natural cavities of large trees 8-27 m above ground. Much information on the diet of B. subcylinuujeus was gained by quietly sitting and watching these nests. "Nest-watch" data used in this paper were collected over the 1983-1984 nesting season during 4 48 h (n - 28) and 12 h (n - 28) sampling periods (700-1100 h and 700- 1900 h). Observations were made from locations 20-60 m from the base of the nest tree using a telescope or 10 x 408 Leitz binoculars. Only information at sealed, occupied nests was analyzed. Thirteen nests were observed for a total of 447 h (Table 3.1). I only watched nests with clearly visible entrances so that birds landing on the nest could be identified and the food items fed into the nest distinguished. The male hornbill’s arrival and departure times at the nest, and the kinds and numbers of items fed into the nest, were recorded. Hornbills are multiple-prey loaders; i.e., more than one food item may be collected before a load is delivered to the nest. Male hornbills regurgitate food items into their bill tips just before they feed them to the female in the nest. These food items are usually easy to see and identify. Each species of fruit was identified and given a rating of 1-4 based on size, with 1 being the smallest and 4 the largest. Examples of species (with approximate mean diameter of the fruit) in each category are: Deltis Durandii (0.5 cm) . l, Djosuyras abyssiujea (1.0 cm) - 2, Mimusous bagshawej (2.0 cm) - 3, Eieus gauei (3.0 cm) - 4. Size ranges (diameter) of fruits for each category are as follows: 1 - (0.3-0.7 cm), 2 - (0.8-1.5 cm), 3 - (1.6-2.5 cm), 4 - 2.6-3.5 .cm). Volumes were calculated for each category, based on the medians: 1 - 0.07 cm3, 2 - 0.52 cm3, 3 -= 4.19 cm3, 4 . 14.14 cm3. These values were derived from the diameters and volumes of samples of fruits. 49 Invertebrates were identified or described and also rated into these same four size categories. W Traps were placed below nests to catch fruits, seeds, insect remains, and other materials dropped by hornbills. These “nest- traps” consisted of burlap fabric suspended 1 m above the ground on a square frame. Each trap had a collecting area of 1 square metre. Rainwater was able to drain well through the small holes in the cloth, while all fruits and debris which fell into the trap remained there. The burlap was replaced as needed. Trap contents were collected, labeled, and sun-dried about every 6th day (Table 3.1). Traps were usually set in place before the hornbill entered the nest cavity. In most cases, the entire nest period was sampled in this way. Nesting hornbills eject their feces and throw debris out the nest hole, so that the cavity remains clean. The percentage of the discarded nest debris caught in the trap varied between nests. The amount of debris falling in the trap depended on foliage cover, nest height, and angle of the nest opening. Because of these variables, the purpose of setting nest-traps was to: (1) identify rare food items and those items usually difficult to see, (2) see how specific food items were "processed” (e.g., defecated or regurgitated), (3) get relative indexes of quantities of various food items fed throughout the nest-cycle, and (4) collect "processed" seeds for germination experiments. 50 Pew mp ~.oee mp Peueh 1 1 eemoep o.—+e.o AN ~ m.em+e.em o.mop N »p_>~e: mp-x eemmep 1 1 apeaegeees ~.P+~.m mp p m.-+~.o~ m.o¢ N -ugmpe .ep-x e.~He.e as . m o Ho.. 9.5 . eaeeopce em-g e._ue.e amp op e.P¢He.em ¢.¢a~ e eaeeapce omen: omueeez wee?» munch omfieeez meguue: eeguue: «exp eeg< meepueeppeu -oeppeu saw: mumez umez Lee -umez we name: umesem xeaam eeexuem name we .ez me .ez muse: acne: we .ez .Aem-mmmpv meeem: .umeceu ep~e_x :P meespmec eepaeeppeu eee meewaeuep eega-umme ecu :eue3-umee mmMHuuqHwammw wumwwmmuxm--.p.m epeep 51 Once trap contents. were dried, the materials were sorted, counted, weighed, identified, and recorded. Volumes of fruits, seeds, and insect parts were measured by displacement in a graduated cylinder, or were calculated from shape and measures of length and diameters. All insect remains were taken to the National Museums of Kenya for identification by John Njoroge. Nutritional Analysis Sun-dried samples of fruits eaten by hornbills were sent to the Range Science Laboratory at Colorado State University, or to Dr. Peter Haterman at the University of Strathelyde, for chemical analysis in order to determine nutritional content. Analyses were carried out on a variety of fruit parts, but only results concerning ripe, fleshy food parts are presented here. Fruits were analyzed for calories per gram, percentage total available carbohydrates (or nonstructural carbohydrates), percentage lipids cn' fat, percentage protein, and total phenolics, including the presence of condensed tannins. Fruit Burpholugy The color, shape, numbers of seeds/fruit, and characteristics of the fleshy fruit parts were recorded for all fruit species. Length, maximum and minimum cross-sectional diameters of seeds, and of seeds plus surrounding flesh, were measured for a sample of fruits of each species. Fruits were grouped into the four morphological types (capsular, drupaceous, husked, and fig) as described by Leighton (1982). "Capsular fruits" are dry, dehiscent, 52 and many-seeded. They are often conspicuously colored and contain oily, arillate seeds. 'Drupaceous fruits" are more or less fleshy, one-celled fruits with one or more seeds. Most of these fruits may be plucked by hornbills and swallowed whole. 'Husked fruits" have a thick, woody pericarp that requires extensive processing. "Figs” (Ejeus spp.) are soft, fleshy fruits with many small seeds. Figs lack a hard covering. R 5 1t nd i ' Eouds and Furaging Fruits and other items were observed to be eaten by B_. subeyljndrjeus in Kibale Forest, Uganda (Appendix 3-A). Foods recorded for this species in south-central Uganda by Kilham (1956) and others are also included in order to make a complete listing of known foods for this species. B. subeyljnuujeus eat a minimum of 67 fruit species from 26 plant families, plus lichens, fungi, mosses, mammals, birds, reptiles, mollusks, and arthropods. Unlike the SE Asian forest hornbills described by Leighton (1982), B. subeyliudrjeus select fruits which are diverse in color and morphological type, with sizes ranging in diameter from the 5 111m Deltis afrieaua to 60 mm Ejeus uaeusu. Also, Be subeylinarjeus eat the same species of fruits consumed by sympatric primates (Struhsaker 1978; Kalina & Butynski (a) in prep). Fruits may have. single large stones, or they may be many-seeded (Appendix 3-B). Like the SE Asian hornbills, B. suBeylingrjeus apparently takes almost any animal prey which it can find, catch, and swallow. 53 Proportions of fruit in the diet varied from year to year and from season to season. During the 1987-88 breeding season, for example, Rauvolfia uyyuhylla fruits comprised most of the diet at some nests. Yet this species accounted for less than 1% of the fruits fed into nests during 1983-84. Also, fruits of [:jehjlja splengjua were not available during the 1983-84 breeding season, although they were highly preferred during the prebreeding months of August and September. B. suBeylindrieus are most frequently observed foraging in the rainforest canopy, hopping or fluttering from branch to branch, leaning forward to pluck a fruit with the tip of the bill. Then, with a backward flick of the head, the bird swallows the fruit. The hornbill’s long, laterally flattened bill is well adapted for frugivory. The long bill extends the bird’s reach, so that a greater number of fruits are available within the radius of the perched bird. Unlike the sympatric crowned hornbills (leekus l t rmi us), which hover above trees and sally in to pluck fruits while remaining in flight, B. subcylindrjcus generally pick fruits while perched. Hornbills are particularly adept at manipulating capsulate fruits. These fruits are often conspicuous, colorful, and attractive to hornbills, which are able to peck and pry them apart at the seams. In this way they get at the edible part of the fruit before less-agile competitors, which must wait for the tough capsules to dehisce. Unlike barbets, starlings, and other small birds which sometimes act as ”fruit-robbers” that tear off 54 fruit pieces and leave seeds to drop below the parent tree, hornbills are high-quality seed dispersers for most trees they forage in (Kalina in prep; Kalina & Butynski (b) in prep). Hornbills swallow large seeds that other birds and mamals cannot handle, ”process" them in the crop, and then regurgitate them cleaned, but intact and viable. Small seeds (< 0.8 cm) pass through the gut and are usually defecated intact (Appendix 3-B). Insects swallowed whole are sometimes alive when they are later regurgitated into the nest, as was the case of one katydid (Euryeuryuha sp.) which I found that had been dropped below a nest. The hornbill beak is strong enough to open thick, leathery fruit capsules, but it is also a sensitive organ used to test the ripeness and quality of fruit before ingestion. Hornbills visually examine fruits, then squeeze and toss them in their bill tips before swallowing good ones or dropping those they reject. Hornbills prefer ripe, insect-free fruits. B. subcylindrieus sometimes forages in lianas (e.g., Phytolaceaceae) or bushy, forest-edge tree's (e.g., Bridelia mierautha). Hornbills also cling vertically to branches and peel lichens from the bark of trees. Mosses and lichens are sometimes passed out in feces. Fungus is fed into nests by male hornbills, and presumably is also ingested by the imprisoned female. Hith repeated, downward thrusts of the bill, hornbills cleave pieces of bark from wood and then use the bill as a lever to remove the bark. Bark and wood strips collected in this way are carried and tossed about by males during displays (Chapters IV, V, 8 VI) and 55 are used for nest lining. Bark peeling and the pecking of rotten wood also expose grubs and other insect prey, which are then crushed and eaten. Hornbills similarly sort through rotting seeds in old, dehisced fleuuguta mflBLaegjl capsules, apparently in search of prey. I have never observed hornbills foraging for mantid oothecae, but these items are often delivered to nests. These well- camouflaged oothecae must be highly sought after despite the gall- like appearance and tough exterior. B. subeyliadrjeus hunts for a wide variety of vertebrate and invertebrate prey (Appendix 3-A). Eggs ranging in size from those of the small mousebird (Kilham 1956) to large chicken-sized eggs like that. of' a guinea fowl (pers. obs.) are sought after and devoured. Lizard eggs are removed from the ground. I observed six of them being fed into a nest, along with the adult red and black skink (3j_Qp_a fernandj). Nestling and adult sunbirds were once regurgitated by a male hornbill into a nest. Hornbills, traveling alone or in flocks, raid weaver bird colonies and eat the eggs, nestlings, and any adults they can catch (J. Corner, pers. comm.; pers. obs.). Egyptian rousette bats (Buusettus_ egyutteus) are sometimes caught in flight (J. Baranga. pers. comm.). I suspected that a roost was raided when one male hornbill delivered several Eutesjgus gauensls bats to a nest. I identified one that it dropped. Captive hornbills feed on rodents (Kilham 1956; pers. obs.), swallowing them whole. Most mammal fur which was dropped by hornbills into traps below nests probably came from rodents. 56 The remains of a lesser galago (Galagu senegaleusis) were found in one nest-trap. The location of this trap was such that the animal could only have fallen into it from the edge of the hornbill nest. The skull bones were crushed in the characteristic way that B. subeyljnurjeus kill their prey. The contents of the skull were gone, but the body was intact. I suspect that the animal was too large to fit into the narrow entrance to the nest. It was certainly too large to swallow whole. The kinds of invertebrates eaten by hornbills are varied, with prey items ranging in size from the fly or bee to the goliath beetle or 6-inch-long millipede. Foraging for. invertebrates is accomplished in many different ways. Hornbills perform aerial hawking and swooping maneuvers to catch alate termites. Flies and bees are snapped from the air as the insects fly by the hornbill’s head. The presence of millipedes, lizard eggs, and other food items suggests that they regularly forage on or near the ground. B. SUD; eylinuuieus hunts for terrestrial prey by scanning the ground from the perch of a low limb. B. subeyljudrieus is omnivorous, but feeds primarily on fruit. Ninety-one percent of the volume of the 13,235 edible items delivered by males at nests (during 1983-84 nest-watches) were fruits; 9% were invertebrates. The remainder was comprised of 16 birds or mamals, 17 eggs, 28 pieces of lichen, and 3 pieces of moss. In addition, 61 pieces of bark were delivered for nesting material. 57 Nest-Hatgh Data Information collected during nest-watches (Tables 3.2 and 3.3) was preferable to nest-trap data for examining relative proportions of the diet. During nest-watches, the exact number and sequence of all items fed into the nest could be recorded. Soft food items such as fungi, papaya (CaLtea uauaya), insect larvae, and grubs were accounted for in nest-watch data, whereas these items were absent from nest-traps. On the basis of volume, invertebrates made up 9% of the B. 522; eyliuuuieus diet. They comprised a greater percentage of the diet for hornbills in the heavily logged study area (15% in K-15) than in unlogged forest (9% in Ngogo) or light-moderately logged forest (5% in K-14). The largest number of insects fed upon were in size categories 1 (509 items) and 2 (272 cases). The greatest total volume of insects was, however, in size category 4, which represented 72% of the total volume of all invertebrates delivered to nests (Table 3.2). Figs (fieus spp.) were the most prominent fruit in the diet, comprising 58% by volume of all fruits fed. A greater volume of figs was eaten in light-moderately logged forest (80% of all fruits in K-l4) than in heavily logged forest (54% in K-15) or unlogged forest (52% in Ngogo). Eteus spp. fruits were sometimes difficult to identify during nest-watches and were, therefore, counted, desdribed, and recorded according to size. Eleus spp. were represented in all size 58 AN V mgeguo Aw V madden muwumw Am V Hfluemumu «Huang Am V “Vauwuflmamm mafia AmpV meuaceeuce>eV AePV mmflefiwwxee mauxewefla .eeV .aam “flea“ Hfluemume wflqamq m»mmma.muflumu mepecemece>eV V we a wquaqwmwa .eem wauflm AAAAA an an to M r— vvvvv .N V meagee AP V Hfleemume wflmamu A“ V Hummewmum weewmafle Am V meuegeeuce>eV .emV mmfleflwwxae «muxqwqfle AKSV .nam «mafia Aese.V . =V mp-x Amm~.~ . cV .p-g A-¢.e . =V oeeez AeV enemas. eee ma.=sm we AmsuV esepe> page» R an Leege gee: .m «Lagee meuegeeuge>=~ .aam meme“ mewwMHmHum mwal eufleflwmamm weuxewefla Hfieemuue F ASVK.M . =V mp-¥ AAAAAA N 5‘ 0’1 N N (D VVVVVV o i V meagee V meuegeeuge>eV V .eem wmwflm V Hfluemume wflwflmu V mwflqflwwxmm wuuxmwuwa Amm~.V . =V SV-¥ mgezuo Hummnwmmm.wmmwmaflm empeceeuce>eV Hfieemume wamauu .eem wmuflm vvvvvv NQMQI—M g VeV AAAAAA “Fm“.e . eV oeomz AeV maeaV uuemew eee “Vegm we geese: pane» a an geese xeam .< .Aem-mmmpV «enema .pmegeu epaer e. megeue3-amee mevgee name: «mu uwuuqflaxummw wumwwqmwau an ewe meanseeage>=V eee mu_=cm Lem geese xee¢-1.~.m «Fee» 59 mepumee meee ece guerpew .meeeeV_F_s .mevuces .megeeem mevuees .mepueee .mgeeeezmmegm .mceppvegmyeu .mevexpex .mmceeem meuwsgep .mseppveceuee .meee .mewpu meme: .meuVELeu .mevpu mezee meek; .mweepacwege mauvu eeueeegevs neweeeemeeeeme .Vezeewwee meemesrz mvmcepeuecimeuwu..Pvcemywzpez were .eercpmmxee «ephemera memUVLea “FeVae .VVecaeae “rupee e m N p e .m 1N p "use: xgemeuee eNVm some e? wereeem gene: pm mop mp ep Pm pep we op we eme e mm m.m1o.N m.N1o.p m.—1m.c N.oum.o a no N up as wNF Nm neN ..o .. , .,. mm.. P pmm 0N F mp mm c on FA Nos mN m m mm m nm em ewwN No N ewe eoN NN mm 1 . pp eem mNm cup N wN mo Nw opop ppmN Fm Np p eNmp Nee mm mp om mN cpm mmw mam ep me P mppp meow eop m an Np NmN open mwep m.m1m.N m.N1m.F m.—1w.c 5.91m.o mNVm eNFm eNVm eNPm ”muce>em eNVm eNFm eNVm muvm "mppecu esspe> a AmEeV esepe> new»? a use: .ez mpug asepe> u AmseV esepe> new». a use: . z epix eE=Pe> a AmEeV eE=Pe> mama? a use»? . e e z AEeV mazes «Npm ep.ep mp.e Nm.c No.o e~.ep mp.e Nm.o No.e AmsoV esepe> e m N P e m N — aceuoaeu e~pm xsemeaeu eNVm meuegeeuge>eV moves; eovmwuceew apesemmeee: we: age: meuegnepce>cm eee waves; .meweeem e» .Neeeaeeu aN.m Na .Aee-mem_V mesuaeziumee mevgee meme: meevce=VFAeeem meucheexm ewe. ewe meuegeeace>=r eee mavegmuu.m.m epeeh 60 categories of fruits. The following analysis of the relative importance of the various fieus spp. in the diet of B. suBeylindri- eus is incomplete, therefore, since it only accounts for those species that could readily be identified. It does, however, give some indication of variation in number and volume of fiteus spp. being delivered to the nest. Overall, for all study areas combined, Ejeus uatalensis ranked first with 44% of all the fieus fruits fed. Ejeus dauei ranked first for volume (14% of all figs fed). [Lotus Brachyleuis ranked first (13% of all figs fed) in K-15, but was ranked third (1% of all figs fed; 5% by volume) in the analysis of all study areas combined. Papaya (B. uauaya) was the only cultivated fruit observed fed into nests. It was only recorded at nests in the selectively logged areas (K-14 and K-15). This fruit, which comprised 2-3% of the fruit fed at nests in these study areas, was delivered as large, amorphous chunks. Papaya accounted for 1% of the total volume of fruits fed for all study areas combined. Hornbills in the logged forest flew to nearby cultivated areas where this fruit could be found. Hornbills at Ngogo would have to fly 11 km for such access. Diasuyuus abyssiniea fruits ranked first in the number of fruit items fed into nests at Ngogo and K-14, and second for K-15 (Table 3.2). D. aByssiuiea fruits ranked second to fieus spp. by volume for all nests in all study areas. D. atyssiuiea is among the four most. common tree species in all study' areas .(Butynski in prep. Skorupa 1988), and it produces fruits throughout the nesting season (unpubl. data). 61 Beltis gurandii fruits comprised 57% of the fruit items fed into nests in K-15, but this was only 9% of the volume of all fruits fed. D. gurandji ranked in the top five fruit species fed by item, or by volume, in all study areas. It was less significant, however, in the diet of hornbills at Ngogo and K-l4 (1% by volume) than it was in K-15. Nest-Hateh Biae Categories (Table 3.3) For Ngogo and K-l4, the greatest number of fruit items was in size category 2, a reflection of the large number of D. abyssinica fruits fed into nests. The greatest number of fruits were in size category 1 for K-15, a reflection of the large quantities of B. Durandii fed into nests. fjeus spp., which made up most of the B. subcylindricus diet, occurred in all sizes. The number of invertebrate items per size category varied between study areas. The greatest volume was always in size cate- gory 4. Lepidoptera larvae constituted many of the items in the smaller categories, whereas larger mantids, beetles, and millipedes made up much of categories 3 and 4. NesL-ILaLDaLa Figs and soft foods such as larvae, flies, and fungus were often fed during nest-watches, but were under-represented in nest- traps (Tables 3.4 and 3.5). Although fig seeds were part of the fecal debris in traps, a relative estimate of the number of ficus 62 .38.. we «gee 3558.8 ego: “Be... meow... on msmum .239 go 23!. 3 =2» gegueg eueemeeeoeemege B eegemeee egez .emm 33... .N .2252 22. e5 .3 33.: eegege ege: use eeesmceu we: ege: .325. meegegx 32252 B eeueuewee ge eeueewogemeg 53 e2 .33? ._. "neuez e5 3 gene... u e: mm H euepeuVNe ewgee ee> u euepeew e .> :ege m u e newewmmmeeg meeem 32:33 a.» N “88 :223 .Ie. l N l I m 32ng 3 :5...» e Hm m N 32:.— me e me». e be m meeem . :egee by e newegw 3w: mmuee ho. op meeem newewmm as .o a meeem newewmm an .o N eeeem newcwmm an .o m meeem newewmm an .o Np meeem newcwmm an .e m— meuegeeuge>eV NN neuegeeuge>=~ oN meuegeeuge>=V up meuegeeuge>=~ Aepw meuegeeuge>eV GS mwgeee Zoe... :8 mwgeee Zoe... :2 mwgeee peuee 3K mwgeee page... :3 mwgeee peuew 2N6 u e m8.— - = 2n n = 3N; .. : «Nee a e AmsoV ess—e> —eueu we a he geege seam .u “N V mmeem mme .11 N meeem wexegm 2 km N utegw eeEmexde B m mmeem mmmmn AN meeem w? esmmmml .z n meeem wezemm mm.h: N meeem e meeem wwecegee .u N n meeem eee_:eV e up. n newegw 32:3. an .e e newegw n8: m N meeem 338% .u e «QwegeeegesLr N we eg eugeamw cw meeeg eugekmw o— meeee newcwmm an .o N neueg eugeemw N newegwmmmmflmmxm1.o em meeem newe—mm ee .9 Nm meeem newewmm an .o Np meuegeeuge>e_ N— meeem newewmm no .a Fe m g e meme AueeV mwgeee —euee .aNnV mwgeee peuew AaNe mwgeee peuew Aunm mwgeee peueg AaNe meeem euwcwmm an .o e3...” 1 : mama .. = N9 a : meg .. : ooN.N .. c 3 232. .38 we a 3 be: is. .e AN m_esw= N meeem wexesm no.4“ N meeem .eem weer» N neuea mama N muwagw newewmmxee .9 AN W c 38... 82. N 3:: :22: 1.1V N I. e muw=g5.eewcwmm an .o m neuegaeuge>:_ N ma.=gu.euwewmm an .o e meeem eye—eeg e .>._ meeem wweeegee .u e memegmeugeamw —— mp meeeg eugeeiw m meeeg eugeemw e meueg eugeamw 2 meeem cog: B. N mmV meeee. 833 me In N «Bee eercwmm me ab N muwegw 3.53 e 4.9 AuNo meeem newewmm an .a Anne meeem newewmm no .a AnemV meeem w—ecegee .u Aumm meeem wweeegee u ANNN meeem newewmm an a egg: .. e BKN - e 5m 1 e NNm.N u e 8o.m u : mseuw we geese: peace we a an geege seem .< $38. 3:85 9&8. 62.128: .8825 Sewn—5V 8588 :< 2 V. 2 3. one. o 9. Se: Segatag :53 .eeam: ege «N1 A acwecemegeeg 253 3.5 .meelw 5 eeeew 2.2.2 :e we eafleeugee we geege cw eexeeg £35130: segw eeuuefieu mewsfig eeew neuwgeewpmueem Sufi—.3311. e. m e32. (53 .ueeew eeewweeeeu ge.wmme>em mauceeeeeue: NmN . esmeeecue .ecwmggeepmen N—N . .eeemw1eeeweegeo ge wppeg eNNeeweem "mp . ecwzggepegu ge ..em mmsecueuqeum .wupeeeen mwsecueueeuem .eewcggegegewn NN . .ueecw ANwEew eeewaegeu go .wa—Nee megeeese: ..em maneuemez .weme>em mauceeeeeue: N— . "mewueem Nege>em eeepeew meeegeeuge>ew we mewgeoeueue £2232 as .8: $3 :3 a .38. NN mg .ez N _N .ez N eeeewppwz AN eemvgeuepm AN w Neme>em meueeeemeue: N Neme>em neuceeeeeue: AN mN .ez N —N .ez m eeeemV N eemmuem: N eeewéeg Am . .ueecm éem eempmegeu N .em memueg_mm N .ee memue—_eu Am wNNeg eNNeuNeeu e wNN—es megeeosez. N .em we ceu e mm .m: m ewe—eegeu N .em mapeeo omega m .eem eeeccuee m eemppeeueu e eemwuem: N mN .ez e eeewweeueu m .3 e_ _eewmew e .9. me Some: m eeewE o 353:». eewmggecmmn m .em a: eueme: N eeewweeueu o eeeeeeem eewgggeeweee o .em memueNme N e ea Np eeeewN—wx N eeewweeueu N wzegge eeeeseee e womasam maucoeo cue: N—V mN .02 eg _ .oz N . .oz 2 g .oz ANNW _ .ez .N N .9. 2 e835: ANNN N .ez NucN N .ez NucN eeeewppw: Anew eN .ez .. .. .Nwfifi eege>eu mNN No ooN Nam mane Nee: we .ez News. a .eee1NN .ueo Np .eee1NN .ueo e— .eewue— .uuo m. .eee1m .ueo eege>eu mean: mp1x ep1x cmnx omen: eeg< meeum .eeuewN ego «N m eewuceeegeeg mewgeoeuee N—eo . ew1mempV mecca: .umegee epeewx .eeegu1umee neuwgecwpaueem meamwceeAN cw neuegaeuge>ew gepeuwagee we eczemae ge eucemeg we Auceaeegw gew eaeueeugee he geege sce¢11.m.m eNeew 64 fruits fed into the nest. was difficult to determine from such samples. The nest-trap method compares favorably with the nest-watch method for other fruit species, particularly ones with large, hard seeds. For example, as is the case for nest-watch results, D. abyssjnjea, B. Mum, and invertebrates represent the top- ranking items in the diet based on nest-trap analysis. Traps are preferable to nest-watches in that less effort must be extended by the researcher for the collection of a large amount of information. Also, foods unknown to the observer during nest- watches can often be collected and identified from traps. During this study, seeds of 20 species were collected in traps and remain unidentified. The seeds are catalogued, however, for possible future reference. They must have come from rare trees or lianas. The nest-trap method is particularly good for the identification of invertebrate remains. Most invertebrate species (Appendix 3-A) were taken from traps. Invertebrates from traps had been defecated, regurgitated, or crushed and dropped by hornbills. Records of the presence or absence of invertebrate categories during trap collections (samples of the diet throughout the 6-month nesting season) can be used to map seasonal changes in abundance of particular species in the diet. It must be remembered, however, that large-bodied, keratinacious Species such as the scarabs are_ perhaps over-represented Since their body parts are more difficult to digest than other species. Scarabeidae are, nevertheless, frequent prey items. These insects are probably very obvious 65 because of their large size and the noisy, buzzing sound of their flight. This family of insects is also extremely diverse and numerous in the Ugandan forests (M. Ritchie, pers. comm.). From the presence or absence data collected, differences between study areas were evident in relative amounts of various invertebrates in the diet of hornbills (Table 3.5). Beetles and mantids, for example, ranked highest at the Ngogo study area, and millipedes were represented in only 2% of the trap collections there. Yet in all the Kanyawara study areas, millipedes comprised 26% of the collections for K-30, 11% for K-l4, and 7% for K-15. Nutrition Based on chemical analysis of selected fruit species (Appendices 3-C, 3-D), hornbills primarily ate ripe fruit items that were relatively low in toxins (phenolics). The chemical composition of fruits in the diet was variable. Arils from capsular Bljgm unijugata and Trichilia sulendida fruits were very high in lipids compared to the pulp of D. abyssiniea and Pseudospgndias mieruearpa drupes. 1D. abyssinica. was high in nonstructural carbohydrates. _i;'_i_c_us spp. were, for a fruit, relatively high in protein (Janzen 1979). The nutritional breakdown of these wild fruits should be informative for aviculturists attempting to breed in captivity birds that are primarily frugivorous in the wild. Hild fruit species have a relatively high protein content for which there is no domestic fruit substitute (M. Underwood, pers. comm.; Hatt & Merrill 1977). 66 Invertebrates were not analyzed nutritionally, since prior research has shown that various insect tissues are similar in terms of caloric value (Slobodkin 1961). Insects and other animals were undoubtedly' important in their' contribution of protein. Other dietary components, such as lichen, may have had minerals that, in addition to calories, were significant to the diet. Fruit species differing in nutritional composition may have been of varying degrees of significance at different stages of the hornbill breeding cycle. ,1. splendiga fruits, for example, comprised much or most of the hornbill diet during prebreeding months. A diet of I. splendiga fruits alone would be too low in protein to sustain hornbill nestlings (M. Underwood, pers. comm.). but the high energy content in these fruits could help bring hornbills into breeding condition. 1. sgleugiga is higher in energy than other fruits, and that energy may be more readily digestible because it is in the form of lipids (M. Underwood, pers. comm.). Perhaps a diet of I. splendiga fruits allows hornbills to store fat on their bodies, helping them to meet energetic demands during the long breeding season. Eieus spp. fruits (figs) comprised much of the diet during the breeding season. Figs, unlike I. sulengjga fruits, are high in protein (Appendix 3-C). At 21% crude protein, figs would provide close to the 25% protein diet that is probably necessary for raising hornbill young (M. Underwood, pers. comm.): Frugivorous, tropical cavity-nesting toucans (family Ramphastidae) need 25% protein in the diet for their nestlings to thrive in captivity (Iwinski 81 67 Iwinski 1984; Svoboda 1988). Since figus spp. and D. aByssjujea fruits comprised approximately 80% of the hornbill diet during the nesting period, further analyses of D. abyssinica should be conducted to determine if these fruits are also high in protein. If total caloric intake is to be determined, the energy content of Eieus spp. must also be ascertained. mansions Four methods were used to study the diet of B. subcyljudrieus, an African rainforest hornbill. Most quantitative data were collected at nests during one breeding season. Although B. s_u_b_- cylinduieus ate a wide variety of fruits, fieus spp. and D. aDys; sluiea fruits accounted for most of the diet. Proportions of fruits represented in the diet were found to vary from year to year, and from season to season. The differences in nutritional values of foods, the energy expended in obtaining the foods, and the degree of assimilation of various nutrients have not been addressed in this study. Volumes, weights, and frequencies of food items are given, however, as approximations of their significance in the diet. Further nutritional analyses of fruits eaten by hornbills are recommended. 68 REFERENCES Butynski, T. M. in prep. Comparative ecology of the blue monkey (Bereuuithecus mutis) in high and low density subpopulations. Eggeling, H. J. and I. R. Dale. 1952. Iue_1ug1geuuus_1uees_ef_the Dgauga_£uuteetuuate. 2nd edition. Government Printer, Entebbe. Hamilton, A. C. 1981. A field Guide tu Dganda Fgrest luees. Makerere University Printery, Kampala. Hubbard, C. E., E. Milne-Redhead, R. M. Polhill and H. B. Turrill, eds. 1952. Flor of Tr 'cal st r' . Crown Agents, London. Iwinski, D. and B. Iwinski. 1984. Too many toucans. The A.F.A. Hat hbi X6:18-20. Janzen, D. H. 1979. How to be a fig. Ann. Rev. Ecol. Syst. 10:13-52. Kalina, J. in prep. Effects of animal dispersers on germination success of five tree species in Kibale Forest, Uganda. Kalina, J. and T. M. Butynski (a) in prep. Comparative feeding ecology of sympatric hornbills and primates in Kibale Forest, Uganda. Kalina, J. and T. M. Butynski (b) in prep. Seed dispersal ecology of lrichilia sulendida in Kibale Forest, Uganda. Kemp, A. C. 1976a. A study of the ecology. behaviour and systemat- ics of Tockus hornbills (Aves: Bucerotidae). Iransv. Bus. Mem. ZD. Kemp, A. C. 1976b. Factors affecting the onset of breeding in African hornbills. Bree, 15th Int, Drn. Cung[., pp. 248-257. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. Ljying Bjud 17:105-136. Kemp, A. C. and M. I. Kemp. 1978. Bucorvus and Sagittarius: Two modes of terrestrial predation. Pr S . fr. tor HEELS, pp° 13'16' 69 Kilham, L. 1956. Breeding and other habits of casqued hornbills (W W) SojtstjsEVLollns 131(9): Leighton, M. 1982. Fruit resources and patterns of feeding, spac- ing, and grouping among sympatric Bornean hornbills. Ph.D. dissertation, Univ. Calif, Davis. Leighton, M. 1986. Hornbill social dispersion: Variations on a monogamous theme, pp. 108-130. In Eeplpgieal Aspects pf Boeial Evolution, eds. D. I. Rubenstein and R. H. Hrangham. Princeton Univ. Press. Skorupa, J. P. 1988. The effects of selective timber harvesting on rainforest primates in Kibale Forest, Uganda. Ph.D. disserta- tion, Univ. Calif., Davis. Slobodkin, L. B. 1961. Preliminary ideas for a predictive theory of zoology. Am. Nat. 45:147-153. Struhsaker, T. T. 1978. Food habits of five monkey species in Kibale Forest, Uganda. In Recent Agvances in Primatplggy, eds. D. C. Chivers and J. Herbert. 1:225-248. Academic Press, London. Svoboda, F. J. 1988. Toucan tending. [he A.F.A, Hatehpirg XV3:36- 38. Hatt, B. K. and A. L. Merrill. 1977. C m osi F od Bay. ProcessedtsPrepaueg. Agriculture Handbook #8, U. S. Dept. of Agriculture, Hashington, D.C. 70 APPENDIX 3-A A SYSTEMATIC LIST OF ALL FOOD ITEMS RECORDED IN THE DIET OF BYBANISTES SDBBXLLNDBLQDB Observations were made in Kibale Forest, Uganda, by the author unless noted otherwise (a or b). The following symbols indicate where the food was recorded: B - below nest, N - nest-watch, F - foraging observation. Food Item Family Species PLANT FOODS (FRUITS) Agavaceae Dracaena steudneri Anacardiaceae Eseudosppugias uieueeaupa Annonaceae Monodoua myristica We 'smgeusis Apocynaceae Aauyelfia pxyphylla Voacanga thouarsii Boraginaceae Duugia abyssinica Burseraceae Danarium sehweinfurthii Butaceae Eagarppsis angolepsis Caricaceae Dauiea papaya Ebenaceae Diospyrps apyssipiea Euphorbiaceae Brigelia migrantha Bapium elliptieum Lauraceae Beilseuuiegia ugangeusis Meliaceae Triehilia splengiga Melianthaceae Beusama abyssiniea Moraceae Antiaris tpuicaria 593999.13 L9__sn bero 71 Family Food Item Species Myristicaceae Myrtaceae Oleaceae Phytolaccaceae Rhamnaceae Rubiaceae Rutaceae Sapindaceae Sapotaceae Simaroubaceae Sterculiaceae Ulmaceae Verbenaceae Unidentified Bhlprpphpua excelsa Eieus braehylepis Eieus prachypoda Eieus eongensis Ficus eyanthistipula Eisus_u_dwi EiSUS exaseerata EIQQS DASHSQ Fieus natalensis Ficus polita U—usor 199193 Musauga rc 101 Eycanthus angolensis Magnum Dlea welwitsehii Phytplacea dodicandra Naesoesis £Diflii Vangueria apieulata 199199 9991119 Blighia unijugata Lychnpdiseus ro mus Beguaertiodendron.oblanceolatum 111—mow 939.29% mm mm Stepeulia dawei Deltis africana 9.9.13.1; QuLL—ndii Euemna augalensis 20 spp. XX XXXXXX X XXXX XX 72 Family Food Item Species Moss Fungus Lichen ANIMAL FOODS Mammalia Chiroptera Primate Ayes Colliidae Nectariniidae Ploceidae Re i ia Scinsidae Chamaeleontidae Mollusea Arthrppoda Diplopoda Insecta Coleoptera Lucanidae Eptesicus capensis Rousettus egyptieus Balago senegalensis Com. sp. Anthreptes epllaris Plpeeus nigerrimus Ploceus nigricollis Mabuya sp. Rippa fernandi Chamaelep sp. unident. ”snail" unident. "millipede" fipmoderus mellyi Homoderus sp. 73 Family Food Item Species Scarabeidae Metopodpntus sayagel Besptopus taua ngus Nigidi ius sp WM Subfam. unident. Cetoniinae Hybosorinae Coccinellidae Anthribidae Elateridae Meloidae Hydrophilidae Cerambycidae Lamiinae Prioninae Detpniinae unident. spp. Dhelorrhina sp. Coelorrhina loricata Dicranorrhina mieans Eccoptocnemis bapaulti W89- M91931 Eudicella spp. Goliathus spp. Neptunides stauleyi Eachnoda arrowi Pachnoda sp. Eedinorrhina subaenae Stephanorrhina sp. Mm SP- genus and spp. unident. genus unident. genera unident. (3 spp.) ietralobus s genus unident. §phaeridium,pietum §L§£DQLQmiS SP- genus unident. Acanthpphozus sp. XXXX X XXXXXXXXXXXXXXX X 74 Food Item Family Species B N F Carabidae Harpalinae Igfflg; sp. x Fam. unident. (several spp.) x Diptera x x x Anthomyiidae Anthomxia sp. x Fam. unident. (puparium) x Lepidoptera Pyralidae genus unident. x Fam. unident. larvae (several spp.) x x x Dictyoptera . x x x Mantidae genus unident. x x x (oothecae) genus unident. x x Blattidae genus unident. x x x Orthoptera x x x Tettigoniidae genus unident. x Hemiptera-Homoptera x x x Cicadidae genus unident. x x x Hymenoptera x x x Apoidea A215 mgllifgra adansgni x x x Insect order indetermined for 9 specimens aObservations in Kampala-Entebbe, Uganda, by J. Baranga. Isabirye Basuta (pers. comm.) and/or L. Kilham (l956). bObservations in Kibale Forest by T. M. Butynski (pers. comm.); M. Ghiglieri (1985). 75 APPENDIX 3-3 DESCRIPTION OF FRUITS IN THE DIET 0F BYCRNISTES SUBCYLINDRICUS HORNBILLS IN UGANDA j Fruit Fruit Fruit No. Part “mt“ 53:3“ "'32: °f 5235 Fruit Species Type Size Color Seeds Eaten Eaten 'Eaten Dispersal Damaged? (an) (m) Antiaris toxicara Drupe 7-13 Red l Drupe Red ’7-13 Regurgitation Mo Beilschmiedia Drupe 25 Purple- 1 Drupe Purple- 25 Regurgitation Mo ugandensis brown brown figguaertiodendron Drupe Red Drupe Red 7 oblanceolatum Bersama abyssinica Capsule Red 4-5 Aril Red + No yellow Blighia unijugata Capsule 32-51 Red 3-4 Aril Yellow + 19 Regurgitation Mo black Defecation Bosqueia phoberos Drupe 17 Green Drupe Green 11 Mo Bridelia micrantha Drupe 5-8 Blue- 1 Drupe Blue- 5-8 No black black Canarium Drupe 25 Purple 1 Drupe Purple 25 7 scfiiEinfurthii Carica 2225!; 100-230 Green Many Pulp Orange Varies Regurgitation Mo Celtis africana Drupe 5 Brown 1 Drupe Brown 5 R urgitation Some- De ecation times Celtis durandii Drupe 9 Yellow 1 Drupe Yellow 9 Regurgitation Some- Defecation times Chlorophora.excelsa Achene 35 Green Many Achene Green 35. Defecation 7 Cola gigantea Capsule 105 Brown- 7 Aril Red 33 Regurgitation Mo red Cordia ggyssinica Drupe 9 Purple 1 Drupe Purple 9 Regurgitation No Defecation Diospyros abyssinica Drupe 9-13 Red 1 Drupe Red 9-13 Regurgitation No Defectation Dracaena steudneri Drupe 9-13 Black Drupe Black 9-13 7 Eugenia Jambolana 7 Fagaropsis Drupe 6-8 Purple 1 Drupe Purple 6-8 Regurgitation No angg ensis Ficus brachylepis Fig 25-38 Green + Many Fig Green + 25-38 Defecation .No purple purple Ficus bracgypgga Fig 25 Many Fig 25 Defecation Mo 76 Col r f 1 Fruit Species i;:;t Eiglt 53i3: 52:3: ::t:n 23::Z SEEE::f ::§EE§S:: 0:£E§:d7 (M) (M) f1§g§,coggensis Fig 25 Reddish Many Fig Reddish 25 Defecation Mo 5152; cyanthistupula Fig 19 Green Many Fig Green 19 Defecation Mo [lgu§_ggggl Fig 27-50 Yellow Many Fig Ye11ow 27-50 Defecation Mo Ficus egasperata . Fig 9-l3 Red Many Fig Red 9-13 Defecation No [1523.maggggl Fig 30-55 Orange Many Fig Orange 30-55 Defecation No Ficus natalensis Fig 9-13 Yellow Many Fig Yellow 9-13 Defecation Mo f1£2é.2211£9 Fig 39-55 Purple- Many Fig Purple- 39-55 Defecation No green green ' Magriggn'i‘aca Drupe 9 Red Drupe Red 9 7 Lychnodiscus Capsule 24-39 Red 3 Aril Red + 18-24 Regurgitation Mo cercospermus white . Maesopsis £51311 Drupe 22-30 Yellow 1 Drupe Yellow 22-30 Regurgitation No Mimusops gagshawei Drupe 25 Orange 1 Drupe Orange 25 Regurgitation Mo Monodora myristica Husk 100-225 Green 125-750 Pulp Mhite 20-30 Regurgitation Mo Mgggs_1ggtgg Drupe 10 Green 3 Drupe Green 10 7 Musaggg_cercopioides Achene 50 Green Many Achene Green 50 Defecation 7 glgg_we1witschii Drupe 6-9 Purple 1 Drupe Purple 5-9 Regurgitation Mo Defectation Phfigolgggara Drupe 5-8 2:2; 5 Drupe Red- 5-8 Regurgitation 7 ge orange Defecation grgmga_angolensis Drupe 9 Green Drupe Green 9 Regurgitation 7 Pseudos ndias Drupe 18-27 Purple 1 Drupe Purple 15-25 Regurgitation No microcarga Pyganthus angolensis Capsule 25-44 :r::;; 1 Aril Pink Regurgitation 7 Rauvolfia oxyphylla Drupe 25 Green >3 ' Drupe Green 25 Regurgitation Mo Defecation Sgpigg.ellipticum Capsule 5 Aril 7 Sterculia gaggi Capsule 75-100 Red 3 Aril Black 19 Regurgitation Mo Igglgg_gg§111§_ Drupe 6-8 Red 1 Drupe Red 6-8 Regurgitation 7 Trichilia splendida Capsule 23-39 Red 3-4 Aril 3;;ck + 17-25 Regurgitation No Uvariogsis congensis Drupe 17-45 Red 3-10 Drupe Red 17-45 Regurgitation No Vangueria apiculata Drupe 25 Green Drupe Green 25 Ma Voacanga thouarsii Drupe 50 Green Drupe Green 50 7 Sources: Eggling (1952). Milne-Redhead and Polhill (1952). Hamilton (1901). and personal observation. 77 .mmmma pgmwoz mew a co mpmapmcm NNum=m mmhmHzm.mzom no m~m>4< so; ugam vamp; mmNuaam came a peach a mmNNoNoo NN.>N== uNaNm ouaNoNou Na ”NuNNa=m Nzom No mNmNN 3 m circumference is higher at Ngogo than at Kanyawara, and these trees are, on average, much larger (1. Butynski, unpubl. data). Nestin ccess Female hornbills emerged with fledglings after being confined between 115 and 142 days in the nest (n - 56) (Appendix 4-D). Mid- March was the latest date for fledging. Although two eggs were laid, one chick was usually raised (Kilham 1956). Only one nest (n = 56) at Kibale contained twins that were ready to fledge. Fledglings emerged from nests located in both logged and unlogged areas (Appendix 4-0). Nesting success was assessed as percentages of nests from which young hornbills fledged. Values 91 G ------------ .\ NGOGO I H S u‘ f ‘ T’ e HORNBILL NEST (ii-30) T F‘. a HORNBILL DISPLAY (n-IO) I CAVITY e ....... ”0......00000..... e ...... ........ eeeee eeeee A» N/ e1 -e ‘ 10 o c 13 O f\ A a IKM #1 5 GRASSLAND SWAMP Fig. 4.3.--Locations of Byganisigs 5gbgyligdrigu§ nests (n - 30) and display cavities (n -'10) at Ngogo (unlogged forest) in Kibale Forest, Uganda (1981-1987). is trail grid, ----- is elephant trail, and ----- is motorable track. K‘3O BUTANZI FOOTPATH ’ err/JAG .9!“ 0 ‘ 00000 0M0 0000’ 0“ 000000 000400009 ”99000”, 9‘.¢0¢ 0.000s\\\00% 00%»? 0“%% . //I’ .00 ' .6 m . run-u. 0.. . «A. s. o. I (n-7) A HORNBILL DISPLAY (n84) CAVITY I CROWNED EAGLE (n- I) O HORNBILL NEST b——l IOOm \ S «I . o.0 _ Qki.0$ 0;. .os../\ ‘ o . . 00‘ .. O soak Woo eéuflex ..«0%m...~.~.%&.vv 0»? t 0494’ 4:00 o 3 ‘00 00”0O~00%ooo¢00-9~o e NEST is display cavities (n . 4) at Kanyawara (K-30 unlogged forest) in Kibale Forest, Uganda (1981-1987). trail grid. -Locations of Bycanistes subcylindricus nests (n = 7) and Fig. 4.4.- 93 K'l4 . e HORNBILL NEST (n-4) TO "\. MlKANA NURSERY 3 "I " FOOTPATH CAMP I .a CENSUS . ROAD .- "'GRASSLAND TRAK BUTANZI FOOTPATH Fig. 4.5.--Locations of Byganistes subcylindricus nests (n = 4) at Kanyawara (K-l4 light-moderately logged forest) in Kibale Forest, Uganda (1981-1987). is trail grid. 94 K‘IS 113 SWAN”D (“1“ IVB 9E: BE \ 7E 6E L\ 6 6‘ 55 SK T ‘hAE 2E: \\ '. IE \\\ 5139 15 as as 45 SS 65 s 85 SS ew I———J IOOrn O HORNBILL NEST (n83) I CROWNED EAGLE (ni- I) NEST Fig. 4.6.--Locations of Bycanistes subcylindricus nests (n - 3) at Kanyawara (K-15 heavily logged forest) in Kibale Forest, Uganda (1981-1987). is trail grid. 95 ranged from minimum to maximum percentages of successful nests. Minimum percentages were determined by counting the number of total nests from which young definitely fledged. Maximum percentages were determined by counting the number of total nests from which young may have, but did not certainly, fledge. Maximum percentages resulted from lack of precision in data collection; i.e., the date on which the nest was first sealed was not determined, or the date and sighting of the chick leaving the nest was not verified. Nesting success was at least 40% (n = 5) in 1982, 36-55% (n = 22) in 1983, and 62-66% (h - 29) in 1984 (Appendix 4-D). Birds attempting to breed had a nesting success of 36-66% or more (i.e., 1982). 8. subcylindricus nesting success in Kibale Forest was considerably lower than that of Tockus spp. hornbills in a South African savannah. Of 249 nesting attempts by Tockus there, more than 90% were successful in fledging young (Kemp 1976). During 1983-84, the only year when all nests apparently had been found and monitored in all study areas, more young were fledged from unlogged than from logged forest. Ngogo had the highest density of chicks fledged (1.82-1.97/km2),followed by K-30 (1.76/km2), K-15 (1.67/km2), and K-l4 (0.9l/km2). Reasons for nest failure were varied. Several nests had entrance holes. that. were too large or faced upward, subjecting them to flooding. One female abandoned her nest on a horizontal limb after Red Colobus (Colobus bagigs) monkeys walked across the nest entrance. 96 Crowned eagles (Stephanoaetus cgronatus) also threatened nesting hornbills. One male hornbill was apparently killed by an eagle while he was trying to feed his mate. His remains were found below his nest half-way through the nesting season (Appendix 4-0). After the male hornbill’s death, the female hornbill abandoned the nest and any chicks in the cavity. Within weeks, the crowned eagle pair moved their nest into the dead hornbill’s nest tree (OJ/47 in K-30) from its prior location 200 m away. As of October 1987, crowned eagles continued to nest in the JJ/47 tree. A pair of hornbills displayed at the old nest cavity but did not nest. Intraspecific intruders were the suspected cause for 20% of the nest failures in 1983 and 80% of the nest failures in 1984 (Appendix 4-0). Harassment by intruding 8. subcylindricus males and females caused nesting birds to abandon their nests during incubation, nestling, and fledging stages of breeding (Appendix 4—D; Chapters V and VI). Percentage of nest failure was higher in the 1982-83 than in the 1983-84 breeding season. In 1982-83, 20% of the failures (n = 10) were attributable to the starvation of chicks that were very near to fledging (Appendix 4-D). In these cases, females broke out of the nest cavity prematurely. Abandoned chicks were heard crying from their nests, but the parents did not feed them. One hungry nestling squeezed out of the nest and fell to her death. Sex-Ratio Bias in Fledglings Sexual dimorphism in casque size (Fig. 4-7) allowed for identification of each chick’s sex upon fledging. Sex ratios were 97 mmumNcwuNm mNmEmN mcwpmuwpm use uNaz chNmuuNN .mzuNNucwp oazm mNmE chNmumNN co acmsaopm>mu ozcmmu :wmzpwn mocmemmuwnii.n.¢ .mwu uNasuN chNmuuNN 98 skewed, favoring different sexes in different study areas. Sex ratios of' 22 fledglings in unlogged forest (Ngogo, K-30) and 5 fledglings in logged forest (K-l4, K-lS) over 3 years (1981-1984) were used for analysis (Appendix 4-D). The sex ratio of 18 males/(18 males + 4 females) in unlogged forest is significantly different from 50:50 (X2 a 7.68, df - l, p < .001). The sample size for logged study areas is small. It appears, however, that the sex ratio of 1 male/(l male + 4 females) is not significantly different from unity (X2 - 0.80, df - l, I) > .20), although females were in a majority. Fledgling sex ratios in Kibale Forest were skewed toward males in unlogged forest and were 50:50 or skewed toward females in logged forest. This is the first account of a skewed sex ratio in young of any species of hornbill. Hypotheses for cause and effect of this phenomenon will be presented elsewhere. DisgussionZRecommendations for Further Study From observations of B. subcylindricus breeding biology in Kibale Forest, it is apparent that rainfall and habitat are factors influencing reproduction of this African hornbill species. Not only does nesting begin with the onset of ‘the rains, but there is evidence that nesting success is also related to rainfall. Starvation of nearly fledged chicks was observed during a drought year. The relatively low percentage of nest success during the 1982-83 breeding season may, therefore, be related to rainfall and food supply. Rainfall during pre-breeding and breeding months of 99 July 1982-March 1983 was less than half as much as in other years (Table 4.2). The 1982-83 drought appears to have been particularly unusual and severe when compared to the relatively regular rainfall patterns typical for Kibale Forest over the past 20 years (Butynski 1988). Relationships between rainfall and food supply will be investigated in a planned analysis of my 1981.-1984 Kibale Forest tree-fruiting phenology data. Activity and feeding data collected at hornbill nests will also be assessed to reveal differences in quantity and quality of foods supplied to nest inmates during the 1982-83 drought year. Table 4.2.—-Total rainfall (mm) during Byganistes subgylindricus pre-breeding and breeding months (July-March) in Kibale Forest, Uganda (1981-1984). Study Area Year Rainfall (mm) Ngogo 1981-82 1339.9 1982-83 644.6 1983-84 1368.1 Kanyawara 1981-82 ' 1229.9 1982-83 710.6 1983-84 1608.5 B. subcylindricus nesting requirements are specific in that the birds only nest in large, natural cavities of very large trees. Nest-site characteristics will be appraised through multivariate analysis to distinguish other important parameters possibly affecting habitat choice. Primary forest is apparently the preferred habitat for breeding, since nest and display cavity 100 density is higher there. The number of young fledged per unit area was also greater in unlogged forest. A more detailed analysis of macro- and micro-habitat at ,8. ,ggbgylindri;u§_ display-sites and nest-sites, and of the success or failure of breeding at those sites, will also be undertaken to estimate the effects of logging and other forestry practices on hornbill reproduction. 101 REFERENCES Bourne, O. and D. Chessell. 1982. Breeding the black-and-white casqued hornbill (Bycgnistes suntylindritus subgnadratns) at the Metro Toronto 200, Canada. Avicglt. Mng. 88:15-23. Brown, L. H. and P. L. Britton. 1980. eedin eas s f ast Afrisan Birds. The East African Natural History Society, Nairobi. Butynski, T. M. 1988. Guenon birth seasons and correlates with rainfall and food. pp. 284-322 in A. Gautier-Hion, F. Bourliere, J. P. Gautier, J. Kingdon and R. M. Martin, eds. Primate Radiation: Evolutionary Biology and the Afrigan Guenons. Cambridge Univ. Press, Cambridge. Butynski, T. M. in prep. Comparative ecology of blue monkeys (Cer- conithecus mitis) in high and low density subpopulations. Ghiglieri, M. P. 1984. The Chimpanzees of Kibale Fnrest. Columbia Univ. Press, N.Y. Harvey, P. M. 1973. Breeding the casqued hornbill at "Birdworld." Ayigult. Mag. 79:23-25. Kemp, A. C. 1973. Factors affecting the onset of breeding in African hornbills. Proc. 16th Int. Orn. Cong.: 248-257. Kemp, A. C. 1976. A Stgdy of the Ecology, Behavior and Systematics of Tockus Hor 'll Aves: Bucero ' . Transvaal Mus. Memoir No. 20. Pretoria, South Africa. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. Livinq Bird 17:105-136. Kilham, L. 1956. Breeding and other habits of casqued hornbills (Bytanistes subcylindrisus). Smithson. Misc. Collns. 131: 1-45. Lwanga, J. S. 1987. Group fission in blue monkeys (Cerconithecus mitis stuhlmanni): Effects on the socioecology in Kibale Forest, Uganda. M.Sc. thesis. Makerere Univ., Kampala, Uganda. Mackworth-Praed, C. N. and C. H. 8. Grant. 1957. Birds of Eastern end North Egstern Africn. Longman, Green and Co., London. Moreau, R. E. and Moreau, H. M. 1941. Breeding biology of the silvery-cheeked hornbill. Ant 58:13-27. 102 Nummelin, M. 1986. The seasonal fluctuations of forest floor insect densities on the areas of different forestry practices in Kibale Forest, Nestern Uganda. t nt. Conf. n Tr ical Entnmology, Nairobi. Porritt, R. and M. Riley. 1976. Breeding the black-and-white casqued hornbill (Bytnnistes subcylindriggs) at Birdworld, Farnham. Int. Zon Itk. 16:104-105. Skorupa, J. P. 1988. The effects of selective timber harvesting on rainforest primates in Kibale Forest, Uganda. Ph.D. dissertation. Univ. of Calif., Davis. Struhsaker, T. T. 1975. T R l s k . Univ. of Chicago Press. Chicago, Illinois. 103 .00.. :2. No 3:... I on 5.2... :3 I o. m :0... No it .29.... use... . 2 N32. 9.. u o. i... .e N... N2...N .3233 2.3. 3.3.3 NEE-N538. -- 8.x .N .- a. N... N... N .. o. N. . .N N.N. N. N..N ILNFJNNNAWLNB N... N .5... .N 2- a. ..N N.. N N N. N. N : ...N N N 8.. E... .8 . 5... N.N... N. .5... .N N.. N.N N... ..N. N . .N w. N.N. N... . N N... .72.... . 3...... NN... .- .5... N. 2. .. N.NN N.NN N . N. .. N.NN N.N. N... 8.. .35.... P85... 3.. N. .5... N. N.- N. N.. o... N 2... e. N. N.N. N.NN ..N N... BREE N... N. N.N... N. .- N .._N N... . 2... N. a... N... N... N.N 8... Emma-13$“ .3. N. . .. a. N... .... ..N a... N. e. N... N..N N.N RN 2.3.... z. N: NN . N.- ..N ..N N.. N .N... N. .. N... N... ..N 8... IVES-N375. .... N. N N.- . N... N... . 2.: N. .. N.N. N.NN N.N N... .NEEE .2. .. I I I -- 1 a -- 1 S I .. I 33 cc .1 3....»th 93 .N B... N N. .. N.NN .6. N . N. S... N... N.N N.N 8.: I: inflame... N.N... NN . 2. N. ...N N... N :5 N. e... .... N.N. N.N N... G: Eufiwwum. .zN... . :8. N. N.- N. N... N... N 2... a. a N... N... N; o... Idmmitfuwwhwfi 2:. o. 3.3... . .- N N... N.N. N 2.5. N. S... N... ...N ... N... 58.8... r. E. .. Ni. . 52.. c. .- N. 3.. N 2... N N N.. N .N N.N N... .. N .N- N.N ...N N... N .5 N. N N..N N... N... N... INHINNI N... .N . :- N.N .... N.NN . 2... N. N N... N... N.N .o... E KNEE... . .3. .N N 2- N ..NN N... . N... NN 8.... N... N... N.N .8.. 9.5.... s .55»: N. . :N .N N. I... N N.- a. N... N... N 2... .. N. N... N... N.. SN 51...:- ¢NNBNN N.N. N. o .N- N. N..N N... N 2... o. N N... N.N. ... N... E33119? .23 N. N .N- N N... N.NN N N N. N N... .2. ..N NN.N 8......1... . :5: N.NN. NN I... N. 2. N N... N... N .N... 8 N ...N N.N. ... 8.. EEG-E . 5. -. .. - N - - . . N. N N... N... ..N 8.. ENNE 2:: - -- - N .- -- - - N. .. N.N. o... -. 8.. nuts: :55 N33. NN . 2. N. ..N. N... N 2. .N a. N... N... ... .8.N LESEPRE m... .N N. N. N .... N... N .2... o. N N... N.NN N.. N... ILHNNNIE-Wai .9“. N. . :- NN N.N. N N... N N. .5 .6. N.N 8.. :Ewuifimlzw. S... .N N .- NN ..NN .... N N... N. .N. N.N. ...N N.N .... EVEN-HEN. N.N. .. .- .5... N. - N -- -- . -. N. N N... ...N N.N 8.. Emma "EN”... .3. .. - - .. - - N 2... 8 a... N... N... N.N 8.. Elm? 2.. N. N 2. N N... ...N N N N. N N... N.N. N.N .8.N :NxzmmNNulzolmwN . NW. .N 8.x N :- NN ..NN ...N N N.N. .N .2... N..N N... N... N..N :55... .2..- N NN. . N. I... N N. a. 9... N 2.: N .. N.N N... ... a... .Eme 2... .. N N. a N... N.N. N .N.NN N. N. N... N... NN.N .8.. 5.3.... 553%. E. N. N 2. N ...N a... . N.N. N. N. N.NN N.N. N.N .8. u. _ N... m . N.- N. N... N... . . a e. N... N.N. N.N 8.. . "w“... .. N N... N N... ...N N . 2... N. a. ..N. N... N.N .8.. ee.wE-_.§:..EmmN-E 2.... .. N N.- N .... N... . 2.2 N. .. N... o... N.N .8... :...-.m..:. 3. :N. ..NNNHN .. N :- N ...N ..N. . N1... 8 .. N... N... N.. N... 5:25... E t. N N. .5... . N- N. N.N. . .6. N N.N N... N... N... N... . Steam-mm“... N.N... .- .8... . - . -- - . . N. N N.N. N.NN N.N N... .5555... N.N. :5. N. :8. N N.- N N.NN ...N N 2... N. N. N.N. N.NN N.. a... L... . .3 N.N.NN .N N 2. N N... ...N . N.... N. .. N... N... N.N .N.. EatNPEmNE-NN E.- .N . 2. a N... ..N. . 2: N. . ..NN N.N. N.N 2.. S :5. . .2. N.N. .- NN N 3. N N... N... N 2.. N. N N..N N.N. N.N 8... .2. NS. «.050» 2.. N... .3. N... N... N... Mn. N... N... N... 8.... Na... 3.... N.N-3 6...... so... 2.... 2. .9. 3.... . '00. ’00 o» w. s... .u when. 8...: .58 :8- ..28 :9. 2... .3. .3. 2h“: : in“: 3...... 8N. in“: No. 8...... c... Eh“: 8...... 2... N c... N... £52.. (iw x~ozmam< um=m mmhwmzm no mumhmmmmhuoao N.N:NNNE aucogounau..u. .vouagN Nuaxpu n o— Novas. N: a o .3... .N «an ape: . N - .N... .N N. N... N..N N.N N..N .Nmmmmmumm.m~mmmaflm .NN N . - 8.... N. ..N N. .N N.NN N.N N... gm... 3 .55 N. N - -- N z . N.NN N.NN N.N N... - .NN. . N - NNNNN N. N. N.NN N.N. N._ . N..N - .NN. NNoNN. v. N.NN on-x N..:. can N N N N N. N. N... N... N.N - - N..N... N N .N N N. . N..N N.NN N.. N..N .NNNNNNN ...N...N NN..NNN . N N N..NN N. N.NN N.N. N.NN N.N N..N 25.....N. .N.....N.N.N .N .NN.N . N - .- N N. N... N.NN N.N N... .. , - .NN N. N u... N..NN N. N N.NN N.NN . N.N N..N .. - .NNN.= N N N.. a N. N. N... N.NN N.N N..N - NNN.. 3., N N - N. N. N..N N.N. .N.N .NN.. ...N..N. N....N=< NNN.N --N - - - - - - - -- - .N.N NN N..N..N . . N H...N NNxNN N. a... N.NN N.N. . N.N .N.N .NNNNNNNNNNNN .N... N.NNN N N - N..N. N x. N... N... N.N .NN.. ......NNN.-.NNNNNNNNM , N.NNNN.N .5. .Eu. 5 z s E one—u vac—10v .004 o—auaEN—u ocvevc too: moo-m aszwz A v huM>wu A v. ucxoo No>ou uNogooz o. o—oz o—o: m>_. N.N: N.N: ago—o: N. ...a:u..u No—uunm on.» =NNNNuoN ...a on.._oN I... uuc....: NNoz u .._a NNoz om.» .Nsau..u a... N.oz HH>um3m mMPmHzm no muHhmHmmho5 F1edgiing Yes H E.5/7.5 12/21/81 69 7 Yes No 7 $12.5 10114184 10114-16 15-17 No 7 2 No data F15.5 9112/82 DISPLAY ONLY 7 No F/9.5 7122/82 1015-10 247 Yes No Nesting hornbill ‘ killed F112 9126/82 10126-30 252 7 7 F.5/O 9119/83 916-12 276 7 N0, tree fell in 1985 F.5/3.5 9116/82 9130-1016 252 7 Yes Yes 817 12121181 9111-18 271 7 Yes Yes N10 9121/83 9130-1016 252 7 7 No data N13.5 9129183 9130-1016 252 7 7 No data H.5/10 9130/83 9124-29 258 7 Yes Yes 1.512 12119181 -- -- -- -- -- -- -- -- -- J1-O.5 8122/83 DISPLAY ONLY 7 No J/O.5 1011182 -- - -- -- -- -- -- -- -- K12 9113/83 9130-1016 252 7 (HOLE 42) Yes Yes (HOLE 13) 104.5 9114/83 SEALING NOT WPLETED Noletoolarge 7 No data K17.5 9112182 DISPLAY ONLY Intruders 7 No L19 919183 9123-29 259 7 Yes No 7 L110 917183 9129-1016 252 7 Yes Yes L.517.9 9114183 9111-18 270 7 7 No N110 9127/83 915-11 277 7 No. tree broke at nest N19 9126183 915-11 277 7 Yes Yes Bigodi Village (forest edge] Kanyawara (K-3O unlogged forest) NH/S 9130/82 9122-26 232 7 Yes Yes J/39 1011182 9122-29 229 7 N0. nest lieb broke off JJ/46 10115181 DISPLAY ONLY No No Crowned eagles nest in tree JJI47 10114182 DISPLAY ONLY No No Eagles nearby 5140 9112183 10120-24 2 1 7 7 N0 .114 1011102 1013-7 219 1 Yes No .110 1011182 10112-19 z 1 7' Cavity a bee hive in 4134 No. trunk broke near mt Kan auara K-l4 1i ht-noderatel 1 ed forest F10 1011182 DISPLAY ONLY No data N124 1015/82 Tree fell before 8184 -- -- '- Y/17 1016182 DISPLAY ONLY N0 "0 Rim H0?" Y117A 12122186 YES Yes Kan aware K-16 heavil 1 ed forest 3E125 10121182 9128 230 7 Yes Yes 36135 10121182 9123-28 230 7 Yes Yes 26175 9130184 9130-1014 a 8 7 Yes Yes 'Dates covered. CHAPTER V NEST INTRUDERS, NEST DEFENSE, AND FORAGING BEHAVIOR IN THE BLACK-AND-NHITE CASOUED HORNBILL mm During the 4.5 month nesting period, the male black-and-white casqued hornbill (Bycanistes subcylindricus) is solely responsible for feeding himself, his mate, and his offspring in the nest. According to optimal foraging theory, he should accumulate food in the minimum amount of foraging time in order to maximize the delivery rate of food to the nest (Andersson 1978; Orians & Pearson 1979; Schoener 1979). This prediction is based on models which assume that there is need only to maximize the net rate of energy intake. The foraging behavior of the male hornbill will, however, be influenced by riskSof predation at the nest (Fig. 5.1) and demands of feeding and protecting the young. In order to account for these additional demands, several authors have investigated time budget conflicts between feeding and other activities such as maintaining vigilance (Caraco et al. l980; Lendrem 1983), minimizing predation risk (Sih 1980), and defending territories (Kacelnik et al. l98l). ’ Martindale (1982) incorporated nest defense and central place foraging into a model which makes different predictions about 110 111 Fig. 5.1.--During the nesting period, the male Byganistes subcylin- dricus is solely responsible for feeding himself, his mate, and his offspring in the nest. At the same time, his foraging behavior will be influenced by risks of predation and demands of protecting the young. 112 foraging behavior. Martindale’s logical assumption for development of' his "nest defense model" is that ”natural selection should produce animals which behave to maximize their net fitness, not just their delivery rate." His nest-defense model predicts that, as predation risks increase, the defending bird will forage closer to its nest, and bring smaller food loads and smaller food items than predicted by models based on delivery-rate maximization. Although not tested under conditions of actual intrusion, Martindale (1982) found that male Gila woodpeckers (Mglanerpes uropygialis) did indeed forage closer to the nest and deliver reduced food loads and smaller food items after attacks from an artificial "intruding" woodpecker. Although Martindale’s experiment was valuable, two factors may have compromised his test of the model: (1) the "intruding" woodpecker at the nest was artificial, and (2) both male and female woodpeckers provisioned and defended the young. Results of his study were difficult to interpret when both male and female woodpeckers joined in foraging and defense. In contrast, the black- and-white casqued hornbill serves as an ideal species fur testing the nest-defense model. The female uses mud to seal herself inside the nest cavity. She remains in the cavity until the nesting attempt ends. The male is solely responsible f0r food provisioning and external defense of the nest. Nest defense is crucial for hornbills since inadequate nest protection against intruding 113 conspecifics may result in the eviction of the resident pair and/or death of the resident pair’s young (Kalina in prep). Nemesis In 1981, I began a long-term study of’ the black-and-white casqued hornbill in Kibale Forest, western Uganda.’ The study area consisted of both primary and selectively logged tropical rainforest. This area has been described in detail by Struhsaker (1975) and Ghiglieri (1984). Since 1981, 45 nests have been found. All are located in the natural cavities of large trees. Except for 1985, nests were monitored each breeding season from 1981 through 1987. Observations were made from locations 20-60 m from the base of the nest tree using a telescope or 10 x 408 Leitz binoculars. The data used in this paper were collected over the 1983-84 nesting season during 4 h (n = 28) and 12 h (n - 28) sampling periods (700-1100 h and 700-1900 h). Only information collected at sealed, occupied nests was analyzed. Thirteen nests were observed for a total of 448 h. Data meeting the requirements necessary for testing Nartindale’s nest-defense: model were obtained during 10 sampling periods from six nests. To reduce sampling error caused by between-day variation in foraging activity, I only utilized data collected on the same day before, during, and after the visit of one or two intruding hornbills. Therefore, these data describe foraging activity during short-term intrusions. An ”intruding hornbill" is~ defined here as one that approaches to within 50 m of the nest and evokes an agonistic response by the resident male and/or female. 114 I only watched nests with clearly visible entrances so that birds landing on the nest could be identified and food items fed into the nest distinguished. The male hornbill’s arrival and departure times at the nest, and the kinds and numbers of items fed into the nest, were recorded. Hornbills are multiple-prey loaders; i.e., more than one food item may be collected before a load is delivered to the nest. ‘ Male hornbills regurgitate food items into their bill tips just before they feed them to the female in the nest. These food items were usually easy to see and identify. The species of fruit was identified to species and given a rating of 1-4 based on size, with 1 being the smallest and 4 the largest. Examples of species (with approximate mean diameter of the fruit) in each category are: 9.9.1.115 durandij (0.5 cm) - 1, 11.050119}. abyssinica (1.0 cm) . 2, Mimgsgps bagshgwgi (2.0 cm) . 3, and £150; ggwgi (3.0 cm) - 4. Size ranges (diameter) of fruits for each category are as follows: 1 = (0.3-0.7 cm), 2 - (0.8-1.5 cm), 3 - (1.6-2.5 cm), and 4_- 2.6-3.5 cm). Volumes were calculated for each category, based on the means: 1 . 0.07 cm3, 2 = 0.52 cm3, 3 = 4.19 cm3, and 4 - 14.14 cm3. These values (were derived from the volumes and diameters of samples of fruits. For each visit, the number of fruits in each size category was totaled and multiplied by the volume for that category, then , added together to determine the "load volume.” For example, 4 Eiggs ggwgi fruits fed during one visit would be multiplied by 14.14 cm3 to get a "load volume" of 56.6 cm3. 115 Data were combined for all 10 sampling periods for the two visits to the nest by the resident male "before the intrusion," for the two visits "after the intrusion," and for all visits ”during the intrusion." Sample sizes vary from table to table because of incomplete observations. For example, on three occasions, data are available for only one (not two) visit to the nest by the resident male prior to the intrusions. For one sample, no data are available concerning visits by the resident male before the intrusion. For each intrusion, the following were determined: (1) total time (min) that intruders remained within 50 m of the nest, (2) mean time the resident male spent perched on the nest entrance, (3) mean time since the resident male’s last visit to the nest, (4) mean "load volume” fed by the resident male into the nest, and (5) mean volume of food brought/min to the nest (Table 5.1). Results Intruding hornbills stay in the nest area for a few seconds to several days or weeks. The "short intrusions” considered in this paper lasted for 1-51 min (n = 10, mean = 20.9, SD = 16.7). Once intruders were detected by the resident male, he remained within the vicinity of the nest, going only short distances (up to 100 m) to forage. Before or after intrusions, males often flew to fruit patches 1 km or more away (pers. obs.). Agonistic responses by resident hornbills to intruding hornbills of either sex included calling and drumming by the imprisoned resident female, and chases and vocalizations by the resident male. The male sometimes 116 .Fm>mp mo. use zepma so no acoowepcmwm appmuFummumpm Ho: u .m.= A.m.cv cm. A a Foo. v a poo. v a Po. v a mooo. v a A.m.cv No. A a A.m.:v on. A a A.m.cv cm. A a A.m.:v op. A a A.m.cv op. A a A.m.:v cm. A a No. v a No. v a mo..v a mcoo. v a =_e\p;m=oen coo» eo mE:_o> Amuummzv a pngwv msapo> use; Aapco awaeev mE=Fo> vac; paw: co meek uwmw> pump emue< .m> memeao empw< .m> mcoemm mcwczo .m> meoemm «m m.ucmu:umv mamms :mmzumn monocoewwo m.pnm.o ¢.pmw.o m.pwu.o Ampucv m.-aw.m~ Ammuev ~.m hm.~ Ampuev A.oeae.~N Aapuev m.~NmP.m~ Amwuev A.m am.- Am_ucv N._eee.- Am_uev m.oe~.m Ammucv m.oeo.P Ampuev «.oee.~ mucwm meek A¢Fucv m.mne.um mem< Aopncv u.pwe.e mavens Aomucv m.n«m.we wgoemm cms\usm=ocm Amsuv noon mo m53~o> Ampumm:_ a peacev AmEuV me=_o> vac; A»_eo .st pwscmv Euv > ewes ammz as “swam mmpacpz uwmw> mzow>oea mucwm mmuacwz .eoccm nemucmpm n memos mew waspm> .muwm “mm: use an move-u imamcou An meowmacuzw Logan ucm .mcmszu .mcommn Aemimwmpv magma: .pmmcoe upmawx cw msu iwgucwpaunsm mmumwcmuam mpms mcwwmmc eo Low>mgmn mcwmmcow mo mucmemcammme Lao» cw mmmcmguii.~.m mpnmh 117 "chuttered" loudly and tapped his bill against the nest entrance while rapidly shaking his head in a ritualistic ”nest-sealing display." The intensity of these behaviors varied with the severity of’ the attack. Resident females "bill-grappled” with intruding hornbills that perched at the nest entrance. During bill-grappling, birds jabbed at each other’s bills, sometimes cutting the other bird’s head or knocking the opponent off balance. Males were observed to bill-grapple only with males, except on rare occasions. Once, an intruding male perched at a nest entrance and grappled with the resident female, who was inside the nest. During the period covered by this analysis, a total of 960 fruits and 11 insects were fed. Insects comprised only 1% of all food items (8% by volume). Load volumes were calculated, both including and excluding insects (Table 5.2). Items fed and not tallied were two pieces of lichen and one piece of bark. Hornbills bring mud to the nest during initial nest-sealing. Later in the nesting period, the male brings mud when the seal is in need of repair. I saw 10 pieces of mud brought to a nest by the male during one sampling period. In this case, all mud was presented to the female by the resident male during and after the intrusion, but not before. Hud was delivered only after the intruders had remained 4-100 m from the nest for 1.4 h. The female hornbill did not accept most pieces of mud offered to her, presumably because the opening into the nest was already very narrow. Nud plastering had ceased two days earlier, when the nest- seal was completed. 118 .Foo. v a .m u we .m.Pmp x i.mco.m=cu:p suave .m> mcvgao N .poo. v a .m u we .n.mu u Nx i meowmacucm emuwm .m> meommm .po. v a .m u en .m.~p u «x i mcowmzcucp acmcau .m> mgoemm .mmmmucmuema mew mommsucmcma c? macamwum «.mm a «.em a, 5mm on o on em Apmv mam Am v A, teae< N.N a e.m mm on on o Apv P Amev mm Aomv NN mcpeso e.m_ a m.a_ o_ NAN Amv m_ Amy FN Aomv em_ eAva mm «Locum up.e_ mp.¢ mm.o mo.o u Amsov .po>.m mm H omo4\muwsem memo; mpwace oo.m oo.~ oo._ om.o u AEuV .2mwu.w .._.c .0: ix. .._.O .02 _.muo._. Awmgm: w AmPIUmEV m AEmIUmEV N APPwEmV _. 3.5: .8 22.52 .mvoop pupa» mo Amazpo> an xmv xp apco acmmmcame amcu mucmm umnspucm no: wen muummcfi .ummmcwp on» aspen u can ummppmsm mg“ mcwmn F sue: .mNPm no comes «up mo oceans m cm>wm mew: avenge .mxumpum cwmchcw empem mam .mcwczu .mgoemn AvwimmmFV macaw: .ummsou mponwx c? mppwnceo; m=Uwcccwannzm mmummcooxm mpme an mama: mg» op usmzocn mcmop peace we copuvmoasou-i.~.m mpamp 119 As predicted by Nartindale’s nest-defense model, nest guarding by the resident male during intrusions altered foraging patterns (Table 5.1): a. Male hornbills had a significantly higher rate of visita- tion to the nest when intruders were in the vicinity of the nest site than when intruders were absent. A greater proportion of these visits was without food (binomial test, 2 - -9.65, p < .0005). All (0 = 16) of the visits before, 36% (n = 25) of the visits during, and 84%. (n = 19) of the visits after intrusions by hornbills resulted in food delivery to the nest. Often the male visited the nest only to perform his nest-sealing display. Such activities indicate that much of the defense against conspecifics is focused on the mud barrier at the nest entrance. b. The mean time the resident male spent perched at the nest entrance was significantly less during an attack by intruders than either before or after the attack. c. The volume of food brought to the nest per visit decreased significantly during the intrusion and increased to about pre- intrusion level again after foreign hornbills left the nest area. Delivery of smaller loads during attacks by intruders was as predicted by the nest-defense model. d. There was no significant change in the delivery rate of food (mean volume brought to the nest/min) before, during, or after‘ intrusions. In short, the presence of intruders resulted in more frequent but shorter visits to the nest by the resident male, with smaller 120 food loads being delivered. llhen intruders remained in the nest area for a short time (5 51 min), the volume of food delivered per unit time did not change with the presente of intruders. There was a significant change in the composition of the fruit load brought to the nest before, during, and after attacks by intruders (Table 5.2). During the intrusions, resident males delivered a greater proportion of small (size 1) and small-medium (size 2) fruits to the nest than any other size category. They brought more large (size 4) fruits before than during or after intrusions. Greater proportions of small-medium (size 2) and medium-large (size 3) fruits were brought after the intrusions. During intrusions, rather than traveling long distances to collect the larger Ficus or flimusops fruits, males gathered the smaller, more common, Cgltis and Oiosgyros fruits that were available nearer the) nest. In summary, during short intrusions, male hornbills foraged closer to the nest and delivered smaller food items in smaller loads. These observations are as predicted by Nartindale’s model. During long, severe intrusions, however, food-delivery rates declined as the resident male hornbill spent. most of' his time defending the nest. At least 4, and probably 6, out of the 10 nest failures in 1984 can be attributed to disturbance of this kind. Imprisoned females were forced to abandon the nest when their mates could not provide enough food. At one such nest, I saw the resident male chase an intruding female 52 times in 8 h. Hhen food was 121 delivered, fruits were rapidly dumped directly into the nest, rather than fed to the female one at a time. In a second case, I found an intrusion already in progress, with the resident female breaking out of her nest. Although most nests contained only one young, this one housed two nestlings that would have fledged in about 10 days. The female flew feebly from the nest after her 4 mos of confinement, but re-entered the cavity within 24 h. She remained there and defended the nest entrance from repeated attacks during the next 4 days. During this 5-day period, the resident male was visibly tired, often chased away from the nest, and frequently driven to the ground by the intruding pair. This harassment led to reduced food-delivery rates to the nest. The nest was watched for 5-7 h each day during the 5 days of intrusion. The volume of food (cm3) brought/min was as follows: Day 1 - 0.5, Days 2 and 3 - 0, Day 4 = 0.3. On Day 5, the intruding female broke into the nest cavity, drove the nesting female out, and attacked the two nestlings. At least one, probably both, chicks died as a result of the attack. Discussion Evidently, the male hornbill changes his foraging behavior markedly as the threat to the nest increases. Changes in foraging activity in response to an intruder may be particularly dramatic in hornbill males because only that member of the pair is responsible for food provisioning and external defense of the nest. Even when the nest is sealed with hard, brick-like mud, the cost of leaving 122 the nest unprotected can be great. It was not unusual to find intruders perched on a nest, chipping the mud-seal away with their bills, while the resident male, unaware of their presence, foraged far away. Most changes in foraging behavior were as predicted by the model. During the short-term presence of intruding hornbills, the resident male altered his foraging pattern so as to maintain surveillance and food delivery to the nest. Throughout this study, competition for nest sites by hornbills was high. In areas where the ratio of hornbill pairs to nest sites was highest, the proportion of unsuccessful nests was greater (Kalina, unpubl. data). Hornbills that did not acquire a nest at the beginning of the breeding season sometimes attacked nests that were already occupied. Some intruders successfully drove off nesting hornbills, then took over the nest site. In this study, the nest-seal appeared to be used more often in defense against conspecifics than against other potential predators. Kemp (1970), whose studies focused on savannah hornbills (M species in particular), hypothesized that the mud barricade at the nest entrance served primarily to keep interspecific predators out of the nest. Unlike the black-and-white casqued hornbill, individuals of Iggkgs spp. appear not to compete for nest cavities (Kemp 1976). Located in small trees, the nest cavities of Igckgs spp. are usually 3-4 m above ground and are relatively vulnerable to predation (Kemp 1976). In contrast, black-and-white casqued hornbill nests, located high in bales of large forest trees (8-30 m 123 above ground), are inaccessible to most predators (Kalina, unpubl. data). It seems, therefore, that, for this species, the nest-seal functions primarily to keep other hornbills out of the nest. 124 REFERENCES Andersson, M. 1978. Optimal foraging area: Size and allocation of search effort. Theor. Peggl. ngl 13:397-409. Caraco, T., S. Martindale and H. R. Pulliam. 1980. Avian flocking in the presence of a predator. Netgre 285:400-401. Ghiglieri, M. P. 1984. Ihe_thimeegzees_et_Kieele_Eetest. Columbia University Press, N.Y. Kacelnik, A., A. 1. Houston and J. R. Krebs. 1981. Optimal foraging and territorial defense in the great tit (Pergs major). Behav. Ecol. Sociobjol. 8:35-40. Kemp, A. C. 1970. Some observations on the sealed-in nesting method of hornbills (family Bucerotidae). Ostrich (suppl.) 8:149-155. Kemp, A. C. 1976. A study of the ecology, behaviour and systemat- ics of Tockus hornbills (Aves: Bucerotidae). Transvaal Mus. Memoir 20. Lendrem, D. 1983. Predation risk and vigilance in the blue tit. Behav. Ecol. Sociobiol. 14:9-13. Martindale, S. 1982. Nest defense and central place foraging: A model and experiment. Behav. Ecol. Seeigbiol. 10:85-90. Orians, G. and N. Pearson. 1979. On the theory of central place foraging. In D. J. Horn, G. R. Staris, and R. 0. Mitchell, eds., Analysis of Ecological Systems. Ohio State University Press, Columbus. Schoener, H. 1979. Generality of the size-distance relation in models of optimal foraging. Am= Net. 114:902-914. Sih, A. 1980. Optimal behavior: Can foragers balance two con- flicting demands? §eience 210:1041-1043. Struhsaker, T. T. 1975. lhe Reg gelgegs. University of Chicago Press, Chicago, Illinois. CHAPTER VI FUNCTION OF NEST-SEALING: WITH SPECIAL REFERENCE TO THE BLACK-AND-HHITE CASOUED HORNBILL (W W) W Most of the 53 species of hornbills, family Bucerotidae, live in forests (Kemp 1979, 1988; Kemp & Crowe 1985). Only 12 species inhabit savannahs. Hornbills vary in body size from the dwarf Tockus species (83-gram female) to the turkey-sized Bueorvus species (up to 6180-gram male) (Kemp in litt.). Their behavior is also diverse; a species may be sedentary or nomadic, territorial or nonterritorial, strictly monogamous or a cooperative breeder (Kemp 1979). Hornbills nest in natural cavities in trees, rock faces, or earth banks. With the possible exception of two species (genus Bucorvus), hornbills seem incapable of excavating their own nest cavities (Penny 1975; Kemp & Kemp 1980). They are, however, able to seal their nest entrances with mud and/or feces, a behavior unique among birds (Kemp 1970). Nest-sealing is accomplished as the hornbill moves its laterally flattened bill side to side in the nest entrance; sealing material is held in the bill tip and is plastered onto the wall of the entrance as it is squeezed out the sides of the 125 126 bill. The material is applied in thin layers until only a narrow vertical slit leads into the nest, and the female is imprisoned inside. She remains in the nest throughout incubation and all, or most, of the nesting period, which, depending on the species, lasts 1.7-4.7 months (Kilham 1956; Kemp 1979; Chapter IV). Confined females and nestlings eject or throw their feces out the nest entrance, keeping the cavity clean. Kemp (1970) has discussed some advantages and disadvantages of the sealed-nest for hornbills, as well as how this behavior may have evolved. It is generally implied, when not stated, that nest- sealing evolved in response to predation at the nest (Kemp 1970; Hatchtel 1982; Helty 1982; Collias 8 Collias 1984; Leighton 1986). Much evidence supports this assumption. Intuitively, authors argue that selection favors those birds able to shield themselves and young from predators (Kemp 1970). Alternative hypotheses for the evolution of nest-sealing have not been considered. The hypothesis that nest-sealing evolved in response to predation has been repeated so frequently that it has begun to be accepted without question. In this case, it is important to make the distinction of whether protecting the nest from predators is the specific function of nest- sealing or merely an incidental consequence. In this paper, I discuss evidence supporting the predation hypothesis and introduce three alternative hypotheses concerning the origin of nest-sealing behavior. These alternative hypotheses involve inter- and intraspecific competition for the nest and increased protection from weather. Information is presented that 127 demonstrates how, for black-and-white casqued hornbills (Bycanjstes sgbcylingrjegs), the nest-seal most often functions to guard against conspecifics. Eredatieg fiypgthesjs: Nest-Sealing Preteets A inst Pr dati n Predation has probably been a major evolutionary force determining the form and structure of nests (Collias A Collias 1984). Lack (1954) suggested that predation is responsible for over 75% of the losses of eggs and young of open-nesting birds. Bird species with enclosed nests tend to have greater nesting success than those with open nests (Nice 1957; Skutch 1966; von Haartman 1971; Oniki 1979). For example, in a study of 16 cavity-nesting bird species in a Central American forest, 60% of the nests produced at least one fledgling. Cavity-nesting species were almost three times as successful as species with open nests in the same forest (Skutch 1966). Among cavity-nesting species, those which seal the entrance may experience the lowest predation rates. Of 249 nesting attempts by lectus hornbills in South Africa, more than 90% were successful in fledging young (Kemp 1976), and predation was not a factor in most of the failures. In addition, other bird species nesting in open holes in the same area during Kemp’s study had a lower nesting success than did hornbills. Nest concealment, inaccessibility, and impregnability have been cited as anti-predator adaptations in avian nest-building (Jackson 1974; Skutch 1976). Hornbill nests have all of these 128 characteristics. Often situated high above ground in the bole of a large tree, the sealed entrances are visually camouflaged and difficult to reach. The sealing material dries brick hard. This makes the nest nearly impregnable, especially as the narrow entrance is guarded from inside by the sharp, snugly fitting beak of the female hornbill. The funk hole, or chimney-like escape tunnel, inside many hornbill nests also works in conjunction with the sealed entrance to help in avoiding predators. Carnivores, apes, monkeys, snakes, and raptors are potential hornbill predators. Yet elephants (Kemp 1976), humans (Hannenburgh 1980), and other hornbills (Poonswad et a1. 1983; Kalina in press) are the only species known to break open hornbill nests. Only the bushmen open the nests. with the intention of' eating the birds, although nest damage usually leads to the death of the nestlings anyway. Elephants inadvertently split the walls of one Ieetes nest when they bent the branch where it was located. Despite attempts by the nesting hornbill to seal up the cracks in the walls, the nest was presumed to have been robbed when Kemp (1976) found it empty a week later. Perhaps it was an accident of this sort that permitted access to a naked black-wattled hornbill (getetegymne ettete) chick by a crowned eagle (Stephanoaetgs egrgnatgs). Keith (1970) saw the eagle carrying an unfledged chick which should certainly have been confined to its cavity. It is unlikely that the eagle could have broken the nest-seal, but it may have reached into a cavity after the female broke out for any of several reasons. 129 Concerning his study’ of three Ieetys species, Kemp (1970) remarked that the "most striking aspect of the study was the lack of predation occurring at the nest." During my 3-year study (65 nest years; 1981-1984) of B. sgbeyltngtiegs ‘Hl Kibale Forest, Uganda, interspecific predation on nestlings was not observed. In short, studies of hornbills have indicated that interspecific predation at the nest is unusually low when compared to that for birds in general. It seems probable that the primary reason for this is that the nests are sealed. Interspecific-Competition Hypothesis: Nest-Sealing Protects Against Interspeeific Competition Competition for nest sites has been noted among birds (Lawrence 1966; Burger & Shishler 1978; Burger 1979; Newton 1979; Trivelpiece & Volkman 1979; Short 1979, 1982) and between birds and other animals (Dennis 1971; Kemp 1976; Short 1982). Competition for nests has probably had a strong influence on the evolution of nest diversity and nest-building behavior (Collias & Collias 1984). Since suitable nest cavities can be limited in supply, competition for breeding sites may be magnified by cavity-nesting species. For example, populations of the house wren (Trgglggytes geese) (Kendeigh 1934) and the pied flycatcher (Fieedula hypgleuca) (von Haartman 1971) increase in response to provisions of nest boxes. Interspecific aggressive encounters at cavities are often intense (Sielmann 1959; Lawrence ‘1966; Kilham '1968, 1969, 1972; Short 1979, 1982). Such competition is likely a factor in the 130 evolution of size-class divergence among different species of birds. Short (1979) suggested that different-sized entrance holes typical for each woodpecker species act to minimize loss of nests to competitors. Hhen competition for nest sites occurs among open-nesting birds, there is a general precedence of larger species over smaller ones. For instance, on certain cliffs in Britain, kestrels (Eeleen tinnuneulus) are usurped by peregrine falcons. (meg peregrjngs) that are in turn supplanted by golden eagles (AguLa ehrysaetos) (Collias & Collias 1984). Similarly, Chinstrap penguins (Pyggseelis antartica) take nest sites from Adele penguins (Pyggsceljs egeliee) (Trivelpiece & Volkman 1979), and herring gulls (LQLUS ergentatus) displace the smaller laughing gulls (Legs etricille) (Burger & Shishler 1978; Burger 1979). If entrance holes are large enough, large birds dominate smaller birds in cavity-nests as well. Hairy woodpeckers (Pjegjdes villosus) replace downy woodpeckers (Pieoides pubescens) (Short 1979), and starlings (Stgyegs, vulgarjs) usurp eastern bluebirds (we stalls) (Zeleny 1977). Small body size is advantageous, however, when it permits a bird to fit into a hole small enough to keep larger competitors out. Biologists have used this concept to help increase bluebird populations across the United States. Artificial nest boxes with exact entrance-hole dimensions effectively prevent the larger starlings from taking over the nests. 131 European nuthatches ($1111..§0£92§£§) prevent intrusions at their nests when they reduce the entrance-hole dimensions with mud. The small nuthatch would be unable to prevent the European starling from taking over its nest were it not for this barrier (Sielmann 1959). The wall is built so that the entrance is just large enough for the nuthatch to slip through and small enough that the starling cannot. Sielmann (1959) described the interaction between representatives of these two species at a nuthatch’s nest as a "battle." As the nuthatch furiously plastered the nest entrance, a starling pecked away at the barrier. The starling was only able to break away pieces of mud that were still damp, and after several days it abandoned the attempt to take over the cavity. The nuthatch had successfully protected its nest by reducing the size of the entrance hole. At least one other nuthatch (Sitte negmeyet), the Hallcreeper (Tricodroma murarje), and two swallows (Petrgehelidon nigtjeens and Eyegee ehelybea) also reduce the entrances to nest cavities with mud (Rowley 1970). Hornbills also» experience interspecific interference at the nest. Unsealed Toekus spp. nests in South Africa may be occupied by the pearl-spotted owl (fileucigium perlatum), purple roller (geracius neeyie), lilac-breasted roller (Coreejus eeggete), greater starling (tempretgtnis eustraljs), monitor lizard (Varanus elbggglaris), python (Pytheg seeae), and bees (Kemp 1976). Bees also occupy unsealed B. subcylindrjegs nests in Uganda, as do red-legged sun squirrels (Heligsejgrgs :ufgbraehigm) and scaly-tailed flying squirrels (Anomelurgs deteieegs) (Chapter IV). Because of the 132 prolonged hornbill nesting period, and the protracted season in Kibale Forest when many of the larger animals breed (Butynski 1988; Chapter IV), arboreal carnivores, prosimians, parrots, and many other species are also potential competitors for hornbill nest cavities. Grey parrots (Psittagus eyjtheegs), for example, nested simultaneously with B. sgbcylindriegs, occupying a nest cavity below that of the hornbills’ in the same tree (pers. obs.). The most serious competition might be expected between sympatric hornbill species with similar nest-cavity requirements. This was suggested in Kemp’s (1976) study of three leetgs species when it was found that species interchanged nest holes. Interspecific competition did not occur, however, since there was no shortage of appropriate nest sites (Kemp 1976). During a study of four forest hornbill species (Bucergs biegrgjs, Bhytjeeres undulatgs, Anthreegcergs elbirgstris, and Etjlglgemgs tittelli) in Thailand, however, competition for nests did occur among the same or different species (Poonswad et a1. 1983). By attacking the nest and breaking the seal with their bills, intruding hornbills caused nesting birds to abandon the site. WWW Prote ain t tr i om eti i n Because of the similarity of their requirements, the most intense competition often occurs between birds of the same species. Agonistic interactions between individuals. oru pairs of“ the same species are COlll'llOll at nest sites early in the breeding season. 133 Possession of preferred nest sites by Canada geese (mm eenedegsis) is attained according to dominance after successive replacements of' different individuals or pairs (Collias & Jahn 1959). Intraspecific competition is undoubtedly magnified for birds with highly specialized requirements, such as those nesting in tree cavities or on cliff edges. Intraspecific competition for nest sites has contributed to the decline of at least one endangered species: the Puerto Rican parrot (Amegeee yittete). One instance of combat between pairs of this tropical rainforest cavity-nesting bird over a single nest site resulted in neither pair producing eggs and injury to the adult birds (Wiley 1980). Nest sites are apparently limiting for B. subeylindriegs in Kibale Forest, Uganda, since competition between conspecifics for traditional sites is evident and often severe (Kalina in press; see Appendix 6-A for descriptions of interactions at nests). Suitable nest cavities seem rare, as the ratio of nonbreeding hornbill pairs to breeding pairs can be as high as 3.4:1 during the pre- and early breeding period (July, Aug., Sept.; Chapter II). Nonbreeding pairs that do not find suitable cavities sometimes guard and display at unsuitable tree holes that may have entrances that are too large, too small, face skyward, or offer no perch at the entrance. The ratio of sealed hornbill nests to these unsealed "display cavities" is as high as 1.06:1 in some areas (Chapter IV). B. syteyljngyiegs are large birds (1.3 kg; Kemp 1979) which, like most other hornbills (Moreau & Moreau 1941; Kemp 1976), cannot excavate their nests and 134 which depend upon traditional sites. The same B. sgbcylingricus pair may occupy the same nest cavity each consecutive breeding season (Chapter IV). At least one pair in Kibale used the same nest for 7 years. Heavily worn rims at entrance holes suggest that some B. subcylindricus nests have been used for decades (pers. obs.). To maintain ownership at such l'preferred" nest sites, resident B. subeylindrjegs must often defend the nest against conspecifics. Fights between pairs at nests occur before and/or after the nest is sealed. At a sealed nest the intruding hornbill(s) will cling to the rim and attempt to knock out the mud wall that blocks access to the nest (Kilham 1956; Appendix 6-A). Such attacks are particularly prominent when the resident male is foraging away from the nest, an activity that can take 1 h or more (Kalina in press; Chapter IV). The nest-seal serves to reduce the impact of such attacks until the male returns to drive off the intruders and help guard the nest. The advantage of the mud barricade is obvious at such times. Since B. subeylindriegs forage in widely spaced fruit patches (Chapters 11 and III), the nest-seal protects the nest and allows the resident (male to forage further away for longer periods of time (Kalina in press). Kilham (1956) described intraspecific interference at 6 of 16 B. subeylindriegs nests he observed in south-central Uganda. It seems reasonable to assume that every hornbill nest in Kibale would be visited by intruders (intending to supplant the resident pair) at least once during each 4.5 month nesting period.1 In the 1983-84 135 breeding season, at least 4, and probably 6, of the 10 nest failures were caused by intraspecific intrusions at the nest. Attacks at these nests were severe, repeated, and directed. The nest-seal was either removed by intruders or, eventually, by the resident female before she abandoned the nest. At one nest, an intruding female hornbill broke into the cavity and attacked the two young inside. This action led to the death of at least one of the two nestlings (pers. obs.). That the goal of intruding hornbills, even those committing infanticide, is nest occupation rather than some other cause (Mock 1984) was shown by the behavior of successful "intruding" birds. On at least three occasions in Kibale, intruding hornbills took over the nest site and guarded it after evicting the resident pair (Appendices 6-A, 6-B, 6-C). Micrpclimate Hypothesis: Nest-Sealing Erotects Against Adverse Neether Nests often protect birds from the physical environment, as well as from other animals (Collias & Collias 1984). Inside tree cavities, birds are sheltered from extreme changes in temperature (Kendeigh 1961). The 10 C temperature range inside a sealed Ipetps hornbill nest is, for example, relatively stable compared to the 29 C temperature range on the outside surface of the nest tree (White et a1. unpubl. report). Most of the temperature stability inside the nest chamber can be attributed to the insulation provided- by thick tree walls. The vertical opening to the sealed hornbill nest also helps to regulate the microclimate inside the cavity. The design of the entrance slit allows for good ventilation since warm 136 air escapes at the top part of the opening, while cool air enters at the bottom. White et al. (unpubl. report) recorded a 0.5 C temperature gradient between top and bottom parts of the opening. They also noted that 02 and C02 levels inside the nest chamber remained within safe limits because of the ventilation system. The sealed entrance also helps protect birds inside the cavity from wind, rain, and hail. Dampness in the nest is a limiting factor for the endangered Puerto Rican parrot (Wiley 1980), which is, like the hornbill, a tropical cavity-nesting bird. During his study of the red-cockaded woodpecker (Dendrocopus borealis), Dennis (1971) noted the advantages that a small nest entrance provides during heavy rains. After 15 red-cockaded woodpecker nests were enlarged by pileated woodpeckers (Byypeppps pjleetus), about half of them were rendered useless by flooding with rainwater. Hornbill nesting behavior is influenced by the effects of wetness in the nest. At least one nest hole was probably abandoned as a result of flooding in the previous season (Kemp 1976). Many bird species select cavities with openings that face away from the direction of approaching storms (Ricklefs & Hainsworth 1968; Balgooyen l973; Conner 1975; McEllin 1979). Cactus wrens (Bampylorhynehus brunneieapillgs) favor nests with openings facing away from cold winds (Ricklefs & Hainsworth 1968), and Austin (1974) found that nest orientation was directly related to fledging success for this species. Parents and young probably conserve energy by avoiding the chilling effects of prevailing winds and rain. 137 Nesting hornbills are faced with adverse weather conditions since the onset of breeding corresponds with the onset of the rains (Kemp 1973; Chapters III and IV). There is a potential conflict here since the rains also correspond with a relative abundance of arthropods (Kemp 1973; Nummelin 1986) and fruit (Butynski 1988). Ground hornbills (Bums spp.) nest during the rainy season in South Africa, and yet most nests open skywards in broken-off tree trunks (Kemp in litt.). Birds in this genus do not seal their nest entrances, but they do have another nest-building behavior that helps to prevent flooding of the nest. Ground hornbills load heaps of dry leaves into the nest as "lining." This serves to level out the uneven nest floors and provides either a drainage or sump for rainwater. There are no records of ground hornbill nests flooding, but there are some where the whole floor apparently collapsed (Kemp in litt.). Hornbill species that seal the nest entrances have not been found to select nests with entrances facing in any particular compass direction (Kemp 1976; Poonswad et al. 1983). In Kibale Forest, Uganda, B. subeylindrieus nest-cavity openings faced in all directions (X2 . 5.62, df . 7, p > .25; Fig. 6.1). It is likely, however, that the sealing of nest entrances permits occupation of nests that would be unsuitable were it not for the mud barricade which keeps out wind and rain. Hhere cavities are limited, selection should favor hornbills with this capability. The fact that some pairs selected cavities with entrances facing skyward indicates that nest holes for B. subcylindriegs in 138 N O O O 0 NW : NE 0 0 w 0.00 000.00 E SVV SE Predominant Direction e1 e e e e e e S Fig. 6.l.--Compass direction of B. sgbcyligdrjegs nest entrances in Kibale Forest, Uganda (n = 44). (If a nest hole was located midway between any of these eight compass bearings, it was recorded as the next clockwise compass direction.) 139 Kibale Forest are limiting and that the nest-seal permits occupation of otherwise unsuitable cavities. Although a significantly higher proportion of "skyward facing” nests failed (n . 10, 90%) compared to nests with "outward facing” entrances (n - 39, 31%; X2 - 24.62, df = l, p < .001), it is notable that, at least once, a nest with a skyward-facing entrance did not fail. The angle to this successful nest was the least severe of all skyward-facing openings. Nevertheless, without the additional protection provided by the nest-seal, it is likely that even this nest would have succumbed to driving rain. Discussion Nest-sealing by hornbills could function to protect nest occupants against predators, prevent inter- and intraspecific competition for nest cavities, or protect nesting birds against bad weather. Each of the possible advantages of sealed nests might be of various degrees of importance to individuals of different hornbill lineages (Kemp pers. comm.). Insight into the advantages and evolution of nest-sealing to various hornbill taxa may best be accomplished by taking diverse approaches to investigation. Future studies might focus on convergent evolution among unrelated bird species which reduce nest entrances with mud. Much information could also be gained from comparative field studies, particularly of hornbills and their relatives the hoopoes (families Phoeniculidae and Upupidae), diverse hornbill taxa, or a single hornbill species in different environments. 140 Convergent Evolution Ampng Unrelated Bjrd Speeies Further studies of the function of nest-sealing behavior might focus on the context and extent to which nuthatches, wallcreepers, and swallows perform the nest-entrance-reduction activity. Convergent evolution among such unrelated species as swallows and hornbills could give insight into the nature of relevant selection pressures. Accounts of intruding female tree swallows (Tachycineta pteplpy; Shelly 1934) and purple martins (m subis; Loftin & Roberson 1983) entering cavities to kill conspecific young are surprisingly similar to intrusions by conspecific B. subcylindricus (Appendices 6-A, 6-8, 6-C). Night infanticide be a selection pressure acting on those other Progne species (e.g., B. ehalybea; Rowley 1970) which reduce nest-cavity entrances with mud? Natural History Studies pf Hopppes Although the green woodhoopoe (Phoenieulus purpureus) has been the subject of much investigation (Ligon & Ligon 1978a, 1978b, 1983; Ligon 1981), little is known about the behavior and ecology of most species of hoopoes. Hoopoes are considered the closest relatives of hornbills (Sibley & Ahlquist 1972; Kemp 1979). Although hoopoes do not seal their nests like hornbills do, they have developed effective methods of nest defense by means of odorous secretions and aggressive behavior. A. comparison of environmental and social factors influencing nest defense in these two families of birds may 141 have bearing on understanding those selection pressures to which ancestral hornbills were subjected (Kemp 1979). m ar tiv ' d St di f H r '1 It may be significant that all observations of intraspecific interference at hornbill nests have been described for two species (B. supeylipdricus and B. unduletus) with a similar social system and feeding ecology (Chapter III; Poonswad et al. 1983; Leighton 1986). One species is African and the other SE Asian, yet both live in tropical rainforests and are monogamous, nonterritorial, semi- nomadic omnivores that feed mainly on fruit. Nesting birds of these species apparently benefit from protection the nest-seal provides while foraging males are collecting fruit far away from the nest. B. subcylindricus and B. undulatus socioecology seems to predispose their nests to threats of attacks by intruders since the males leave the nests unguarded and vulnerable for long intervals (1 visit/60 min). Ipstus socioecology may tend to prevent threats of intrusions because male Ipstus spp. return to their nests frequently to deliver single items of animal prey (l visit/8 min; Kemp 1976). Their nests are not, therefore, left unguarded for long (Kemp pers. comm.). In addition, Ms spp. maintain exclusive territories which are established before breeding. These territories, which are maintained intra- but not interspecifically, probably also serve to repel any potential enemies (Kemp pers. comm.) A Selection pressures acting on nesting ancestral hornbills would certainly be different depending on similarity of habits to either 142 of the two foraging patterns just described. Leighton (1986) suggested that foraging constraints are the most important ecological factor for Bornean hornbills and are responsible for the evolution of the varied hornbill social systems described in his study. He argued that cooperative breeding by two species and the maintenance of year-round territories by five of seven sympatric hornbill species function to protect fruit resources for these primarily frugivorous birds. Leighton (1986) assumed that suitable cavities were not in limited supply. He presented little data, however, on nesting birds. Although cooperative breeding and territoriality may be social behaviors developed for protecting limited food resources, they may also help prevent competition for nest cavities. Territories meant to protect food resources will also protect encircled nests. Perhaps the availability of fruit is the most important factor influencing the development of hornbill social systems, but it is equally reasonable to suggest that nesting constraints were evolutionarily significant as well. Analysis of the distribution, abundance, suitability, and use of nest cavities in the Bornean study site is necessary to continue the argument either way. Unfortunately, these data are not available at this time. Contrary observations concerning reproductive success of B. unduletus on Bornean and Thai study sites should prompt a new study' of ‘this species--with emphasis on the comparison of B. unduletus nesting behavior in two different natural communities. Poonswad et a1. (1983), as mentioned previously, observed 143 B. undulatus abandoning their nests after disturbance by conspecific intruders. Leighton (1986) observed nothing similar for this species in Borneo. In fact, several 3. unduletus pairs were able to nest successfully within territories of other Bornean hornbill species to which they were subordinate. Hhat environmental factors differ' between habitats ‘that are altering the behavior of this species? The identification of factors influencing reproductive strategies should be relevant to the testing of nest-seal hypotheses. ummar Most hornbills seal the entrance to their nest cavity so that the female and young are imprisoned inside. Previous speculations about this curious behavior focused on selective advantages obtained against interspecific predation of the nest occupants. Evidence presented here supports the predation hypothesis. Additional information was provided, however, which attempted to qualify three new hypotheses for nest-sealing (inter- land intraspecific competition for nests, and protection against the weather). Any of several possible benefits could result from chance rather than natural selection. Nest-sealing behavior might be attributable to selection for any one, or a combination of some or all, of these benefits. The last three hypotheses, unlike the predation hypothesis, are based on the assumption that nest cavities for hornbills are limited in supply. Evidence .was provided that supported this 144 assumption in Kibale Forest, Uganda, and perhaps in Thailand (Poonswad et a1. 1983) but not for hornbills studied in some other areas (Kemp 1976; Leighton 1986). If nest cavities are limited for hornbills, there should be selection for mechanisms that reduce intra- and interspecific competition for nests, or that make more cavities suitable as nest sites. The nest-sealing specialization could be attributed to a long period of selection in this regard. 145 REFERENCES Austin, G. T. 1974. Nesting success of the cactus wren in relation to nest orientation. Beugpy 76:216-217. Balgooyen, T. G. 1973. Behavior and ecology of the American kestrel (Falco sparverius L.) in the Sierra Nevada of California. Univ. Celif. Publ. 1001. 103:1-83. Burger, J. 1979. Competition and predation: Herring gulls versus laughing gulls. Queue; 81:267-277. Burger, J. and J. Shishler. 1978. Nest site selection and competi- tive interactions of herring and laughing gulls in New Jersey. Aut 95:252-266. Butynski, T. M. 1988. Guenon birth seasons and correlates with rainfall and food. In A. Gautier-Hion, F. Bourliere, J. P. Gautier, J. Kingdon, and R. M. Martin, eds. Primate Radiation: Evolutionary Biology and the African Buenons. Cambridge Univ. Press, Cambridge. Collias, N. E. and E. C. Collias. 1984. N st 8 il in and ird Behavior. Princeton Univ. Press, New Jersey. Collias, N. E. and L. R. Jahn. 1959. Social behavior and breeding success in Canada geese (Branta eanadensis) confined under semi-natural conditions. gut 76:478-509. Conner, R. N. 1975. Orientation of entrances to woodpecker nest cavities. Aug 92:371-374. Dennis, J. V. 1971. Species using red-cockaded woodpecker holes in northeastern South Carolina. Birg Bending 42:79-163. Jackson, J. A. 1974. Gray rat snakes versus red-cockaded woodpeck- ers: Predator-prey adaptations. Aug 91:342-347. Kalina, J. in press. Nest intruders, nest defense and foraging behavior in the black-and-white casqued hornbill (Bypanistes subcylinerieus). 1015- Keith, A. R. 1970. Crowned hawk-eagle raids hornbill nest. J. of E. Afr. Nat. Hist. Sec. end Net. Mus. 1:64. Kemp, A. C. 1970. Some observations on the sealed-in nesting method of hornbills (Family Bucerotidae). Ostrich (suppl.) 8:149-155. 146 Kemp, A. C. 1973. Factors affecting the onset of breeding in African hornbills. Bree, Ifitp Int. Que, Beng.: 248-257. Kemp, A. C. 1976. A study of the ecology, behaviour and systemat- ics of Ipetus hornbills (Aves: Bucerotidae). lreusveel Mus, fiemoir Ne, 29. Pretoria, South Africa. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. The Living Bire 17:105-136. Kemp, A. C. 1988. Bucerotidae. - In E. Urban, 5. Keith, and H. Fry, eds. [h e Bi [gs pt Afrjee, vol. . Academic Press, London. Kemp, A. C. and T. H. Crowe. 1985. The systematics and zoogeog- raphy of Afrotropical hornbills (Aves: Bucerotidae). In K. L. Schuchman, ed., Proc. Intern. SyupesAfrigen Vertebr. Bonn. Kemp, A. C. and M. I. Kemp. 1980. The biology of the southern ground hornbill, Bucorvus leadbeeteri (Vigors) (Aves. Bucero- tidae). Ann. Trensvael Hus. 32: 65- 100. Kendeigh, S. D. 1934. The role of environment in the life of birds. Eeolpgjee! Mpnog. 4:299-417. Kendeigh, S. C. 1961. Energy of birds conserved by roosting in cavities. Hilson Bull. 73:140-147. Kilham, L. 1956. Breeding and other habits of casqued hornbills (Bycanistes subeylindricus). Smitpspn. Mjsc. Belles. 131:1-45. Kilham, L. 1968. Reproductive behavior of hairy woodpeckers, II. Nesting and habitat. leson Bull. 80:286-305. Kilham, L. 1969. Reproductive behavior of hairy woodpeckers, III. Agonistic behavior in relation to courtship and territory. Hilson Bull. 81:169-183. Kilham, L. 1972.f Habits of the crimson-crested woodpecker in Panama. Hilson Bull. 84:28-47. Lack, D. 1954. e N r ula 'on Ani 1 N mb r . Clarendon Press, Oxford. Lawrence, L. D. K. 1966. The comparative life-history study of four species of woodpeckers. anith, Mpnog. 5:1-156. Leighton, M. 1986. Hornbill social dispersion: Variations on a monogamous theme. pp. 108- 130. In D. I. Rubenstein and R. N. Hrangham, eds. i ' ' . Princeton Univ. Press, New Jersey. 147 Ligon, J. D. 1981. Demographic patterns and communal breeding in the green woodhoopoe (Ehoeniculus purpureus). pp. 231-243. In R. 0. Alexander and D. H. Tinkle, eds., Neturel Belection end Soeial Behevipr: Regent Researeu eug New Theory. Chiron Press, N.Y. Ligon, J. D. and S. H. Ligon. 1978a. Communal breeding in the green woodhoopoe as a case for reciprocity. Netuke 176:496- 498. Ligon, J. D. and S. H. Ligon. 1978b. The communal social system of the green woodhoopoe in Kenya. Liyiug_Biye 17:159-197. Ligon, J. D. and S. H. Ligon. 1982. The cooperative breeding behavior of the green woodhoopoe. Seient. Am. 247:126-134. Ligon, J. D. and S. H. Ligon. 1983. Reciprocity in the green woodhoopoe (Ehpenieulus purpureus). Anjm, Behev. 31:480-489. Loftin, R. H. and D. Roberson. 1983. Infanticide by a purple martin. Hilso l . 95:146-148. McEllin, S. N. 1979. Nest sites and population demographics of white breasted and pigmy nuthatches in Colorado. Epngor 81:348-352. Mock, D. H. 1984. Infanticide, siblicide and avian nestling mortality. pp. 4-30. In G. Hausfater and S. B. Hrdy, eds. Infenticide: Comparetive_end Evolutiunery Perspeetjves. Aldine, N.Y. Moreau, R. E. and H. M. Moreau. 1941. Breeding biology of the silvery-cheeked hornbill. Auk 58:13-27. Newton, I. 1979. Pppuletjpn Ecplogy pf Reptprs. Buteo Books, Vermillion, S. Dakota. Nice, M. 1957. Nesting success of altricial birds. Auk 74:305- 321. Nummelin, N. 1986. The seasonal fluctuations of forest floor insect densities on the areas of different forestry practices in Kibale Forest, Hestern Uganda. f. n Tr ical Entomolpgy, Nairobi. Oniki, Y. 1979. Is nesting success of birds low in the tropics? Bietkupiee 11:60-69. Penny, C. G. 1975. Breeding the Abyssinian ground hornbill at San Diego Hild Animal Park. Ipt’l pr Xypk 15:111-115. 148 Poonswad, P., A. Tsuji and C. Ngarmpongsai. 1983. A study of the breeding biology of hornbills (Bucerotidae) in Thailand. Pro- ceeding BelagourZIFEB Symposium on Breeding Birds in Ceptivity, pp. 239-265. Ricklefs, R. E. and F. R. Hainsworth. 1968. Temperature dependent behavior of the cactus wren. Eeplpgy 49:227-233. Rowley, I. 1970. The use of mud in nest building--A review of the incidence and taxonomic importance. Bstrjeh (suppl,) 8:138- 48. Shelly, L. O. 1934. Tree swallow tragedies. Bird Bending 5:134. Short, L. L. 1979. Burdens of the Picid hole-excavating habit. kilson Bull. 91:16-28. Short, L. L. 1982. Npgdpeckers of the Hprld. Delaware Museum, Delaware. Sibley, C. G. and J. E. Ahlquist. 1972. A comparative study of the egg white proteins of non-passerine birds. Bull. Peabody Mus. uet. flist. 39. Sielmann, H. 1959. My Yeer with the Nppdpeckers, translation by S. Lightman. Barrie and Rockcliff, London. Skutch, A. F. 1966. A breeding bird census and nesting success in Central America. ijs 108:1-16. Skutch, A. F. 1976. Parent Birds and Their Young. Univ. of Texas Press, Austin. Trivelpiece, H. and N. J. Volkman. 1979. Nest-site competition between Adelie and Chinstrap penguins: An ecological interpre- tation. Auk 96:675-681. von Haartman, L. 1971. Population dynamics. pp. 392-461. In D. S. Farner and J. R. King, eds. Avjep Biplpgy 191, 1. Academic Press, N.Y. . Hannenburgh, A. 1980. Ine_Bushmeu. Country Life Books, N.Y. Hatchtel, P. S. 1982. Hornbill, a most peculiar bird. Int’l kildljte 12:36-41. Helty, J. C. 1982. [he Ljfe pf Bikes. Saunders College Publish- ing, Philadelphia. 149 White, F. N., J. Kinney, H. R. Siegfried, and A. C. Kemp. Thermal and gaseous conditions of hornbill nests. . r N ’ Geog, Sec. (unpubl. report). Hiley, J. H. 1980. The Puerto Rican parrot (Aeezppe yjttete): Its decline and the program for its conservation. p. 133. In R. F. Pasquier, ed. Conservetipn of New Npklg Parrots. ICBP Tech. Publ. No. l. Zeleny, L. 1977. Nesting box programs for bluebirds and other pas- serines. pp. 55-60. In S. A. Temple, ed. Eudangered Birds. Univ. of Hisconsin Press, Nadison. 150 APPENDIX 6-A DESCRIPTIONS OF BYCANISIES SUBCILINQRIEUS BEHAVIORS, SOUNDS, AND VOCALIZATIONS MENTIONED IN APPENDICES 6-8 AND 6-C ehaviors Beeg;sheke--The rapid movement of the bill side to side as is done during nest-sealing, only the head-shake is a display performed without mud. The head-shake is executed by males and females as a component of courtship displays (directed at each other) and during intrusions at the nest by conspecifics (defending residents directing display at intruder). Head-shake is often observed near or on the nest, but is also performed away from the nest. Sometimes combined with chutter or purr vocalizations. Feathers above the eyes may be erected during head-shakes performed for courtship, but they are flattened against the head during agonistic encounters. During intrusions by conspecifics, the male often will strike his bill against a branch, break off a 4-9 inch long piece of bark or wood, and head-shake while holding the bark or wood in his bill. Spunds and Vocalizetjons Sounds and vocalizations have been tape-recorded and will be spectrographically analyzed. The hornbill’s vocal repertoire is extensive and is not fully represented here. Brumming--Rapid drumming sound made by female from inside the nest as she head-shakes so that her bill hits against one wall of the nest entrance and then the other. If the sound resonates in the cavity, it can be heard 300 m away. Usually heard during nest- sealing, but also during agonistic encounters with conspecific hornbills intruding at the nest. Bhutter--A soothing "chuckle" (re: Kilham 1956) usually made by a male while transferring food to his mate. Flying long-caI|--The long-call (described below) given during flight, which causes a change in the rhythm as it is timed with the beats of the wings. Can be heard 2 km away. flunk--Commonly heard, brief calls which may be soft (quiet contact calls between mates), moderate or loud, depending (M1 the context. Soft honks often made by males before they approach the nest to ‘feed. Honks may be given singly or in very rapid succession. fluh-huh--An "ugh" sound given in twos. 151 Lung-eel --The drawn-out 'ka-waaack" most characteristic of this species. A call that descends and then ascends in succession. Used often and in many circumstances (e.g., pairs advertising their territory at the nest and hornbills at fruiting trees). Can be heard for 2 km. Long-calls give the listener the location and activity of the bird (whether it is stationary or moving). The intensity, timing, and duration of the call, and whether or not it is combined with repeated hi-pitched screams, indicate whether the caller(s) are foraging or are at the nest site. Bonkey-hoo--This monotone, muffled call sounds like it could be made by a monkey. It is an alarm call made by hornbills as other hornbills approach or as a human approaches. The first note (hoo) is longer than the hoos that follow in rapid succession. Purr--A soft sound made by females most often while accepting food or during courtship displays. Repeeted hi-oitched scree_--Shrill rapid screams most often given by males and females during territorial counter-calling bouts. Often following long-calls and rapidly repeated loud-honks. Because females have higher-pitched voices than males, this call is most expressive among females. Ruah--Ascending scream of panic and pain heard only once, from resident male during a physical attack by an intruding pair of hornbills. Bumbling--Low, guttural sound. Beef--Soft, full sound made with bill closed. Vocalizations Made N tlin s Ety--Squeaky whimpers. Hiss--As the hiss of a snake. The chick flattens the head feathers, opens the bill wide in defensive posture, exposing the red spiny tongue, and hisses. Given to humans when approached while on the ground. Since chicks on the ground are totally vulnerable to predators, this is the only form of defense. Rhythmic sgueak--Given by nestlings and females while inside the nest. Sounds like the squeaking of a wooden sawhorse as it is being rhythmically rocked under the movement of a saw. Associated with feeding, and lack of feeding, by the male. 152 --Ascending and then descending distinct notes given by females and nestlings during periods of great excitement. Females often make this call loudly if intruders arrive while the resident male is far away from the nest. Apparently a long-distance contact ca . Bee--Immature version of the long-call. Bee--Hi-pitched call which is given repeatedly by chicks when the male parent approaches or lands on the nest. ”Begging” call which continues throughout feeding and often for a short time afterward. 153 APPENDIX 6-B EXAMPLE OF INTRUDING BYCANISTES SUBCYLINQRICUS SUPPLANTING RESIDENTS EARLY IN THE NEST-CYCLE (KIBALE FOREST, UGANDA) Nest location: HH/5; Undisturbed forest in K-30, Kanyawara Date female sealed inside nest: Oct. 1, 1983 Number of days female sealed in before nest failure: 1-4 Time is read: hours: minutes(seconds) Oct. 2, 1983 The resident pair (RP) at nest is the same one that nested here successfully in 1982-83. 7:50 RP male is long-calling in the nest tree. RP female is inside the cavity, where she spent her first night last night. She is tapping the remainder of the mud-seal in place with her bill. 8:OO(45)-8:Ol(58) RP male lands on nest, clings vertically to the nest rim, and head-shakes with bill pointed into the nest hole, while making a low rumbling sound. No food is delivered to the nest. Male flies off nest, into the nest tree crown. He woofs and counter long-calls to hornbills which are long-calling 250 m away. 8:05(58)-8:O9(27) RP male lands on nest, head-shakes, puts his bill deep into nest hole, then repositions his head as he apparently tries to see inside cavity. 8:09(26) RP male makes a low rumbling sound and flies to 15 m from nest and long-calls. 8:15 All is calm when I leave the area. RP female is still tapping mud from her position inside the tree. 10:40 I check the nest again and the situation is the same as at 8:15. The female is still tapping from inside the nest. Oct. 4, 1983 I return to the nest site after a technician has informed me that there is a conflict at the nest. - 15:59 I arrive at the site to find that the RP female has broken out of the nest and the RP is confronting an intrusion by conspecifics. One hornbill pair long-calls 5 m from the nest, and a second hornbill pair is 50 m from the nest tree. Unfortunately, neither pair was seen clearly enough to identify at the time. Dried 154 mud is smeared on outside of nest entrance. Both pairs flutter back and forth, flying toward and away from each other. Vigorous vocal exchanges include mmnkey-hoos, long-calls, and repeated hi-pitched screams. 16:19 I get my first good look at the intruding pair (IP) when they fly (long-calling in flight) from approximately 70 m away into the nest tree. The IP hornbills are easily distinguished from the RP hornbills, as the male has unique casque markings and the female’s casque and bill are unusually long with a wide gape (even when her mouth is closed). The female also has a black spot on her white abdomen. He fleshy eye rings are bright red and very swollen, an indication of her excited state. She has mud on the top of her bill at the top, suggesting that she may have knocked out the RP nest- sea . 16:19-16:57 IP and RPs continue to counter long-call. RP moves farther away from the nest. 16:57(28) IP male lands on nest rim. 16:57(43) IP female lands on nest and wildly head-shakes into the nest hole, tapping her bill on its edges. IP male self-preens while watching the female, then makes soft huh, huh sounds. 17:00 IP flies to 1.5 m from nest. They purr, perform the head- shake display with bills tapping together, then make huh, huh vocalizations. 17:02(22) The IP female lands on the outside rim of the nest and head-shakes 10 times. The IP male makes huh, huh sounds and the female head-shakes, tapping her bill in the hole as if she had mud in her bill. l7:O3(13) While the IP female flies off the nest and lands 0.5 m away, the IP males makes huh, huh sounds. l7:O4(OO) IP female makes four low, drawn-out long-calls and IP male huh, huhs. 17:05(OO) IP continues to call and remains within 50 m of the nest. RP has flown to > 100 m from the nest and they occasionally counter long-call with the IP from distances of up to 250 m from the nest. 18:18 I leave the area with IP in the nest tree and RP still away. 155 Oct. 5, 1983 17:36 I arrive at the nest site to find the IP long-calling from in the nest tree. IP female’s eye rings are now tan-grey and have lost their swelling, an indication that she is less excited than yesterday. Red mud stains her back and is smeared on her body where she was spotless yesterday. She has obviously been inside the nest cavity. The IP males is breaking off bits of bark from the branch where he is perched, and he is manipulating it about in his bill before drgpping it and breaking off another piece. RP hornbills are not aroun . 17:40-17:59 IP male and female quietly preen themselves and then each other. IP male occasionally utters soft woofs. 17:59 I leave, having been convinced that they are the new residents at that site. - Oct. 10, 1983 15:53-17:00 IP male fed IP female while she was inside the nest cavity. They attempted to seal the nest with mud, but the female later joined her mate outside the cavity. Both were long-calling from their perch on a limb of the nest tree when I left. Nov. 2, 1983 IP still resident, guarding the nest site, but the female is not sealed inside the nest cavity. Nov. 21, 1983 18:11-18:39 Five individuals in a sub-adult hornbill group are perched in the nest tree. I? is not seen in area. 551m The RP attempted to nest in the HH/S cavity as they had done successfully the year before. On Oct. 2, 1983, the RP female was sealed into the nest. The RP was supplanted at the nest site by IP 4 days after RP female had initially entered the nest. The RP abandoned the site to IP, which then occupied the site for at least 1 month. IP attempted to, but did not successfully, seal the nest. 156 APPENDIX 6-C EXAMPLE OF INTRUDING BYCANISTES SUBCYLINBRICUB SUPPLANTING RESIDENTS LATE IN THE NESTING CYCLE (INFANTICIDE) Nest location: N/9.5; Undisturbed forest, Ngogo Date female was sealed inside nest: < Sept. 26, 1983 Number of days female sealed in before failure: 92+ Jan. 3, 1984 Observations made from: 10:30-12:05; 14:40-18:05 This nest is the only one observed in the wild to contain two nestlings. A male and female chick could be seen inside the cavity. At the time of this intrusion by conspecifics, the nestlings were approximately 1-2 weeks from fledging. 10:33 The male of the resident pair (RP) is found clinging to the nest hole, pecking away the mud of the nest-seal when an intruding pair (IP) of hornbills chases him off the nest. The IP lands on the nest, begins pecking at the seal, and then the RP male chases them away. 10:44 RP male lands on the nest and resumes opening it by pecking at the mud. (Note: When nesting females and young are ready to leave the nest, they typically break the seal themselves.) At 10:45, RP male finishes opening the seal and RP female comes out of the nest. Both RP male and female long-call loudly and fly around the immediate area of the nest site together. 'The chicks remain inside the tree cavity, peck at the nest rim, waa-call, and occasionally stick their bills out the hole. Although the RP rattempts to approach the nest, 1P aggressively guards the nest tree. IP male and/or female lands on the nest rim, attempt to feed the chicks inside, and then peck and jab at the nest rim and at the chicks. IP can be distinguished from RP based on appearance and sound. IP female has a raspy voice. Her tail is frayed, an indication that she has also been sealed into a nest this season. Hhen IP is on or near the nest, the chicks are quiet. Hhen RP long-calls or honks, the chicks respond immediately with a waa call or 6-note call. RP keep their distance when IP is in the immediate vicinity of the nest. RP male "sneaks" in to feed the chicks in the nest when IP is away from the nest site, presumably on foraging trips. During the hours of observation, intruders (IP male, female, or both together) land on the nest rim on 9 separate occasions, for a total of 19 min., 22 sec. IP male drops 2 dark unidentified items and 4 157 Ficus sp. fruits into the nest. RP male lands on the nest 9 times, for a total of 12 min., 22 sec. and feeds 128 Biospyrps epyssiniea and 5 Mimusops Bagspewei fruits into the nest. Jan. 4, 1984 Observations made from: 7:30-13:02 7:30 I arrive to find IP calling in the crown of the nest. The male chick is pecking on the inside rim of the nest hole. RP female has re-entered the nest cavity, and the three hornbills are moving inside. 7:35(56)-7:40(58) IP female lands on nest. RP female head-shakes in hole with bill parted, even before IP female lands. IP female and RP female clack bills and head-shake. IP male perches 4 m away, monkey hoos, woofs, and long-calls. IP female jabs RP female, head- shakes, and pecks hard on the nest rim. Chicks and RP female 6-note ca . 7:40(36)-7:44(35) IP male lands on nest four times, once joined by IP female. All birds head-shake and counter-call to RP male, who is loud-honking 100 m away. Vigorous and excited vocal exchanges follow between IP and RP. The chicks are active in the nest hole. One throws a wood chip out the hole and the other chick then does the same. One nibbles on the other chick’s bill tip. Chicks wee, 6-note call, and then make repeated hi-pitched screams as IP and RP male jockey for positions around the nest. The intensity of vocal exchanges increases to a high level. 8:21(OO)-9:25(20) and 8:46(29)-8:53(10) IP female is on the nest, head-shaking, pecking at nest rim, and attacking the RP female by jabbing with her bill. IP male and RP are screaming at a high level from their positions. Finally, the RP female pecks so hard at the IP female that IP female loses her balance. She pecks back with much force and flies off the nest. - 8:59(27) Male chick in nest is holding a long wood chip in bill tip and is manipulating it back and forth in his bill as hornbills do to crush their prey. This behavior is the same as used by adult males during displays against intruders, and also during courtship displays. 9:02 The male chick drops that piece and picks up another 3-inch piece of wood to do the same thing. 9:O6(38)-9:O7(OO) IP male is on the nest, holding an insect in his bill tip. He head-shakes but does not feed the chicks. He flies to 5 m. RP male lands at 10 m and IP chases him. 158 Chicks appear agitated and active in nest with much wood flipping. IP female lands on nest and RP female and both chicks face her in defense posture with bills parted. All head-shake and peck at each other. RP male dives low through the trees, swooping near the nest. Much vigorous counter-calling follows. RP male is no match for the IP, and the fighting at the nest continues. RP female knocks IP female so hard that IP female screams and falls off nest rim. IP female does the vigorous fighting. IP male head-shakes and backs her up by remaining near her, but he does not jab the RP female as viciously as IP female does. 9:47 RP male is fluttering about 70 m from the nest when IP fly long-calling to him, perch, huh, huh, and then chase him to a spot 15 m from me, low (about 3 m up) in the vegetation. They attack him together and he tumbles to the ground. RP male cries a ruah ruah ascending scream, a sound of panic and pain. I have never heard this vocalization on any other occasion. ' 11:02 RP male is self-preening 50 m from the nest. His tail is frayed and feathers are in disarray. IP chases him and RP female drums her bill on the walls of the nest entrance. Chicks wee. Today, one or both intruding hornbills are on the nest on 21 separate occasions, for a total of 46 min., 42 sec. RP male does not land on the nest, and no food is fed in to RP female and chicks. Jan. 5, 1984 Observations made from: 8:00-12:05 RP female is still inside the nest guarding her chicks against frequent attacks by IP. RP male is able to feed the chicks a small amount (approximately 5 small items) of food on one brief occasion. Jan. 6, 1984 Observations made from: 7:59-15:05 7:59 I arrive at the nest site to find RP female 6-note calling and repeatedly screaming inside the nest cavity. I can see the chicks as they waa call. IP are 4 m from the nest, woofing and long- calling. RP male is fluttering about 100 m away. . 8:02(25)-8:02(55), 9:OO(20)-(40) IP male is on the nest jabbing at RP female. 9:27(50)-9:28(20) RP male lands on nest and dumps about 10 B. abyssiniea fruits very quickly into nest. IP male approaches and RP male dives off the nest, only to be chased by IP female. 159 9:38(59)-9:49(38) IP female is on nest fighting with RP female. RP female’s bill is wide open in defense posture while IP female jabs aggressively with bill closed. 9:48(48) IP male also lands on nest. 9:49(08) RP male lands in tree crown and IP flies off nest. 9:50 RP male flies out of nest tree. lO:Ol(47)-lO:02(30) RP male lands on nest and feeds 12 fl. pegshewei fruits very quickly. He dives off nest as IP approaches him. Vigorous counter-calling by all hornbills continues off and on throughout. 10:05(29)-10:08(54) IP female is on nest fighting with RP female. RP female knocks IP female off nest. lO:lO(29)-lO:ll(33) IP female fighting with RP female at nest. RP male is fluttering about 80 m away. 10:30 All three hornbills in nest are tapping their bills on the nest rim very quietly. RP male loud-honks from 70 m away and chicks immediately waa loudly in response. Much counter-calling between all hornbills follows. ll:38(44)-11:41(44) IP on nest. IP female vigorously head-shakes and purrs coarsely while tapping her bill against the tip of IP male. Then IP female jabs at RP female. RP male is quietly approaching closer and closer to nest. 11:57(20)-ll:59(ll), ll:59(42)-12:Ol(52) IP male is on the nest jabbing at the RP female. 12:OO(06) RP male flies into nest tree and loud-honks with fig in bill tip. Chicks answer RP calls with waas and 6-note call. RP female long-calls and 6-note calls. IP female chases RP male down through the canopy and IP male follows in the chase. RP male’s bold approach to the nest attracts the IP’s attention away from the nest and toward himself. 12:36 Chicks wee and cry rhythmic squeak. 12:44(52)-12:45(50) IP male chutters and feeds 1 fl. pegshawei fruit. RP male dives at IP male but is chased away by IP female. 14:03 RP male is inching quietly toward the nest with a B. epyssiniee fruit in his bill tip. Hhen he is 20 m from the nest, IP dives down very quickly and chases him away. Loud vocal exchanges between hornbills continue. . During hours of observation today, one or both intruders are on the nest on 12 separate occasions, for a total of 30 min., 34 sec. RP male lands on the nest 3 times for a total of 2 min., 37 sec. He rapidly dumps several fruits at a time into the nest. He remains very alert and dives down off the nest quickly as soon as the wings 160 of approaching IP are heard. RP male feeds approximately 37 B. ab ssinic , 20 B. begshawei, and 1 insect into the nest. Jan. 7, 1984 Observations made from: 8:20-10:05, 15:08-17:40 8:20 IP is in the nest tree long-calling. A chick is wee calling in the nest. The RP female has again left the nest. She may have left in order to obtain food, since the RP male has been unable to provide adequate quantities during the intrusion. 8:30-8:32 IP female lands on nest rim, puts her head deep into the hole, and fights with a chick in the nest. 8:37 A chick pecks on the side of the hole, places his beak outside the hole, and cries softly. 8:39 IP female enters the hole. The chicks waa loudly, crying continuously. There is much movement and circling inside the cavity as the IP female is chasing and jabbing the chicks. The IP male is quietly perched in the nest tree crown. RP is long-calling from approximately 150 m away the entire time IP female is inside the nest. 8:48 IP female comes out of the nest hole and chicks stop vocalizing. 8:50 IP female lands on nest and pecks on rim, but does not enter. 9:03 IP male lands on the nest, looks inside, and soft-honks. Chicks do not vocalize in response. 9:04 IP male flies to the north. 9:15 IP female follows him. 9:19 Hhile IP is still away, the female chick sticks her head outside the nest hole, but she pulls it back in again. The male chick is not seen or heard. 9:20 RP male and female fly while long-calling to 20 m from the nest. RP calls to the chicks, and the female chick responds with waa calls. 9:33 Female chick squirms her way out of the nest and quietly lands 5 m below the nest on a teltis eutepeii branch. 9:40 The chick waa calls from her perch. The IP suddenly appears and flies at her, knocking her off the branch. At this time, RP flies away to approximately 100 m. The IP female attacks the chick with her claws, and the chick falls to the ground approximately 30 m 161 from the nest. As she falls, IP female follows close behind her. The chick waa cries as she hits the ground. The IP female lands atop her, biting and pecking at her. During this attack, all adults are quiet and only the screams of the chick can be heard. IP male remains 2 m away. 9:45 My technician rushes over to rescue the chick. As he approaches, IP monkey hoos, then flies quietly away to the north. Chick hisses, snaps, and waa calls as he picks her up. Then she immediately calms down and becomes very weak and docile. Chick has bloody puncture wounds on her back from the claws of IP female. She also has a large swelling on her right side and a superficial wound on the right leg. There are no broken bones. She weighs 750 grams and she accepts food readily. She is wounded and too weak to fly, and would surely fall victim to predators (intra- or interspecific). Judging from the immature condition of her feathers, I suspect that shed should have remained in the nest for one more week before e ging. There is no sign of the male chick, and I suspect he was badly injured or killed inside the nest by the IP female. He was not heard calling, and there is no indication that he is with the RP. He take the chick back to camp for treatment. ' 15:08 There is no activity at the nest site. 15:30 RP male lands on nest with fruit in bill tip. There are no vocalizations from the male and no sounds from the nest. He search the area for a dead male chick on the ground, but find nothing. 16:53 IP lands in nest tree, but not on the nest. 17:05 IP flies from area. 17:40 We leave nest area with no activity at the site. Jan. 8, 1984 Observations made from: 8:45-14:40 IP long-calls from the nest tree and remains around the nest site. No chicks are seen or heard. RP is not seen, but there are hornbills counter long-calling at 100-200 m from the nest which may be them. On two occasions, IP female lands on the nest and looks inside. Once she holds a fruit in her bill tip, and once she only pecks at the nest rim. IP male also lands on the nest twice. Once, he soft-honks and looks inside, and once he counter long-calls to the hornbill pair at 100 m. 162 Jan. 9, 1984 Observations made from: 16:09-17:00 IP remains at the nest-site long-calling. RP and chicks are not seen or heard. IP male lands on the nest once, with a fruit in his bill tip. There is no activity inside the nest. CHAPTER VII SUMMARY AND RECOMMENDATIONS Byeanjstes subcyljpgkieus habitat is restricted to forested areas, which in Uganda are experiencing rapid destruction and conversion. Basic information provided here on the ecology and behavior of black-and-white casqued hornbills birds permits recommendations to be made concerning the conservation of this species and of its rainforest habitat. In Kibale Forest, B. supeylingrieus movements, spatial dispersion, and habitat use vary seasonally. Hornbills are present in selectively logged areas, but in lower numbers than in primary forest in the core of the Reserve. Primary forest is the most suitable habitat for these birds. Hith an estimated mean density of 12.9 birds/km2 at Kanyawara and 27.8 birds/km2 at Ngogo, hornbills in Kibale are at greater densities than any other forest hornbill species on record. Hornbill movements are closely related to their diet. Particularly during dry seasons, B. subeyltuetieus travel long distances (probably > 10 km) in search of fruiting trees. Because they exploit fruit resources that are widely dispersed, B. sueeyliuerjcus require very large areas of forest for survival. These birds are primarily frugivorous and feed on a wide variety of 163 164 fruit species. Since seeds of most plant species consumed are passed out in feces or are regurgitated intact, these birds mediate seed-dispersal of tropical rainforest trees. The conservation and regeneration of this threatened ecosystem are certainly influenced and possibly maintained by these birds. B. subcylindricus have a complex reproductive ecology. Data indicate that rainfall, intra-specific competition for nest-sites, predation, and food supply influence nesting success. Hornbill reproduction is dependent on the availability of large, over-mature trees for nesting. Primary forest is apparently the preferred habitat for breeding, since nest and display-cavity density is highest there. The number of young fledged per unit area is also highest in unlogged forest. Logging alters the forest habitat for hornbills by reducing the number of large, over-mature trees available for nest sites. Therefore, it might be expected that, for each large nest tree removed, one less breeding pair is able to produce young (Fig. 7.1). Observations during this study indicated, however, that in forest tracts subjected to intensive logging of nest trees, far more than one breeding pair would be affected for each nest tree removed. B. subcylindricus are long-lived (probably > 20 yrs.) and compete for nest-sites with fervor. As nest trees are removed, the severe competition for remaining sites would likely decrease nesting success exponentially (Fig. 7.1, line b), rather than linearly (Fig. 7.1, line a). Today, it is important to note age distributions of B. subcylindricus when determining the population status of this 165 HORNBILL NEST SUCCESS (N° CHICKS FLEDGED/UNIT AREA) a N° NEST TREES REMOVED/UNIT AREA Fig. 7.l.--Hypothetical relationship between Byeenistes supeyljneyi; eus nest success and number of trees removed per unit area. Line "a" represents the linear response that might be expected if, for each nest tree removed, one hornbill pair was prevented from breeding. Line "b" represents the exponential reduction in nest success that is predicted based on expected increase in competition for remaining nest-sites. 166 species in degraded forested areas. Although birds may seem abundant, the population may be aged due to limited nest sites and low recruitment of young. Numbers of individuals would make a precipitous decline as aging birds die. More data analyses are needed for the effects of logging on the reproductive ecology of B. suttpylingLieus to be better understood. Based on the findings of this study, general recommendations can be made to reduce the negative effects of logging on this species. Recommendations made here for the conservation of B. subeylindricus in Africa agree with those of Johns (1987) and Kemp and Kemp (1974) for the conservation of hornbills in SE Asian rainforests. 1. Degradation of primary forest correlates with reduction in hornbill numbers. Large tracts of primary forest should, ideally, be left intact and undisturbed. 2. Should timber trees be removed, forestry practices should be as follows: a. Leave as many large trees standing as possible. b. Prevent incidental damage to the forest during timber extraction. c. Leave snags, hollow trees, and large or over-mature trees likely to produce nest cavities for hornbills. Often these trees are of no commercial value anyway. d. Leave trees known to contain hornbill nest or display cavities. These sites are used by hornbills year after year. 167 e. Leave large fruit trees (especially figs, Ejeus spp.) standing, which are known to produce mass crops for hornbills. Fig trees are an extremely important food source for hornbills and many other animals, and they are of little or no commercial value. Fig trees fruit aseasonally and may be a key factor in hornbill survival. f. The collection or hunting of hornbills must continue to be prohibited. 168 REFERENCES Johns, A. D. 1987. The use of primary and selectively logged rain- forest by Malaysian hornbills (Bucerotidae) and implications for their conservation. Biple_§puseyy. 40:179-190. Kemp, A. C. and M. I. Kemp. 1975. Report on a study of hornbills in Sarawak, with comments on their conservation. Kuala Lumpur, Horld Hildlife Fund Malaysia. (unpublished report.) REFERENCES REFERENCES Andersson, M. 1978. Optimal foraging area: Size and allocation of search effort. [hepr, Popul, ijl 13:397-409. Austin, G. T. 1974. Nesting success of the cactus wren in relation to nest orientation. Bender 76:216-217. Balgooyen, T. G. 1973. Behavior and ecology of the American kestrel (Falco sparverius L.) in the Sierra Nevada of California. Univ. Calif. Publ. 1991. 103:1-83. Bolin, 8. 1977. Changes of land biota and their importance for the carbon cycle. Bcienee 196:613-615. Bourne, D. and D. Chessell. 1982. Breeding the black-and-white casqued hornbill (Byeenistes subcylindricus subguedratus) at the Metro Toronto Zoo, Canada. Avigult. Meg. 88:15-23. Brown, L. H. and P. L. Britton. 1980. The Breeding Seasons of East Africen Birds. The East African Natural History Society, Nairobi. Burger, J. 1979. Competition and predation: Herring gulls versus laughing gulls. gongpr 81:267-277. Burger, J. and J. Shishler. 1978. Nest site selection and competi- tive interactions of herring and laughing gulls in New Jersey. Auk 95:252-266. Butynski, T. M. 1988. Guenon birth seasons and correlates with rainfall and food. In A. Gautier-Hion, F. Bourliere, J. P. Gautier, J. Kingdon, and R. M. Martin, eds. 'm R ia ' n: v 1 ti n r '01 n f ° an n . Cambridge Univ. Press, Cambridge. Butynski, T. M. in prep. Comparative ecology of blue monkeys (Eercopjthecus mitts) in high and low density subpopulations. Caraco, T., S. Martindale and H. R. Pulliam. 1980. Avian flocking in the presence of a predator. Aetuye 285:400-401. 169 170 Collar, N. J. and S. N. Stuart. 1985. threatened Birds of Africe end keleted isle ds. IUCN. Cambridge, UK. Collias, N. E. and E. c. Collias. 1984. "1 ' n ' d Behevipr. Princeton Univ. Press, New Jersey. Collias, N. E. and L. R. Jahn. 1959. Social behavior and breeding success in Canada geese (Breute eenadensis) confined under semi-natural conditions. Auk 76:478-509. Conner, R. N. 1975. Orientation of entrances to woodpecker nest cavities. Auk 92:371-374. Dennis, J. V. 1971. Species using red-cockaded woodpecker holes in northeastern South Carolina. Bird Bending 42:79-163. Eggeling, H. J. and I. R. Dale. 1952. The indigenous Trees of the Bganda Protectorate. 2nd edition. Government Printer, Entebbe. Emlen, J. T. 1984. An observer-specific, full-season, strip-map method for censusing songbird communities. [be Auk 101:730- 740. Ghiglieri, M. P. 1984. e i an e f Ki e st. Columbia Univ. Press. N.Y. Hamilton, A. C. 1981. A Field Guide to Ugende Eorest Trees. Makerere University Printery, Kampala. Hamilton, A. C. 1984. Befprestation in Ugande. Oxford Univ. Press. Nairobi, Kenya. Harvey, P. M. 1973. Breeding the casqued hornbill at "Birdworld." Avicult. Meg. 79:23-25. Hubbard, C. E., E. Milne-Redhead, R. M. Polhill and H. 8. Turrill, eds. 1952. Flora pf Trppical East Afriea. Crown Agents, London. IUCN. 1980. Save the rainforests. IBQN Bulletin: 11(5). Gland, Switzerland. Iwinski, D. and B. Iwinski. 1984. Too many toucans. The A.F.A. Hetchpird X6:18-20. Jackson, J. A. 1974. Gray rat snakes versus red-cockaded woodpeck- ers: Predator-prey adaptations. Auk 91:342-347. Janzen, D. H. 1979. How to be a fig. Ann. Bey. Eepl, Byst. 10:13-52. 171 Johns, A. O. 1983. Ecological effects of selective logging in a Hest Malaysian rain forest. Ph.D. dissertation, Cambridge Univ. Cambridge, U.K. Johns, A. D. 1985. Selective logging and wildlife conservation in tropical rainforest: Problems and recommendations. Biol. Eeeserv. 31:355-375. Johns, A. D. 1987. The use of primary and selectively logged rain- forest by Malaysian hornbills (Bucerotidae) and implications for their conservation. B1e1t_§eesery. 40:179-190. Kacelnik, A, A. 1. Houston and J. R. Krebs. 1981. Optimal foraging and territorial defense in the great tit (Eerus major). Behav. Eeo I. §Q§IQDIQI- 8: 35-40. Kalina, J. in press. Nest intruders, nest defense and foraging behavior in the black-and-white casqued hornbill (Byeenistes 50991113111015)- MS- Kalina, J. in prep. Effects of animal dispersers on germination success of five tree species in Kibale Forest, Uganda. Kalina, J. and T. M. Butynski (a) in prep. Comparative feeding ecology of sympatric hornbills and primates in Kibale Forest, Uganda. Kalina, J. and T. M. Butynski (b) in prep. Seed dispersal ecology of iriehilia spleudide in Kibale Forest, Uganda. Keith, A. R. 1970. Crowned hawk-eagle raids hornbill nest. J. of E. Afr. Net. Hist. Soc. end Nat. Mus. 1:64. Kemp, A. C. 1970. Some observations on the sealed-in nesting method of hornbills (Family Bucerotidae). Qstrieh (suppl.) 8:149-155. Kemp, A. C. 1973. Factors affecting the onset of breeding in African hornbills. Bree, 16th Int, Ore, Eeng.: 248-257. Kemp, A. C. 1976a. A study of the ecology. behaviour and systemat- ics of [oekus hornbills (Aves: Bucerotidae). lreusvee! Bus, Memoir Ne. 20. Pretoria, South Africa. Kemp, A. C. 1976b. Factors affecting the onset of breeding in African hornbills. r h n rn -. PP 248- 257. Kemp, A. C. 1979. A review of the hornbills: Biology and radia- tion. The Living Bird 17:105-136. 172 Kemp, A. C. 1985. IBBP Bernpill Spee. Bp, Bemm. 5. Kemp, A. C. 1988. Bucerotidae. In E. Urban, S. Keith, and H. Fry, eds. Ih§_fii£Q§_Qf;AI£i£§, vol. . Academic Press, London. Kemp, A. C. and T. M. Crowe. 1985. The systematics and zoogeog- raphy of Afrotropical hornbills (Aves: Bucerotidae). In K. L. Schuchman, ed., Bree, Intern. Symp. Africen Verteer. Bonn. Kemp, A. C. and M. I. Kemp. 1975. Report on a study of hornbills in Sarawak, with comments on their conservation. Kuala Lumpur, Horld Hildlife Fund Malaysia. (unpublished report.) Kemp, A. C. and M. I. Kemp. 1978. Bueeryus and Beg_tterius: Two modes of terrestrial predation. Bree, Symp. Afr, Predatory Birds: 13- 16. Kemp, A. C. and M. I. Kemp. 1980. The biology of the southern ground hornbill, Bueervus ieadbeateri (Vigors) (Aves: Bucero- tidae). Ann. Transvael Bus .32: 65-100 Kendeigh, S. D. 1934. The role of environment in the life of birds. Eeelegieel Meneg. 4:299-417. Kendeigh, S. C. 1961. Energy of birds conserved by roosting in cavities. Hilson Bull. 73:140-147. Kilham, L. 1956. Breeding and other habits of casqued hornbills (Byeanistes subcylindrieus). Smithsen. Bise, Cellns. 131:1-45. Kilham, L. 1968. Reproductive behavior of hairy woodpeckers, II. Nesting and habitat. Hilson Bull. 80:286-305. Kilham, L. 1969. Reproductive behavior of hairy woodpeckers, III. Agonistic behavior in relation to courtship and territory. NilfiQfl_Bflll- 81: 169- 183. Kilham, L. 1972. Habits of the crimson-crested woodpecker in Panama. kilson Bull. 84:28-47. Kingston, 8. 1967. orki P n f r b nd twara n ral For v . Uganda Forests Department, Entebbe, Uganda. Lack, D. 1954. Ihe Naturel Regulatien ef Animal Numpers. Clarendon Press, Oxford. Lawrence, L. D. K. 1966. The comparative life-history study of four species of woodpeckers. anttne_fleneg; 5:1-156. 173 Leighton, M. 1982. Fruit resources and patterns of feeding, spacing and grouping among sympatric Bornean hornbills. Ph.D. dissertation, Univ. Calif., Davis. Leighton, M. 1986. Hornbill social dispersion: Variations on a monogamous theme. pp. 108-130. In D. I. Rubenstein and R. H. .Nrangham, eds. Eeelegieal Aspeets ef Seeiel Eyolutien. Princeton Univ. Press, New Jersey. Lendrem, D. 1983. Predation risk and vigilance in the blue tit. Behev. Ecel. Seeiebiel. 14:9-13. Ligon, J. D. 1981. Demographic patterns and communal breeding in the green woodhoopoe (Bhoenieulus purpureus). pp. 231-243. In R. 0. Alexander and D. H. Tinkle, eds., Neturel Seleetien end Social Behavier: Reeent Reseereh end New Theory. Chiron Press, N.Y. Ligon, J. D. and S. H. Ligon. 1978a. Communal breeding in the green woodhoopoe as a case for reciprocity. Nature 176:496- 498. - Ligon, J. D. and S. H. Ligon. 1978b. The communal social system of the green woodhoopoe in Kenya. Living Bird 17:159-197. Ligon, J. D. and S. H. Ligon. 1982. The cooperative breeding behavior of the green woodhoopoe. Seient. Am. 247:126-134. Ligon, J. D. and S. H. Ligon. 1983. Reciprocity in the green woodhoopoe (Phoeniculus purpureus). Anim. Behav. 31:480-489. Loftin, R. H. and D. Roberson. 1983. Infanticide by a purple martin. Hilson Bull. 95:146-148. Lwanga, J. S. 1987. Group fission in blue monkeys (Cercopithecus mitis stuhlmanni): Effects on the socioecology in Kibale Forest, Uganda. M.Sc. thesis. Makerere Univ., Kampala, Uganda. Mackworth-Praed, C. H. and C. H. 8. Grant. 1957. Birds ef Eestern end North Eestern Afriee. Longman, Green and Co., London. Marsh, C. H. and H. L. Hilson. 1981. A survey of primates in peninsular Malaysian forest. Final report for the Malaysian primate research programme. Univ. Kebangsaan, Malaysia and Cambridge Univ., Cambridge, U.K. Martindale, S. 1982. Nest defense and central place foraging: A model and experiment. Behev. Eeei. Seeiebiol. 10:85-90. 174 McEllin, S. M. 1979. Nest sites and population demographics of white breasted and pigmy nuthatches in Colorado. gender 81:348-352. Mock, D. H. 1984. Infanticide, siblicide and avian nestling mortality. pp. 4-30. In G. Hausfater and S. B. Hrdy, eds. f t d : m ar v n i n r P e ti Aldine, N.Y. Moreau, R. E. and H. M. Moreau. 1941. Breeding biology of the silvery-cheeked hornbill. Auk 58:13-27. Myers, N. 1984. lee Primery Seuree, lrepieel Eerests end Bur future. H. H. Norton, N.Y. Newton, 1. 1979. Po 1 tor . Buteo Books, Vermillion, S. Dakota. Nice, M. 1957. Nesting success of altricial birds. Auk 74:305- 321. Nummelin, M. 1986. The seasonal fluctuations of forest floor insect densities on the areas of different forestry practices in Kibale Forest, Hestern Uganda. nf. n T o ical Entomology, Nairobi. Oates, J. F. 1977. The guereza and man: How man has affected the distribution and abundance of Beleeus guereze and other black colobus monkeys. pp. 419-467 in Prince Rainer and G. H. Bourne (eds.), Brimate Eonservetien. Academic Press. London and N.Y. Oniki, Y. 1979. Is nesting success of birds low in the tropics? Bjotrepice 11:60-69. Orians, G. and N. Pearson. 1979. On the theory of central place foraging. In D. J. Horn, G. R. Staris, and R. 0. Mitchell, eds., An 1 i f i a1 S . Ohio State University Press, Columbus. Penny, C. G. 1975. Breeding the Abyssinian ground hornbill at San Diego Hild Animal Park. Int’l Zee trek 15:111-115. Poonswad, P. A. Tsuji and C. Ngarmpongsai. 1983. A study of the breeding biology of hornbills (Bucerotidae) in Thailand. Bro- di . Ie . o_ F : .. i m . : -d n- rc '1 1!.V.t , pp. 239- 265. Porritt, R. and M. Riley. 1976. Breeding the black-and-white casqued hornbill (Byeenistes subeylindrieus) at Birdworld, Farnham. lntt_1ee_1ek. 16:104-105. 175 Ricklefs, R. E. and F. R. Hainsworth. 1968. Temperature dependent behavior of the cactus wren. Eeelegy 49:227-233. Rowley, I. 1970. The use of mud in nest building--A review of the incidence and taxonomic importance. Qstrieh (supp|,) 8:138- 148. ' Schoener, H. 1979. Generality of the size-distance relation in models of optimal foraging. Ame_Net. 114:902-914. Shelly, L. O. 1934. Tree swallow tragedies. Bird Banding 5:134. Short, L. L. 1979. Burdens of the Picid hole-excavating habit. Hilson Bull. 91:16-28. Short, L. L. 1982. Hoodpeekers of the Horld. Delaware Museum, Delaware. Sibley, C. G. and J. E. Ahlquist. 1972. A comparative study of the egg white proteins of non-passerine birds. Bull. Beabody Mus. Nat. Hist. 39. Sielmann, H. 1959. Y r ith th 0 d r , translation by S. Lightman. Barrie and Rockcliff, London. Sih, A. 1980. Optimal behavior: Can foragers balance two con- flicting demands? Seienee 210:1041-1043. Skorupa, J. P. 1985. Responses of rainforest primates to selective logging in Kibale Forest, Uganda: A summary report. Skorupa, J. P. 1988. The effects of selective timber harvesting on rainforest primates in Kibale Forest, Uganda. Ph.D. disserta- tion, Univ. Calif., Davis. Skutch, A. F. 1966. A breeding bird census and nesting success in Central America. this 108:1-16. Skutch, A. F. 1976. i n T i n . Univ. of Texas Press, Austin. Slobodkin, L. B. 1961. Preliminary ideas for a predictive theory of zoology. Am. Net. 45:147-153. Sommer, A. 1976. Attempt at an assessment of the world’s tropical moist forests. Unesylye 28:5-24. Struhsaker, T. T. 1972. Rainforest conservation in Africa. Primetes 13:103-109. 1 176 Struhsaker, T. T. 1975. [he Bed Colopus Nenkey. Univ. of Chicago Press, Chicago, Illinois. Struhsaker, T. T. 1978. Food habits of five monkey species in Kibale Forest, Uganda. In Recent Adyenees in Primateiogy, eds. D. C. Chivers and J. Herbert. 1:225-248. Academic Press, London. Struhsaker, T. T. 1981. Forest and primate conservation in East Africa. Afr, J, Eeel. 19:99-114. Struhsaker, T. T. 1987. Forestry issues and conservation in Uganda. Biele_tensery. 39:209-235. Svoboda, F. J. 1988. Toucan tending. [he A.F.A. Hetehbird XV3:36- 38. Trivelpiece, H. and N. J. Volkman. 1979. Nest-site competition between Adelie and Chinstrap penguins: An ecological interpre- tation. Auk 96:675-681. von Haartman, L. 1971. Population dynamics. pp. 392-461., In D. S. Farner and J. R. King, eds. Avian Bielegy Vol. 1. Academic Press, N.Y. Hannenburgh, A. 1980. The Bushmen. Country Life Books, N.Y. Natchtel, P. S. 1982. Hornbill, a most peculiar bird. Int’l Hildlife 12:36-41. Watt, 8. K. and A. L. Merrill. 1977. tamposition of Foods . . . Raw, Proeessed, Prepered. Agriculture Handbook #8, U.S. Dept. of Agriculture, Hashington, D.C. Helty, J. C. 1982. if f Bi . Saunders College Publish- ing, Philadelphia. White, F. N., J. Kinney, H. R. Siegfried, and A. C. Kemp. Thermal and gaseous conditions of hornbill nests. R ort N ’ Beeee_See. (unpubl. report). Hiley, J. H. 1980. The Puerto Rican parrot (Ameeene yittete): Its decline and the program for its conservation. p. 133. In R. F. Pasquier, ed. Eonservatien of New Herld Berrets. ICBP Tech. Publ. No. l. Hilliams, J. G. 1967. A field Buide te the Netional Perks ef E. Africa. Collins, London. 177 Wilson, H. L. and A. 0. Johns. 1982. Diversity and abundance of selected animal species in undisturbed forest, selectively logged forest and plantations in East Kalimantan, Indonesia. Biel, Qensery. 24:205-218. Hing, L. D. and I. O. Buss. 1970. Elephants and forests. Nildlife Neneerepts: No. 19. The Hildlife Society, Hashington, D.C. Zeleny, L. 1977. Nesting box programs for bluebirds and other pas- serines. pp. 55-60. In S. A. Temple, ed. Endengered Birds. Univ. of Hisconsin Press, Madison.