Comparative Anatomy of the Gastrointestinal Tract in Eutheria II: Taxonomy, Biogeography and Food. Laurasiatheria 9783110562217, 9783110560473

Citation preview

Handbook of Zoology Mammalia Comparative Anatomy of the Gastrointestinal Tract in Eutheria Taxonomy, Biogeography and Food Volume 2: Laurasiatheria, General Discussion

Handbook of Zoology Founded by Willy Kükenthal Editor-in-chief Andreas Schmidt-Rhaesa

Mammalia Edited by Frank E. Zachos

DE GRUYTER

Comparative Anatomy of the Gastrointestinal Tract in Eutheria

Taxonomy, Biogeography and Food Volume 2: Laurasiatheria, General Discussion Peter Langer

DE GRUYTER

Author Peter Langer Institut für Anatomie & Zellbiologie Justus-Liebig-Universität Aulweg 123 D-35392 Giessen [email protected] Scientific Editor Frank E. Zachos Naturhistorisches Museum Wien Säugetiersammlung Burgring 7 A-1010 Wien, Österreich

ISBN 978-3-11-056047-3 e-ISBN (PDF) 978-3-11-056221-7 e-ISBN (EPUB) 978-3-11-056067-1 ISSN 2193-2824 Library of Congress Cataloging-in-Publication Data A CIP catalogue record for this book is available from the Library of Congress. Bibliografic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. © 2017 Walter de Gruyter GmbH & Co. KG, Berlin/Boston Typesetting: Compuscript Ltd. Shannon, Ireland Printing and Binding: Hubert & Co. GmbH & Co. KG, Göttingen ∞ Printed on acid-free paper Printed in Germany www.degruyter.com

To the memory of my parents

Motto: “Der derzeitige Wissenschafts- und Universitätsbetrieb leidet nicht nur unter einer bisweilen schon ins Absurde gehenden Spezialisierung, sondern auch unter dem Verlust der historischen Kontinuität” (F. M. Wuketits, 2015).

Comparative Anatomy of the Gastrointestinal Tract in Eutheria Taxonomy, Biogeography and Food Volume 1: Introduction, Afrotheria, Xenarthra and Eucharchontoglires (ISBN 978-3-11-052615-8) Volume 2: Laurasiatheria, General Discussion (ISBN 978-3-11-056047-3)

Contents of volume II V Laurasiatheria xix 21 to 23 General remarks on Erinaceomorpha, Soricomorpha and Pholidota 309 309 21 Erinaceomorpha 21.1  Introductory remarks and notes on food of Erinaceomorpha 309 21.2  Anatomy of the stomach of the Erinaceomorpha, including arterial supply 310 21.3  Small intestine and colon in Erinaceomorpha 311 21.4  Blood vessels supplying or draining the colon of Erinaceomorpha 311 22 Soricomorpha 312 22.1  General remarks 312 22.2  Short remarks on the digestive physiology of Soricomorpha 312 22.2.1 Family Soricidae, Subfamily Crocidurinae and Myosoricinae 313 22.2.1.1 Food of Crocidurinae and Myosoricinae 313 22.2.1.2 Gastric anatomy and remarks on digestion 313 22.2.2 Family Soricidae, Subfamily Soricinae 315 22.2.2.1 General remarks 315 22.2.2.2 Food of Soricinae 315 22.2.2.3 Gastric anatomy and remarks on digestion 315 22.2.2.4 Rectum-licking 316 22.2.3 Family Talpidae, short overview 316 22.2.4 Concluding remarks on gastric elongation in the Soricomorpha 317 22.2.5 Blood vessels of the digestive tract in Soricomorpha 317 22.2.6 Colon of the Soricomorpha 318 22.2.6.1 Colonic villi in Soricomorpha 318 318 23 Pholidota 23.1  General remarks 318 23.2  Remarks on systematics, biogeography and food in Pholidota 319 23.3  Gastric anatomy of Pholidota 319 23.4  Form of the small and large intestines in Pholidota 321 23.5  Arteries of the gut in Pholidota 321 23.6  Concluding remarks on the colon of Erinaceomorpha, Soricomorpha and Pholidota 321

322 24 Chiroptera 24.1  Introductory remarks 322 24.2  Systematics, phylogeny, zoogeography of Chiroptera 323 24.3  Food of Chiroptera in general 325 24.4  General remarks on the gastric anatomy of Chiroptera 327 24.5  Chiroptera, Pteropodidae 328 24.5.1 General remarks on Pteropodidae 328 24.5.2 Gastric anatomy of Pteropodidae 328 24.5.2.1  Pteropus sp. 329 24.5.2.2  Harpyionycteris sp. 329 24.5.2.3 Scotonycteris sp. 331 24.5.2.4 Eonycteris sp. 331 24.5.2.5 Rousettus sp. 331 24.5.2.6 Megaloglossus sp. 331 24.5.2.7 Epomophorus sp. 332 24.5.2.8 Concluding remarks on pteropodid stomach anatomy 332 24.5.2.9 Gastric digestion in Pteropodidae 332 24.5.3 Small intestine of the Pteropodidae, form and function 332 24.5.4 Large intestine in the Pteropodidae 333 24.6  Chiroptera, Microchiroptera 333 24.6.1 Gastric anatomy 333 24.6.1.1 Gastric anatomy of the Megadermatidae 333 24.6.1.2 Gastric anatomy of the Rhinopomatidae 334 24.6.1.3 Gastric anatomy of the Hipposideridae 334 24.6.1.4 Gastric anatomy of the Rhinolophidae 334 24.6.1.5 Concluding remarks on the gastric form of Megadermatidae, Rhinopomatidae, Hipposideridae and Rhinolophidae 334 24.6.2 Gastric anatomy of the Noctilionidae and Mormoopidae 334 24.6.3 Gastric anatomy of the Thyropteridae 335 24.7  Chiroptera, Phyllostomidae 336 24.7.1 General remarks 336 24.7.2 Feeding types within the Phyllostomidae 336 24.7.3 Gastric anatomy of the Phyllostominae 337 24.7.4 Gastric anatomy of the Carollinae 338

viii 

 Contents of volume II

24.7.5 Gastric anatomy of the Stenodermatinae 338 24.7.6 Gastric anatomy of the Glossophaginae 339 24.7.7 Gastric anatomy of the Brachyphyllinae and Phyllonycterinae 341 24.7.8 Gastric anatomy of the Desmodontinae 341 24.8  Chiroptera, Molossidae 342 24.8.1 Introductory remarks 342 24.8.2 Gastric anatomy of the Molossidae 342 24.9  Chiroptera, Emballonuridae 343 24.9.1 Introductory remarks 343 24.9.2 Gastric anatomy of the Emballonuridae 343 24.10  Chiroptera, Natalidae 344 24.11  Chiroptera, Vespertilionidae 345 24.11.1 Introductory remarks 345 24.11.2 Gastric anatomy in the subfamily Vespertilioninae 345 24.11.3 Gastric anatomy in the subfamily Antrozoinae 346 24.11.4 Gastric anatomy in the subfamily Myotinae 347 24.11.5 Gastric anatomy in the subfamily Miniopterinae 348 24.11.6 Gastric anatomy in the subfamily Kerivoulinae 348 24.12  Small intestine of Chiroptera 349 24.12.1 Villi in the chiropteran small intestine 349 24.13  The caecum in Chiroptera 349 24.13.1 Distribution and anatomy of the chiropteran caecum 349 24.13.2 Digestive physiology of chiropterans under special consideration of the caecum 351 24.14  Colon of Chiroptera 352 24.15  Arterial supply of the gastrointestinal tract of Chiroptera 352 24.16  Concluding remarks on the digestive tract of Chiroptera 353 355 25 and 26 Carnivora General remarks on Carnivora 355 Types of food in Carnivora 356 25 Feliformia 356 25.1  General remarks on Feliformia 356 25.2  Gastric anatomy of Feliformia 356 25.3  Small intestine of the Carnivora 359 25.4  General remarks on the caecum of the Carnivora 359 25.5  Anatomy of the carnivoran caecum 359

361 25.5.1 Caecum of the Carnivora, Feliformia 361 26 Caniformia 26.1  Fissipeds, land-living Caniformia 361 26.1.1 General remarks on fissipeds 361 26.1.2 The food of fissipeds 362 26.1.3 Gastric anatomy of fissipeds 362 26.2  Pinnipeds, aquatic Caniformia 364 26.2.1 General remarks on pinnipeds 364 26.2.2 Gastric anatomy of pinnipeds 365 26.2.3 The small intestine of Pinnipedia 366 26.2.4 Caecum of the Carnivora, Caniformia 366 26.2.4.1 Caecum of the Canidae 366 26.2.4.2 Regio ileocolica of the Mustelidae 366 26.2.4.3 Regio ileocolica of the Ursidae (including Ailuropoda melanoleuca and Ailurus fulgens) 367 26.3  Caecum of the Pinnipedia 369 26.4  Colon of the Carnivora 369 26.4.1 Topography and morphology of the colon in Carnivora 369 26.4.2 Notes on the wall of the colon 372 26.4.3 Functional characteristics of the carnivoran colon 373 26.4.4 Blood vessels supplying or draining the carnivoran colon 375 26.4.5 Concluding remarks on the colon of Carnivora 376 378 27 Perissodactyla 27.1  General remarks on the Perissodactyla 378 27.2  Food of Perissodactyla in general 378 27.3  Form and function of the post-oesophageal digestive tract in the Perissodactyla 380 27.4  Equidae 380 27.4.1 Considerations on the Phylogeny of horses 380 27.4.2 Food of the genus Equus 381 27.4.3 Wild asses 382 27.4.4 Gastric anatomy of the Equidae 382 27.4.4.1 Internal mucosal lining of the equine stomach 384 27.4.4.2 Gastric digestion in the horse 385 27.5  Rhinocerotidae 385 27.5.1 Food of the Rhinocerotidae 386 27.5.2 Anatomy of the stomach of Rhinocerotidae 387 27.6  Tapiridae 388 27.6.1 Food of the Tapiridae 388 27.6.1.1 Food of Tapirus bairdii 388 27.6.1.2 Food of Tapirus pinchaque 389



27.6.1.3 Food of Tapirus terrestris 389 27.6.1.4 Food of Tapirus indicus 390 27.6.1.5 Summing up the information on tapir food 390 27.6.2 Anatomy of the stomach of Tapiridae 390 27.7  The small intestine of Perissodactyla 391 27.8  Caecum of the Equidae – recent horses, donkeys and asses 391 27.8.1 Functional remarks on the equine caecum 393 27.8.2 General remarks on the caecum of the Tapiridae 396 27.8.2.1 Anatomy of the caecum of the Tapiridae 396 27.8.3 General remarks on the caecum of Rhinocerotidae 397 27.8.4 Anatomy of the caecum of the Rhinocerotidae 397 27.9  Topography and morphology of the colon in Perissodactyla 399 27.9.1 Notes on the wall of the colon in Perissodactyla 402 27.9.2 Blood vessels supplying or draining the colon 403 27.9.3 Functional characteristics of the colon in Perissodactyla 404 27.10  Remarks on the digestive tract of Perissodactyla, with anatomical and functional considerations 405 27.11  Concluding remarks on Perissodactyla 409 28 to 30 Cetartiodactyla 410 410 28 Cetartiodactyla, Artiodactyls 28.1  General remarks 410 28.1.1 Systematics and types of food 412 28.2  Camelidae 415 28.2.1 General remarks 415 28.2.2 Food of the Camelidae 416 28.2.3 Anatomy of the stomach of Camelidae 416 28.2.3.1 Terminological questions 416 28.2.3.2 Descriptive anatomy of the camelid stomach 418 28.2.3.2.1 General gastric anatomy and topography 418 28.2.3.2.2 Specific remarks on the rumen of Camelidae 420 28.2.3.2.3 Specific remarks on the reticulum of Camelidae 422 28.2.3.2.4 Specific remarks on the gastric tube and hindstomach of Camelidae 422

Contents of volume II 

 ix

423 28.2.3.2.5 Gastric mesenteries of Camelidae 28.2.3.2.6 Architecture of the tunica muscularis in Camelidae 423 28.2.3.2.7 Histology of the tunica mucosa in Camelidae 425 28.2.3.2.8 Gastric blood vessels in Camelidae 426 28.2.3.2.9 Functional overview of the stomach in Camelidae 426 28.2.3.2.10 Concluding remarks on the camelid stomach 429 28.2.4 Small intestine of Camelidae 429 28.2.5 Colon of Camelidae 430 28.2.6 Functional anatomy of the caecum in Camelidae 430 28.3  Suidae 431 28.3.1 Introductory remarks 431 28.3.2 Food of Suidae 433 28.3.3 Gastric anatomy of the Suidae (Babirusa follows separately) 433 28.3.3.1 Mucosal lining of the suid stomach (Babirusa follows separately) 435 28.3.3.2 Gastric mesenteries (Babirusa follows separately) 436 28.3.3.3 Architecture of the tunica muscularis (Babirusa follows separately) 437 28.3.3.4 Blood vessels of the porcine stomach (Babirusa follows separately) 438 28.3.3.5 Ontogenetic development of the porcine stomach 439 28.3.3.6 Short functional remarks on the porcine stomach 440 28.3.4 Small intestine of Suidae 440 28.3.4.1 Glands and villi in the small intestine of Suidae 440 28.3.5 Topography and morphology of the colon in Suidae 440 28.3.5.1 Blood vessels of the colon in Suidae 441 28.3.6 The caecum of the Suidae, general remarks 442 28.3.6.1 Anatomy of the porcine caecum 442 28.3.6.2 Remarks on the functions of the porcine caecum 443 28.4 Babyrousa sp. 444 28.4.1 Introductory remarks 444 28.4.2 Food eaten by the babirusa 444 28.4.3 Anatomy and mucosal lining of the stomach  445 28.4.3.1 Gastric mesenteries of the babirusa 447 28.4.3.2 Blood vessels of the babirusa stomach 448

x 

 Contents of volume II

28.4.4 Small intestine of babirusa 448 28.4.5 Colon of babirusa 448 28.4.6 Remarks on the caecum of babirusa 449 28.5  Tayassuidae 449 28.5.1 Introductory remarks 449 28.5.2 Remarks on the biology of peccaries 450 28.5.3 Food of the peccaries 451 28.5.4 Anatomy of gastric compartments 451 28.5.4.1 Internal mucosal lining 455 28.5.4.2 Muscle architecture of the tunica muscularis 456 28.5.4.3 Arteries of the gastric region 457 28.5.4.4 Mesenteries of the peccary stomach 457 28.5.4.5 Functional remarks on the peccary stomach 459 28.5.5 Small intestine of Tayassuidae 460 28.5.6 Colon of the Tayassuidae 460 28.5.7 Caecum in Tayassuidae 460 28.5.8 Remarks on digestion the Tayassuidae 461 28.5.9 Comparative remarks on digestion in Suidae and Tayassuidae 462 28.6  Tragulidae 463 28.6.1 Introductory remarks 463 28.6.2 Fossil and surviving Tragulidae 464 28.6.3 The food of Tragulidae 464 28.6.4 Anatomy of gastric compartments 465 28.6.4.1 Rumen 465 28.6.4.2 Reticulum 466 28.6.4.3 Isthmus reticuloabomasicus 466 28.6.4.4 Abomasum 468 28.6.4.5 Mucosal lining of the tragulid stomach 468 28.6.4.6 Blood vessels of the tragulid stomach 469 28.6.4.7 Muscle architecture of the tunica muscularis 469 28.6.4.8 Mesenteries 469 28.6.4.9 Digestion and functional remarks on the stomach of the Tragulidae 470 28.6.5 Small intestine of the Tragulidae 471 28.6.6 Colon of the Tragulidae 472 28.6.7 Caecum of the Tragulidae 472 28.6.8 Remarks on digestion in Tragulidae 473 28.7  Pecora 473 28.7.1 Introductory remarks on Pecora 473 28.7.2 General considerations of gastric digestion in the Pecora 474 28.7.3 Remarks on food and general gastric function in Pecora 475

476 28.7.4 Gastric anatomy in the Pecora 28.7.4.1 The ruminoreticulum in Pecora 476 28.7.4.2 Anatomy of the rumen in Pecora 476 28.7.4.3 Internal macroscopic surface differentiations in the pecoran rumen 477 28.7.4.4 Anatomy of the reticulum in Pecora 479 28.7.4.5 Internal macroscopic surface differentiations in the pecoran reticulum 479 28.7.4.6 Functional remarks on the whole ruminoreticulum in Pecora 480 28.7.4.7 Anatomy of the omasum in the Pecora 480 28.7.4.8 Internal macroscopic surface differentiations in the pecoran omasum 480 28.7.4.9 Functional remarks on the omasum in Pecora 481 28.7.4.10 Anatomy of the abomasum in Pecora 482 28.7.4.11 Internal macroscopic surface differentiations in the pecoran abomasum 482 28.7.4.12 Functional remarks on the abomasum of Pecora 482 28.7.4.13 Relative volumes of gastric compartments in Pecora 482 28.7.4.14 Blood vessels of the stomach in Pecora 482 28.7.4.15 Gastric mesenteries in Pecora 483 28.7.4.16 Muscle architecture of the gastric tunica muscularis in Pecora 483 28.7.4.17 Functional remarks on the total stomach of Pecora 485 28.7.5 Small intestine of Pecora 486 28.7.5.1 Glands and villi in the small intestine of Pecora 487 28.7.6 Colon of Pecora 488 28.7.6.1 The tunica mucosa in the colon of Pecora 489 28.7.7 Caecum of Pecora 490 28.7.7.1 Anatomy and topography of the pecoran caecum 490 28.7.7.2 The caecum in Bovidae 491 28.7.7.3 The caecum in Cervidae 492 28.7.7.4 The caecum in Giraffidae 492 28.7.7.5 Remarks on the histology of the pecoran caecum 493 28.7.7.6 Embryology of the pecoran caecum 493 28.7.7.7 Microbiology of the pecoran caecum 494 28.7.7.8 Remarks on the physiology of the pecoran caecum 494



28.7.7.9 Remarks on the arterial supply, venous drainage and innervation of the pecoran caecum 495 28.8  Hippopotamidae 496 28.8.1 General remarks 496 28.8.2 Fossil Hippopotamidae 497 28.8.3 Food of the Hippopotamidae 497 28.8.4 Gastric anatomy in Hippopotamidae 498 28.8.4.1 General subdivision and position of the stomach in Hippopotamidae 498 28.8.4.2 Visceral blindsac in Hippopotamidae 499 28.8.4.3 Parietal blindsac in Hippopotamidae 500 28.8.4.4 Connecting compartment in Hippopotamidae 500 28.8.4.5 Glandular stomach in Hippopotamidae 500 28.8.4.6 Volumes of gastric compartments in Hippopotamidae 500 28.8.4.7 Macroscopic internal surface differentiations in hippopotamid forestomachs 502 28.8.4.8 The mucosa of the glandular stomach in Hippopotamidae 504 28.8.4.9 Mesenteries of the stomach in Hippopotamidae 504 28.8.4.10 Muscle architecture of the hippopotamid gastric tunica muscularis 504 28.8.4.11 Blood vessels of the gastric region in Hippopotamidae 506 28.8.4.12 Remarks on gastric digestion in Hippopotamidae 507 28.8.5 Small intestine of Hippopotamidae 509 28.8.6 Colon of Hippopotamidae 509 28.9  Concluding remarks on functional characteristics of the colon in different Artiodactyla 510 28.9.1 Microbial population 510 28.9.2 Alloenzymatic digestion (fermentation) 511 28.9.3 Importance of fermentation in the colon of Artiodactyla 511 28.9.4 The colon as fermentation volume 511 28.9.5 Colonic spirals 512 28.9.6 Taeniae, haustra and semilunar folds as differentiation of the colonic wall 512 28.9.7 Differentiation of the colonic tunica mucosa 512 28.9.8 Lymphoid tissue and innervation of the colon in Artiodactyla 513

Contents of volume II 

 xi

514 29 and 30 Cetartiodactyla, Cetacea 29 Mysticeti 514 29.1  Introductory remarks 514 29.2  Systematics and Phylogeny of the Whippomorpha (Hippopotamidae plus Cetacea) 514 29.3  Types of food in Mysticeti 518 29.4  Types of food in Odontoceti 519 29.5  Anatomy of the stomach of Cetacea 521 29.5.1 General remarks 521 29.5.2 Stomach of Mysticeti 522 29.5.2.1 Stomach of Balaenidae, right whales 522 29.5.2.2 Stomach of the Balaenopteridae, rorquals 522 29.5.2.3 Stomach of Neobalaenidae 526 29.5.2.4 Stomach of Eschrichtidae 526 527 30 Odontoceti 30.1  Stomach 527 30.1.1 Stomach of Odontoceti 527 30.1.2 Stomach of Delphinidae 527 30.1.3 Stomach of Monodontidae 530 30.1.4 Stomach of Phocoenidae 531 30.1.5 Stomach of Physeteridae 533 30.1.6 Stomach of Platanistidae 535 30.1.7 Stomach of Iniidae 537 30.1.8 Stomach of Inia geoffrensis 538 30.1.9 Stomach of Pontoporia blainvillei 538 30.1.10 Stomach of Lipotes vexillifer 538 30.1.11 Comparative remarks on the stomach of the Iniidae and Platanista 539 30.1.12 Stomach of Ziphiidae, general remarks 540 30.1.13 Stomach of Hyperoodon 540 30.1.14 Stomach of Mesoplodon 541 30.1.15 Concluding and comparative remarks on the cetacean stomach 542 30.2  Small intestine 545 30.2.1 Small intestine of Mysticeti 545 30.2.2 Small intestine of Odontoceti 546 30.3  Morphology and topography of the colon in  546 Cetacea, introductory remarks 30.3.1 Colon of Mysticeti 546 30.3.2 Colon of Odontoceti 547 30.3.3 Notes on the wall of the cetacean colon 551 30.3.4 Concluding remarks on the colon of Cetacea 553

xii 

 Contents of volume II

30.4  The cetacean Caecum 553 30.4.1 Caecum in Mysticeti 553 30.4.2 Caecum in Odontoceti 555 30.5  Functional considerations concerning the cetacean gut 555 VI 1 1.1 1.2

557 General discussion 559 The digestive tube of Eutheria Reactor models 560 Form, size and function in the digestive tract in Eutheria 560 1.3 General remarks on fermentation and alloenzymatic digestion 561 1.4 Functional differentiation within the gut 562 2 Some criteria that are used to define anatomical sections of the gut 562 2.1 Blood vessels 563 2.2 Intraperitoneal and retroperitoneal positions 563 2.3 Tunica mucosa 564 2.4 Tela submucosa 565 2.5 Tunica muscularis 565 2.5.1 Taeniae, haustra and semilunar folds 565 2.5.2 Architecture of the gastric tunica muscularis 566 2.6 Anatomy of the gastrointestinal tract, comparative aspects 566

2.7

Food quality and regional differentiations of compartments of the post-oesophageal digestive tract 569 Combinations of morphological 2.8 differentiations 573 577 3 Stomach Separate glands in the stomach of some 3.1 Eutheria 578 4 Small intestine 579 4.1 Duodenum 580 Jejuno-ileum with remarks on their 4.2 differentiation and on their mesentery 582 583 5 Colon Terminological aspects 5.1 584 5.2 Sections of the colon 584 5.3 Colonic spirals and loops 586 Digestive function of the colon – 5.4 comparative aspects 587 589 6 Caecum 6.1 Comparative remarks 589 6.2 The appendix vermiformis, a differentiation of the caecum in some eutherian mammals 591 6.3 Functional remarks on the Appendix vermiformis 593 595

Literature Index

683

Contents of volume I Preface

vii

1 I Introduction 1 An appreciation of exemplary investigators – past and present 3 2 Outline of eutherian systematics 3 2.1 Definition and delineation of gut sections 4 2.2 Functional differentiations of gastrointestinal tract sections 5 2.3 General remarks on sections of the gastrointestinal tract 7 2.4 Food index and volume 7 2.5 Food quality eaten by eutherian taxa

9

61 II Afrotheria Description of the post-oesophageal digestive tract in the eutheria 63 1 Afrosoricida, Tenrecomorpha 63 1.1  Geographical distribution and types of food eaten by species of the four tenrecomorph subfamilies 63 1.2  Remarks on gastric and small intestinal, as well as on colon anatomy in Tenrecomorpha 64 2 Afrosoricida, Chrysochloridea 65 2.1  Type of food eaten by species of the two chrysochlorid subfamilies 65 2.2  Remarks on gastric and small intestinal, as well as on colon anatomy in Chrysochliridae 65 3 Macroscelidea 66 3.1  Type of food eaten by Macroscelidea 66 3.2  Remarks on gastric and small intestinal, as well as colon and caecum anatomy in Macroscelidea 67 4 Tubulidentata 68 4.1  Type of food eaten by Tubulidentata 69 4.2  Remarks on gastric and small intestinal, as well as colon and caecum anatomy in Tubulidentata 69 5 Hyracoidea 70 5.1  Types of food eaten by Hyracoidea 70 5.2  Anatomy of the stomach of Hyracoidea 71 5.3  Functional considerations related to the stomach and other sections of the digestive tract of Hyracoidea 73

5.4  The small intestine, especially the duodenum of Hyracoidea 74 5.5  Colon and caecum of the Hyracoidea 75 5.6  Arterial supply of the caecum and of other parts of the digestive tract in Hyracoidea 77 5.7 Terminology of the digestive tract of Hyracoidea under special consideration of the caecum 77 5.8  Descriptive and functional anatomy of the caecum of Hyracoidea 78 6 Proboscidea 79 6.1  Introductory remarks 79 6.2  Food of elephants 80 6.3  Anatomy of the stomach of the Proboscidea 81 6.4  Functional considerations concerning the stomach of the Proboscidea 81 6.5  Anatomy of the small intestine of Proboscidea 82 6.6  Anatomy of the colon of Proboscidea 82 6.7  Anatomy of the caecum of Proboscidea 82 6.8  Functional considerations concerning the large intestine of Proboscidea 83 7 Sirenia 84 7.1  Introductory remarks 84 7.2  Food of Dugong dugon (living in salt water) 85 7.3  Food of Trichechus manatus (living in variable salinity) 86 7.4  Food of Trichechus inunguis (living in freshwater) 86 7.5  Food of Hydrodamalis gigas (marine and extinct) 86 7.6  Gastric anatomy in the Sirenia 86 7.6.1 Mucosal differentiations of the sirenian stomach 90 7.6.2 Mesenteria of the sirenian stomach 90 7.6.3 Arterial supply of stomach and ampulla duodeni 91 7.6.4 Functional remarks, emphasising the sirenian stomach and ampulla duodeni 92 7.7  Small intestine of the Sirenia 93 7.8  Colon anatomy in the Sirenia 93 7.9 Caecum anatomy in the Sirenia, as described for Trichechus manatus, by Snipes (1984b) 94

xiv 

 Contents of volume I

7.9.1 Vascularisation of the caecum in Trichechus manatus, described by Snipes (1984b) 96 7.9.2 Transmission electron microscopy in Trichechus manatus, described by Snipes (1984b) 97 7.10  Remarks on functional aspects of the total post-oesophageal digestive tract in Sirenia 97 7.11  Final remarks on the Paenungulata (Hyracoidea + Proboscidea + Sirenia) 98 99 III Xenarthra Introductory remarks, Xenarthra in general 101 8 Cingulata 101 8.1  Systematics, phylogeny and physiology 101 8.2  Type of food eaten by Cingulata, Dasypodidae 102 8.3  Gastric anatomy of Cingulata, Dasypodidae 104 8.4  Anatomy of the small intestine of Cingulata, Dasypodidae 105 8.5  Anatomy of the colon of Cingulata, Dasypodidae 105 8.6  Anatomy of the caecum of Cingulata, Dasypodidae 105 9 Pilosa, Folivora 106 9.1  Introductory remarks 106 9.2  Food of the Pilosa, Folivora 107 9.3  Gastric anatomy of Pilosa, Folivora 108 9.3.1  Mesenteries of Pilosa, Folivora 112 9.3.2  Architecture of the gastric tunica muscularis in Bradypus sp. 114 9.3.3  Mucosal lining of the stomach 115 9.3.4  Blood vessels of the gastric area in Bradypus tridactylus 115 9.3.5  Concluding functional remarks on the stomach of the Pilosa, Folivora 116 9.4  Anatomy of the small intestine of the Pilosa, Folivora 117 9.5  Anatomy of the colon of Pilosa, Folivora, including remarks on the absence or presence of a caecum 117 10 Pilosa, Vermilingua 118 10.1  Introductory remarks, also considering fossil forms 118 10.2  Recent species 119 10.3  Food of the Pilosa, Vermilingua 119 10.4  Gastric anatomy of Pilosa, Vermilingua 120 10.5  Anatomy of the small intestine of Pilosa, Vermilingua 121

10.6  Anatomy of the colon of Pilosa, Vermilingua 122 10.7  Anatomy of the caecum of Pilosa, Vermilingua 122 123 IV Euarchontoglires 125 Introductory remarks 11 Scandentia 125 11.1 Food of Scandentia 126 11.2 Gastric anatomy of Scandentia 127 11.3 Anatomy of the small intestine of Scandentia 127 11.4 Anatomy of the colon of Scandentia 126 11.5 Anatomy of the caecum of Scandentia 127 11.6 Short remarks on digesta transit in Scandentia 130 12 Dermoptera 130 12.1 Introductory remarks 130 12.2 Food of Dermoptera 131 12.3 Gastric anatomy of Dermoptera 131 12.3.1 Arterial supply of the stomach in Cynocephalus volans (Schultz, 1972) 131 12.4 Anatomy of the small intestine of Dermoptera 131 12.5 Anatomy of the colon of Dermoptera 132 12.6 Anatomy of the caecum of Dermoptera 133 12.7 Functional remarks on the gastrointestinal tract in Dermoptera 133 13 and 14 Primates 133 General overview 133 13 Strepsirrhini (“wet-nosed” primates) 136 13.1 Introductory remarks 136 13.2 The food of Strepsirrhini 137 13.3 The stomach of Strepsirrhini 139 13.3.1 Gastric anatomy of Strepsirrhini 139 13.3.2 Internal lining of the strepsirrhine stomach 142 13.3.3 Functional remarks concerning the strepsirrhine stomach 142 13.4 General remarks on small and large intestines of Primates 143 13.4.1 Anatomy of the small intestine of Strepsirrhini 146 13.4.2 Anatomy of the colon of Strepsirrhini 147 13.4.3 Concluding remarks to the colon of Strepsirrhini 149 13.5 Anatomy of the caecum of Strepsirrhini 150 14 Haplorrhini (“dry-nosed” primates) 151 14.1 Introductory remarks 151 14.2 General remarks on food 151 14.2.1 Food of the Tarsiiformes, Tarsiidae 152



14.2.2

Food of the Simiiformes, Platyrrhini 152 14.2.3 Food of the Simiiformes, Catarrhini: Cercopithecidae 153 14.2.4 Food of the Simiiformes, Catarrhini: Hylobatidae and Hominidae 153 14.3 Anatomy of the stomach of Haplorrhini 153 14.3.1 Tarsiiformes and Simiiformes: Platyrrhini 153 14.3.2 Simiiformes: Catarrhini 154 14.3.2.1 Cercopithecoidea: Cercopithecinae 154 14.3.2.2 Hominoidea, Hylobatidae and Hominidae. Discussion of the mesogastria 154 14.3.2.3 Musculature of the gastric wall, mainly of the human stomach 156 14.3.2.4 Remarks on the histology of the tunica mucosa in the unilocular, mainly human, stomach 159 14.3.2.5 Remarks on vascularisation and innervation 160 14.4 Small intestine of the Haplorrhini 161 14.5 Colon of the Haplorhini 161 14.6 Caecum of the Haplorrhini 165 14.7 Appendix vermiformis in Primates 167 14.8 Cercopithecoidea: Colobinae (mainly dealing with the stomach) 168 14.8.1 Introductory remarks 168 14.8.2 Food of the Colobinae 168 14.8.3 Short account of previous publications on the macroscopic anatomy of the stomach in Colobinae 170 14.8.3.1 General subdivision and position of the colobine stomach 171 14.8.3.2 Form of the colobine stomach 174 14.8.3.3 Macroscopic and microscopic internal surface differentiations of the colobine stomach 177 14.8.3.4 Volumes of the stomach regions in Colobinae 178 14.8.3.5 Gastric blood vessels in Colobinae 179 14.8.3.6 Gastric mesenteries in Colobinae 180 14.8.3.7 Muscle architecture of the tunica muscularis 180 14.8.3.8 Superficial layer 180 14.8.3.9 Deep layer 180 14.8.3.10 Functional remarks on the stomach of the Colobinae 181 15 to 20 Glires, short overview 182 15 to 19 Rodentia 183 General remarks 183

Contents of volume I 

 xv

185 15 Sciuromorpha 15.1 General remarks 185 15.2 Food of the Sciuromorpha 186 15.3 Gastric anatomy of the Sciuromorpha 188 15.4 Small intestine of Sciuromorpha 189 15.5 Colon of Sciuromorpha 190 15.6 Caecum of Sciuromorpha 191 16 Castorimorpha 194 16.1 Introductory remarks 194 16.2 Food of the Castorimorpha 194 16.3 Gastric anatomy of the genus Castor 195 16.4 A few remarks on the castorimorph families Geomyidae and Heteromyidae 196 16.5 Small intestine of Castorimorpha 197 16.6 Colon of Castorimorpha 198 16.7 Caecum of Castorimorpha 198 17 Myomorpha 199 17.1 General remarks 199 17.2 Food of the Myomorpha 199 17.3 Gastric anatomy in the Myomorpha 201 17.3.1 Terminological remarks 201 17.3.2 General remarks on gastric form, mucosal lining and taxonomic relationships in the Myomorpha 202 17.3.2.1 Remarks on the stomachs of Dipodidae 204 17.3.2.2 Remarks on the stomachs of Nesomyidae 205 17.3.2.3 Remarks on the stomachs of Cricetidae 206 17.3.2.3.1 Subfamily Arvicolinae 206 17.3.2.3.2 Subfamily Sigmodontinae 207 17.3.2.3.3 Subfamily Cricetinae 208 17.3.2.3.4 Subfamily Neotominae 208 17.3.2.3.5 Subfamily Lophiomyinae 209 17.3.2.3.6 Subfamily Tylomyinae 209 17.3.2.4 Remarks on the stomachs of Muridae 210 17.3.2.5 Remarks on the stomach of Spalacidae 213 17.4 Small intestine of Myomorpha 214 17.5 Colon of Myomorpha 215 17.6 Caecum of Myomorpha 220 17.6.1 Muridae 220 17.6.2 Cricetidae, Genus Microtus 222 Cricetidae, Genus Ondatra 17.6.3 223 17.6.4 Caecum of Myomorpha, Cricetidae, Genus Mesocricetus 223 17.6.5 Caecum of Myomorpha, Cricetidae, diverse genera 224

xvi 

 Contents of volume I

17.6.6 Caecum of Myomorpha, Spalacidae and Nesomyidae 225 17.6.7 Concluding remarks on caecal digestion 225 18 Anomaluromorpha 226 18.1 General remarks 226 18.2 Gastric anatomy of the Anomaluromorpha 226 18.3 Small intestine of Anomaluromorpha 227 18.4 Colon of Anomaluromorpha 227 18.5 Caecum of Anomaluromorpha 227 19 Hystricomorpha 228 19.1 General remarks 228 19.2 Form and function of the gastric region in Hystricomorpha 228 19.2.1 Infraorder: Ctenodactylomorphi 229 19.2.1.1 Family: Ctenodactylidae 229 19.2.2 Infraorder: Hystricognathi 229 African and Eurasian hystricognath families 229 19.2.2.1 Family: Bathyergidae 229 19.2.2.1.1 Food of some bathyergid species 229 19.2.2.1.2 Remarks on the gastric anatomy of some bathyergid species 229 19.2.2.2 Family: Hystricidae 230 19.2.2.2.1 Remarks on the gastric anatomy of some hystricid species, combined with short notes on food 230 19.2.2.2.2 Arterial supply of the stomach of Hystrix cristata 231 19.2.2.3 Family: Petromuridae 231 19.2.2.4 Family: Thryonomyidae 231 American hystricognath families 232 19.2.2.5 Family: Erethizontidae 232 19.2.2.6 Family: Chinchillidae 233 19.2.2.7 Family: Dinomyidae 233 19.2.2.8 Family: Caviidae 233 19.2.2.9 Families: Dasyproctidae and Cuniculidae 236 19.2.2.10 Family: Ctenomyidae 237 19.2.2.11 Family: Octodontidae 237 19.2.2.12 Family: Abrocomidae 239 19.2.2.13 Family: Echimyidae 239 19.2.2.14 Family: Myocastoridae 239 19.2.2.14.1 Food of the nutria (Myocastor coypus) 239 19.2.2.14.2 Anatomy of the nutria stomach 239 Caribbean hystricognath families 240 19.2.2.15 Family: Capromyidae 240 19.3 Hystricomorpha, small intestine 241 19.4 The colon of Hystricomorpha 241

247 Compilation of colonic differentiations Mesenteries of the rodent colon 248 Arterial supply of the rodent colon 250 Macroscopic configuration of the rodent colon 254 19.4.5 Differentiations of the colon wall in rodents 256 19.4.6 Macroscopically visible internal differentiations of the colon 260 19.4.7 Histological differentiations of the colonic internal lining 264 19.4.8 The colonic separation mechanism and colonic anatomy 266 19.5 Caecum of Hystricomorpha 268 19.5.1 The caecum of Bathyergidae 268 19.5.2 The caecum of Hystricidae and Erethizontidae 270 19.5.3 The caecum of Thryonomyidae 271 19.5.4 The caecum of Chinchillidae 271 19.5.5 The caecum of Caviidae 272 19.5.6 The caecum of Octodontidae 275 19.5.7 The caecum of Echimyidae 275 19.5.8 The caecum of Myocastoridae 275 20 Lagomorpha 276 20.1 Leporidae 278 20.1.1 General remarks 278 20.1.2 Food of the Leporidae 278 20.1.3 Coprophagy and caecotrophy, general remarks 280 20.1.3.1  Coprophagy and caecotrophy in Lagomorpha 280 20.1.3.2 Microbial population in the stomach of Oryctolagus cuniculus and its ontogenetic differentiation 281 20.1.4 Anatomy of the stomach of the genus Oryctolagus 281 20.1.4.1 Short remark on the arterial supply of the rabbit stomach 283 20.1.5 Anatomy of the stomach of the genus Lepus 283 20.2 Ochotonidae, general considerations 284 20.2.1 Food of the Ochotonidae 284 20.2.2 Gastric anatomy of the Ochotonidae 285 20.3 The small intestine in Lagomorpha 285 20.4 Colon configuration and arterial supply in Lagomorpha 285 20.4.1 Taeniae in the colon wall 292 20.4.2 Macroscopically visible internal differentiations of the colon in Lagomorpha 292 19.4.1 19.4.2 19.4.3 19.4.4

Contents of volume I 



20.4.3

Histology of the tunica mucosa in the colon of Lagomorpha 293 20.4.4 Functional differentiations in the colon of the rabbit 293 20.4.5 Mesenteries of the colon of the rabbit 294 20.5 Anatomy of the caecum of the Lagomorpha, especially the rabbit 294 20.5.1 Microbes in the caecum of lagomorphs 298 20.5.1.1 Hard and soft faeces, refection and caecotrophy in rabbits 298

 xvii

20.6 Anatomy of the caecum in hares and jackrabbits 301 20.6.1 Hard and soft faeces, refection and caecotrophy in hares 302 20.7 The caecum in Ochotonidae 302 20.7.1 General remarks 302 20.7.2 Anatomy of the caecum in pikas 302 Index

305

V Laurasiatheria

21 to 23 General remarks on Erinaceomorpha, Soricomorpha and Pholidota The Laurasiatheria (Murphy et al., 2001a, b; Kriegs et al., 2006; Hu et al., 2012) consist of the orders Carnivora, Perissodactyla, Artiodactyla and Chiroptera, which will be presented later, as well as of the following orders with genera and species, as listed by Wilson and Reeder (2005): 1.  Erinaceomorpha (hedgehogs) with 10 genera and 24 species. 2.  Soricomorpha (shrews) with 45 genera and 428 species. 3. P  holidota (pangolins) have only one genus, including eight species. Due to the admirable efforts of Schultz (1965), there is remarkable amount of information referring to the Tab. 5.1: Schematic compilation of caecal anatomy in Laurasiatheria. The numbers to the right of each drawing refer to the following literature sources: 4. Huntington (1903), 8. Jacobshagen (1937), 12. Takahashi and Yamasaki (1972), 16. Starck (1982), 17. Amasaki et al. (1989). Order

Suborder

Erinaceomorpha

Erinaceidae

Soricomorpha

 

Chiroptera

 

Pholidota

Manidae

Carnivora

Feliformia

Carnivora

Caniformia

Perissodactyla Ariodactyla

Caecum

     

 

 

 

 

Catacea

Mysticeti

Cetacea

Odontoceti

DOI 10.1515/9783110562217-001

     

digestive tracts of the Chiroptera, but for the other three orders, especially for the Erinaceomorpha and the Pholidota, there is no reliable comparative information with detailed illustrations, especially not for the large intestine. Information on the morphology of the digestive tract in these orders is very unbalanced. It represents an additional complication that a caecum in Laurasiatheria can be totally absent in some orders (Erinaceomorpha, Soricomorpha and Pholidota) (Tab. 5.1), or present in some and absent in other species of an order (Chiroptera, Cetacea: Odontoceti).

21 Erinaceomorpha 21.1 Introductory remarks and notes on food of Erinaceomorpha This order consists of 10 recent genera and includes 24 species, 16 belonging to the subfamily Erinaceinae (Hutterer, 2005a), which are spiny hedgehogs, and eight to the Galericinae or moonrats and Gymnures (Gould, 1995; He et al., 2012) without spines. The latter authors consider Erinaceidae as “the oldest known living placental mammals”, “extending back to the early Paleocene of North America” (He et al., 2012, page 1). The monophyletic Erinaceinae can be found in Eurasia and Africa, the Galericinae, which are also monophyletic (Gould, 1995), live in South and Southeast Asia, as well as on Philippines and Indonesian islands. Maps showing the geographical distribution of spiny hedgehogs have been published by Reeve (1994). A short account of the history of African Erinaceidae and their two subfamilies Erinaceinae and Galericinae was published by Butler (1985). Food of Erinaceus europaeus, the western hedgehog (Erinaceomorpha) consists largely of invertebrates at ground level, e.g. earthworms, carabid beetles, caterpillars, spiders, slugs. Occasionally vertebrates are eaten, e.g. frogs, lizards, young rodents, and nestlings as well as bird’s eggs and carrion including fish and some plants, fruits and fungi (Macdonald and Barrett, 1993). Southern African hedgehogs (Atelerix frontalis) are omnivores (Skinner and Chimimba, 2005). Invertebrates (e.g. beetles, earwigs, grasshoppers, termites, slugs, snails, centipedes, millipedes, moths and earthworms) form the bulk of their diet. They also consume small vertebrates (e.g. mice, lizards, frogs, the eggs and chicks of ground-living birds) and some vegetable matter, including fungi.

310 

 V Laurasiatheria – 21 Erinaceomorpha

Describing the biology of Atelerix albiventris (four-toed hedgehog of East Africa, Kingdon, 1974a writes Erinaceus albiventris) that the author identifies insects, earthworms, snails and slugs as favourite food, but a wide range of other animal and even vegetable materials are also eaten, such as eggs and ground-nesting birds, small mammals, frogs, reptiles, crabs, fruit, fungi, roots and groundnuts have been reported. These findings have more recently been corroborated by the investigations of Santana et al. (2010). This means that this species has to be classified as omnivorous. Very similar data for the food of Atelerix frontalis, the Southern African hedgehog, are given by Skinner and Chimimba (2005). Erinaceus concolor, the Eastern European hedgehog, is also called “omnivore” by Qumsiyeh (1996) – they “eat almost everything given to them” (page 63); KryŠtufek and Vohralík (2001) speak of omnivory. Grosshans (1983) mentions for Erinaceus europaeus, the West European hedgehog, that beetles, Coleoptera, represent the staple food in Schleswig-Holstein (northern Germany). Plant material is only eaten accidentally; sometimes seeds and fruits are taken, but during the rest of the year they do not contribute significantly to the food. Only in autumn fruits play a role as food constituents (Holz and Niethammer, 1990a). Investigating western European hedgehogs in England Dickman (1988) found that young animals take the food from the entire spectrum of available prey types, but older animals sample prey from a narrow range of material. According to Holz and Niethammer (1990b) coleopterans are important food material in Erinaceus concolor, the southern white-breasted hedgehog. Atelerix algirus, the North African hedgehog (according to Saint Girons [1969] the latter species can also be found in southern France) prefers insects, but small reptiles can also be ingested (Holz and Niethammer, 1990c). In contrast to their appearance in the diet, adult beetles were less important in E.  europaeus in total energy terms, but retained an important position along with caterpillars and earthworms (Reeve, 1994). In the subfamily Galericinae, the food seems to be similar to that of the spiny hedgehogs. For example, Sheng et al. (1999) mention that Neotetracus sinensis (authors write Hylomys sinensis) feeds “on various insects and a variety of roots and stems” (page 4). Recently, Pereira et al. (2016) published a comparison of the gastrointestinal tracts of the rodent Jaculus jaculus, the lesser Egyptian yerboa, and Paraechinus aethiopicus, the desert hedgehog, an erinaceomoph species. The information presented in that paper is rather limited. The hedgehog eats mainly insects with some additional plant material and the surface enlargement factor in its stomach is significantly lower than

in the yerboa, but for the gut (small plus large intestines), such differences are not clearly given. A caecum is absent in the hedgehog, but the other sections of the gastrointestinal tract do not show great differences between both unrelated species.

21.2 Anatomy of the stomach of the Erinaceomorpha, including arterial supply Some information is available for the gastric anatomy of Erinaceinae. According to Oppel (1896), there is no prominent incisura angularis on the lesser gastric curvature of Erinaceus europaeus. A cardiac gland zone is present, but very small and not well differentiated. Around the cardia, there is a 2- to 3-mm narrow zone of cardiac mucosa. This has been illustrated by Pernkopf (1937) (Fig. 5.1 A, black), but cannot be found in the picture supplied by Zhukova (2001) (Fig. 5.1 B). The zone of proper gastric glands (Pernkopf, 1937) extends over 90% of the total internal gastric surface; the pyloric gland zone, close to the effluent aperture of the stomach, is small. A hedgehog will eat a third of its body weight in one night (Kingdon, 1974a), according to Rossolimo (1955), fide Zhukova (2001), the daily food consumption in Erinaceus europaeus equals approximately 42% of the body weight. Erinaceidae have a voluminous and saccular stomach (Pernkopf and Lehner, 1937, Fig. 5.2) and it is possible “to consume large amounts of food in short periods of

Fig. 5.1: Gastric outlines and mucosal linings from species belonging to two subfamilies of Erinaceomorpha. The different signatures refer to cardiac mucosa (black, only in A), proper gastric and pyloric gland mucosa. Modified from: Pernkopf (1937) (A) and Zhukova (2001) (B, C, D).

21 Erinaceomorpha 

 311

obtained 25% as fibrous material from chitin. A food with a high load of “fibrous” component is highly indigestible, but can be degraded because of the chitinase in the digestive tract, especially in the stomach. In addition, the stomach provided the only site for digesta retention (Clemens, 1980), and thus might enable chitin degradation through gastric chitinases.

21.3 Small intestine and colon in Erinaceomorpha

Fig. 5.2: Internal aspect, showing the tunica muscularis, of the opened stomach of Erinaceus europeaus. Adapted from Pernkopf and Lehner (1937).

time” (Zhukova, 2001, page 393). Girgiri et al. (2015) published an illustration of the stomach of Atelerix albiventris, the four-toed hedgehog, and gave some measurements on 12 animals: The lesser curvature is considerably shorter (mean: 3.32 ± 0.37 cm) and the greater curvature is 7.42 ± 0.95 cm long, so that the organ forms a pouch-like bag. Zhukova (2001) also published illustrations of two species of the Galericinae, namely Echinosorex gymnura (moonrat, Fig. 5.1 C) and Hylomys suillus (short-tailed gymnure, D). In all four panels of the compiled illustration, the border between the proper gastric mucosa (dark rectangles) and the pyloric glands (lighter octagons) has been drawn at the aboral end of the transitional zone that was identified by Pernkopf (1937) and Zhukova (2001). Only in the drawing published by Pernkopf (1937) cardiac glands are depicted as a narrow “ring”, marked black, around the oesophageal opening. Cornelius et al. (1975), working on Carnivora, write that most vertebrate species that eat organisms containing chitin (e.g. insects, fungi), synthesise chitinolytic enzymes in their digestive systems, principally secreted by the gastric mucosa, in some species also by the pancreas. Chitinolytic enzymes are not of bacterial origin (Jeuniaux, 1961). In the higher vertebrates, the capacity of a given species to synthesise chitinases seems also to be related to the nature of the diet: Species which are strongly adapted to a diet entirely devoid of chitin, do not secrete chitinases in their digestive system (Cornelius et al., 1975). On the other hand, species with chitinases in the gastric mucosa, such as the western hedgehog in Europe, can potentially use chitin as a source of dietary “fibre” (Jeuniaux, 1962b). Graffam et al. (1998) determined in the four-toed hedgehog, Atelerix albiventris, that the diet dry matter of the food

The morphology of small intestine and colon in hedgehogs is not sufficiently documented. It is difficult, and in most cases even impossible, to differentiate these sections of the gut that can be so easily identified in humans and many other mammals. Recently, Girgiri et al. (2015) published a short and cursory study intended to provide “a baseline on the morphometry of the visceral organs” (page 32). The small and large intestines seem to differ in diameter according to their names – the diameter of the small intestine is slightly less than that of the colon. As has already been mentioned in the introductory words on Erinaceomorpha, the border between both intestinal sections is not marked by the presence of a caecum. This fact has already been indicated by an older, but now invalid taxonomy, where Lipotyphla (including hedgehogs) were characterised by the absence of a caecum. In addition, information on the colon of Erinaceomorpha is from the publication of Mitchell (1905). The hindgut consists of a short, straight, and rather wide rectal portion. Zhukova (2001) believes that the ancestral state of the intestinal differentiation in Erinaceomorpha is indicated by the presence of villi in the large (!) intestine.

21.4 Blood vessels supplying or draining the colon of Erinaceomorpha Information on the vascular supply or drainage in the Erinaceomorpha is very rare. Malinovský and Bednářová (1985) published information on the arterial supply of the stomach in the southern white-breasted hedgehog, Erinaceus concolor. From their data, a diagram of the branches of the A. coeliaca was drawn (Fig. 5.3). They showed that four arteries, the A.  hepatica communis, the A.  gastrica sinistra and two vessels, which they call right and left lienal artery, arise from the coeliac artery; here, one could speak of a “quadrupes”, different from the “tripus”, as can be found in humans (Lippert and Pabst, 1985). In Erinaceus europaeus, a common truncus coeliacomesentericus can be found (Barone, 1972). It divides into

312 

 V Laurasiatheria – 22 Soricomorpha

Fig. 5.3: Branches of the A. coelica in Erinaceus concolor. Adapted from Malinovský and Bednárová (1985).

the A. coeliaca and the A. mesenterica cranialis or superior. This is comparable with the situation in humans, where Lippert and Pabst (1985) found a common trunk with branches to the stomach and into the upper mesenterial artery in 2% of the investigated cases. In his detailed study, Mitchell (1905) deals with the venous drainage and general pattern of another representative of the Erinaceomorpha, the Algerian hedgehog, Atelerix algirus. However, an illustration of the venous drainage is not supplied, but Mitchell (1905) mentions the pattern of the intestinal tract of this species. The anterior (superior) mesenteric vein supplies the small intestine and the posterior (inferior) mesenteric vein drains the straight section of the colon. The aboral branches of the V. mesenterica superior meet the oral branches of the V.  mesenterica inferior, which means that this straight part represents the colon transversum. However, a delineation of the colon ascendens against the jejunoileum and against the colon transversum is not possible with the help of the information supplied by Mitchell (1905).

22 Soricomorpha 22.1 General remarks All phylogenetic analyses based on nuclear data strongly support the monophyly of a group including hedgehogs (Erinaceidae), shrews (Soricidae) and moles (Talpidae), as Cabria et al. (2006) write. Together with the family Selenodontidae the Soricidae and Talpidae belong to

the order Soricomorpha (Hutterer, 2005b). The Solenodontidae consist of one genus with four species (Wilson and Reeder, 2005). Within the family Soricidae or shrews 26 genera and 376 species can be discerned and there are 17 genera and 39 species in the family Talpidae, moles and their kin, as is listed in Wilson and Reeder (2005). Overviews for taxonomy and systematics of Soricomorpha have been published by Shinohara et al. (2003), Berman et al. (2007) and Jenkins (2013). In the detailed handbooks published by Niethammer and Krapp (1990) and Braun and Dieterlen (2005), different authors give information on the type of food of Talpidae and Soricidae. Both insectivorous groups consume protein-rich food of animal origin, which is mainly composed of invertebrates and only small amounts of plant material are taken (Myrcha, 1967): Talpidae prefer to eat Lumbricidae (earthworms), as well as insect larvae, slugs and snails. On the other hand, the Pyrenean desman, Galemys pyrenaicus, living close to open water, ingests one-fifth of its body weight per day in the form of aquatic articulates, as well as fish (Juckwer, 1990). Species belonging to the family Soricidae eat beetles, other insects and small vertebrates and some avoid earthworms, such as Sorex minutus, the pygmy shrew (Hausser, 1995). On the other hand Lumbricidae contribute importantly to the food of some species, such as Crocidura leucodon (bicoloured shrew), as Nagel and Nagel (2005) remark. The relative amount of daily intake, compared with body weight, can be remarkable. For example, Sorex minutissimus, the lesser pygmy shrew, ingest the double to fivefold amount of their body mass (Sulkava, 1990). Neomys fodiens, the northern water shrew, has a daily intake of 116% of its body mass (Spitzenberger, 1990) and adult common shrews, Sorex araneus, take 45–77% (Hausser et al., 1990). Dealing with the Soricomorpha in North America Berman et al. (2007) remarked that members of this order live in forested habitats where precipitation is high, for example, in the Pacific Northwest and Southern Appalachians. In studies of the biodiverse fauna of the Himalayan Mountain soricid community, Jenkins (2013) found a broad range of ecomorphological adaptations within this eutherian family.

22.2 Short remarks on the digestive physiology of Soricomorpha Several species of the order Soricomorpha are at or near the lower size limit for homeothermic animals, which implies a high metabolic rate (Reumer, 1989). The two subfamilies Soricinae and Crocidurinae behave

22 Soricomorpha 

metabolically different in relation to their environment. Vogel (1980) gave an extensive overview of metabolic levels and biological strategies of shrews: Soricinae (“hot shrews”) have high metabolic rates compared to Crocidurinae (“cold shrews”). That author believes that this difference may be related to different geographical origins of both subfamilies: Holarctic region for Soricinae, palaeotropical region for Crocidurinae, thus representing ecophysiological adaptations to differing climatic conditions. The Crocidurinae live in a warm, tropical climate and a cold, seasonal climate is inhabited by the Soricinae (Genoud, 1988).

22.2.1 Family Soricidae, Subfamily Crocidurinae and Myosoricinae 22.2.1.1 Food of Crocidurinae and Myosoricinae Apart from a negligible amount of molluscs, the diet of the Soricomorpha, for example, Crocidura suaveolens (the lesser white-toothed shrew), contained mainly arthropods, especially insects (Bauerová, 1988). According to Rowe-Rowe (1986), Crocidura flavescens, the greater red musk shrew, and Myosorex varius, the forest shrew, eat only arthropods. The insect orders most frequently recorded in the stomachs of Myosorex varius were Coleoptera (adults and larvae), Hymenoptera, Orthoptera, and larvae of Lepidoptera. In the family Soricidae, the subfamily Crocidurinae can be found, which is named after Crocidura, the largest mammalian genus of no less than 164 species (Motokawa et al., 2000), widely distributed in the Palaearctic, Ethiopian and Oriental regions. In his taxonomic reference, Hutterer (2005) differentiated 172 species of the genus Crocidura. Beetles, other insects and in some cases also Lumbricidae, as well as small vertebrates, contribute to the food of Soricidae (Hausser, 1995; Nagel and Nagel, 2005). The food of some “cold shrew” species (sensu Vogel, 1980), members of the genus Crocidura, was also studied on continents other than Europe. Boonzaier (2012) and Boonzaier et al. (2013) studied the reddish-grey musk shrew from southern Africa (Crocidura cyanea), which eats beetles, crickets and pseudoscorpions, as well as a small amount of leaf, seed or other plant material. Small lizards, along with termites and spiders have also been reported as food of Crocidura cyanea. Brahmi et al. (2012) mention for the Algerian greater white-toothed shrew, Crocidura russula, that insects contribute 48.9%, centipedes 29.7% and lizards 14.9% to the food biomass. The authors classify this soricid species as a generalist and opportunistic insectivore. Suncus murinus, the crocidurine Indian

 313

musk shrew, eats a similar food, consisting of insects, spiders, small mammals, worms, small frogs and rodents (Balakrishnan and Alexander, 1979). 22.2.1.2 Gastric anatomy and remarks on digestion Allison (1948) is one of the few authors who give a description of the stomach of a crocidurine species, Suncus varilla, lesser dwarf shrew (Allison writes “S. orangiae”). He mentions that the dilatation in the region of the fornix gastricus is followed in aboral direction by a long, constricted, tubular portion (Fig. 5.4). In many shrews, the pyloric part of the stomach is much prolonged. In different species of Suncus, the stomach was found to vary somewhat in size and shape (Allison, 1948). The fundic dilatation is followed aborally by a long, thin, tube-like region. At first sight, this tubular portion of the stomach appears to be the most proximal part of the duodenum, but closer inspection by that author revealed a slight pyloric constriction. According to Allison (1948), the wall of the stomach is unusually thin because the muscular coat is poorly developed. However, a small pyloric sphincter is formed. A most comprehensive account of the stomach types that can be found in the order Soricomorpha – and in addition also in the Erinaceomorpha – has been published by Myrcha (1967). This author gives information on 19 species of the Soricidae and five species of the Talpidae. In practically all 11 species of the genus Sorex depicted by that author, the pars pylorica of the stomach is considerably elongated (see the gastric outline of Sorex araneus in Fig. 5.5, adapted from Myrcha, 1967). This author defined elongation with the help of the length of the gastric region

Fig. 5.4: Outlines of stomachs in three species of Crocidurinae. Adapted from: Allison (1948) (A), Boonzaier (2012) (B) and Oppel (1896) (C).

314 

 V Laurasiatheria – 22 Soricomorpha

lined with pyloric mucosa plus the length of a “transition zone” between the area of proper gastric glands (he speaks of “fundic glands”) and the pyloric gland zone. The total length of the stomach is given in the illustration in centimetres, but Myrcha (1967) does not characterise the precise points between which the total length of this organ had been measured. The data in Fig. 5.5 have therefore to be taken as approximations. The contribution of the pars pylorica plus transition zone, expressed as percentage of the total gastric length is generally relatively high in the Soricidae. For example, the percentage of the “tubiform” section of the stomach can contribute more than half of the gastric length in Sorex araneus, but can also be as low as about 30% in Neomys fodiens). At the end of this chapter dealing with Soricomorpha, some general remarks on gastric elongation in that order will be made, based primarily on the discussion published by Myrcha (1967).

Fig. 5.5: Total length of the stomach and percentage of the pars pylorica in relation to the total gastric length, shown for two families of the Soricomorpha. Adapted from drawings and calculations by Myrcha (1967).

Not only the general shape of the stomach of the lesser dwarf shrew (Suncus varilla) has been described by Allison (1948), that author also gives an account of the mucosal lining of that organ. Four main regions of mucosa may be distinguished (Fig. 5.6). The squamous stratified epithelium of the oesophagus ends at a narrow zone of cardiac glands, some 8 mm wide, which surrounds the cardiac opening. “Farther from the cardia, chief cells become rapidly more abundant until, in the true fundus, they make up the lower half of each gland” (Allison, 1948). The proper gastric gland zone represents the internal mucosal lining of the fornix (= fundus) gastricus. “As the limit of the fundic region is approached, chief cells diminish in number, and parietal cells occur in the lower third of glandular tubules. In the transitional region there are no chief cells;…parietal cells are found in the basal part of the glands, and mucoid cells make up the rest of the glandular epithelium. Parietal cells diminish in number as the true pyloric region is reached, and the glands of the pylorus have only mucoid cells” (Allison, 1948, page 253). It is common in the literature to describe the junction between the fundus and the pylorus as abrupt. However, in each of the insectivore species examined, the fundopyloric transition is gradual. In Fig. 5.5, an abrupt demarcation line between the proper gastric mucosa (rectangles) and the pyloric mucosal region (lighter octagons) is given, but in Fig. 5.6, an extended transition zone is shown according to Allison (1948). Takeuchi and Yoshioka (2004) speculate that a valvelike mucosal fold protruding into the oesophageal lumen of Suncus murinus, the Asian house shrew, facilitates the regurgitation of the stomach contents. According to Horn et al. (2013), Suncus murinus is able to vomit. It is not clear whether the differentiation of a “gastric groove”, lined by columnar epithelium might also contribute to vomiting in that species.

Fig. 5.6: Gastric mucosal lining in Suncus varilla. The transition zone between proper gastric and pyloric gland mucosa is very wide. Adapted from Allison (1948).

22 Soricomorpha 

22.2.2 Family Soricidae, Subfamily Soricinae 22.2.2.1 General remarks Sorex minutissimus (Eurasian least shrew, body mass between 1.4 and 2.8 g, Silva and Downing, 1995) has an intensively elongated stomach (Hanski, 1984) with considerable development of the gastric mucous membrane (Myrcha, 1967). This is consistent with a maximal or near maximal metabolic rate. In S.  minutissimus, throughput time through the digestive tract varies from 12 min to 1 h, depending on the diet (Pernetta, 1976). Neomys fodiens, the European water shrew, is larger (12.0–16.8 g, Silva and Downing, 1995) and in this species 80% of the fed larvae of Tenebrio molitor (mealworm beetle) were excreted within the first two postprandial hours (Kostalecka-Myrcha and Myrcha, 1964c). A tube-like section of the digestive tract without special dilatations or folds, but with peristaltic waves, is a structure well adapted to efficient transport (Langer, 1991; Langer and Takács, 2004). The highest food requirements per body weight are found in the smallest species (Fig. 5.7) (Hanski, 1984). This corroborates statements by Myrcha (1967). Sorex minutissimus approaches a size limit (weight generally below 2.5 g and maximally 4 g, Sulkava, 1990) below which it becomes impossible to assimilate food fast enough to compensate for high metabolic rate. Shrews that weigh less than 5 g eat proportionately much more than the larger species (Saarikko, 1989); for example, Sorex minutissimus ingests up to 200% of its body weight in 24 hours (Hanski, 1984). 22.2.2.2 Food of Soricinae Sorex araneus, the common shrew and Sorex minutus, the Eurasian pygmy shrew, are clearly insectivorous, but show an essentially opportunistic character and reflect

Fig. 5.7: Ants and sawfly pupae as food in six species of Soricinae with different body sizes, expressed as percentage of grams dry food per grams of body mass. Adapted from Hanski (1984).

 315

the major traits of the supply of invertebrates in the study area (Bauerová, 1984). Two North American species of the genus Blarina eat a similar, but not completely identical type of food: Food of B. carolinensis, the southern shorttailed shrew, consists mainly of insects. Earthworms represent just 14.8% of the total gastric contents (Genoways and Choate, 1998). On the other hand, earthworms and other annelids, as well as millipedes, make up the major portion of the food of B.  brevicauda, the northern short-tailed shrew (George et al., 1986). The diet overlap between Sorex araneus, the common shrew, and S. minutus, the Eurasian pygmy shrew, is small (Pernetta, 1976); both species live in the same habitat. The two species have dietary areas peculiar to themselves, and competition for food between them is reduced: The pygmy shrew is shown to feed on the surface fauna, whilst the common shrew feeds more underground, eating more burrowing animals, depending on a varied diet of invertebrates with a preponderance of earthworms, molluscs, beetles and spiders (Churchfield et al., 2012).

22.2.2.3 Gastric anatomy and remarks on digestion Saarikko (1989) writes on page 415 of his publication that – in relation to body size – Soricinae have a relatively larger stomach than other insectivores. The pyloric gland section of the stomach, where protein rich meat is digested, is particularly elongated and large in Sorex (Myrcha, 1968) (Fig. 5.8). The extensive “tubiform” section of the stomach is not only formed by the pars pylorica, but also encompasses the distal part of the corpus region. A comparative illustration of stomachs of Sorex araneus and Neomys sp., based on illustrations from Pernkopf (1937) and Zhukova (2001), can be found in Fig. 5.8. Stomach size and the ratio of the surface area of the stomach to the body weight are largest in the smallest species of shrew. According to Myrcha (1968), the main reason for the differences in the relative stomach sizes lies in the different amounts of food eaten. In practically all species of the Soricomorpha, but especially in the Soricidae, subfamily Soricinae, the pars pylorica of the stomach is considerably elongated (Fig. 5.5). Myrcha (1967) quantified this elongation by determining the length of the gastric region lined with pyloric type of mucosa plus the length of a “transition zone” between the area of proper gastric glands (he speaks of “fundic glands”), which are marked in the illustration with small quadrangles, and the pyloric gland zone. The total length of the stomach is given in the illustration in centimetres, but Myrcha (1967) does not characterise the precise points between which the total length of this organ had been measured. As in the case of the subfamily Crociduridae,

316 

 V Laurasiatheria – 22 Soricomorpha avoid “coprophagy” and “refection” and speak of “rectumlicking” in Soricidae. It should be mentioned that Geraets (1980) observed shrews to eructate gastric contents, chew it and swallow it again. In this case, the term “rumination” (“Wiederkauen”), as it was applied by that author, seems to be appropriate. The observation is surprising because shrews are not herbivores.

22.2.3 Family Talpidae, short overview

Fig. 5.8: Stomachs of two species of the subfamily Soricinae. Adapted from: Zhukova (2001) (A) and Pernkopf (1937) (B).

the results published by that author and depicted in Fig. 5.5 for the Soricinae, can only be considered as approximations. Pfeiffer and Keith (1985) investigated the ultrastructure of the stomach of Blarina brevicauda and came to the conclusion: “In general the cell types present and cytologic character of the gastric mucosal, submucosal, and muscularis cells were similar to that reported from other mammalian species” (page 315). 22.2.2.4 Rectum-licking In studies on Sorex araneus, Crowcroft (1952) made the following observations (page 627): During the process of rectum-licking the rectum emerges as a stout firm tube, 5–10 mm long, which curves slightly forward towards the mouth. The open end of the everted rectum is nibbled and licked for a period which does not exceed 10 minutes. Immediately after this the stomach and the first few centimetres of the intestine, but not the remainder of the intestine, are found to be filled with a milky fluid containing numerous fat globules and small fragments of undigested food. Geraets (1980), who studied the genera Crocidura and Suncus, was also able to observe rectumlicking. This process took place during day or night, but generally only happened once during an observation period of 8 to 12 hours. In Sorex araneus, Loxton et al. (1975) speak of “coprophagy”. This term, as well as the term “refection”, also used by Crowcroft (1952), are inappropriate because faeces are not swallowed, as in truly coprophagous rodents (Hörnicke and Björnhag, 1980). Hirakawa and Haberl (1998) observed a “milky white fluid” to be licked from the everted rectum. They

There are three recent subfamilies within the Talpidae, the New World moles or Scalopinae (seven species according to Hutterer, 2005b), the Talpinae (28 species) with moles and desmans of the Old World, as well as the Chinese shrew moles or Uropsilinae (4 species). An overview given by Shinohara et al. (2003) deals with the phylogeny of the Talpidae. The authors suggest that seven major clades evolved in the New and Old World; they give names of exemplary genera: 1. Asiatic shrew-like moles (Uropsilus), 2. North American partly aquatic, partly fossorial (digging) moles (Condylura), 3. North American fossorial moles (Parascalops, Scalopus, and Scapanus), 4. North American semi-fossorial shrew moles (Neurotrichus), 5. Japanese semi-fossorial shrew moles (Dymecodon and Urotrichus), 6. European semi-aquatic desmans (Desmana), and 7. Eurasian fossorial moles (Euroscaptor, Mogera, and Talpa). In a genomic study, Cabria et al. (2006) concentrated their study on a desman species, the rare Iberian endemic Galemys pyrenaicus, but discussed the Laurasiatheria in total and elucidated the phylogeny of the whole superorder. It has already been mentioned in the introductory section on Soricomorpha that Talpidae, represented by Talpa europea, prefer to eat earthworms, but they also eat insect larvae, slugs and snails (Niethammer, 1990; Muschketat, 2005). However, Hartman et al. (2000) mentions for Scalopus aquaticus, the eastern mole from North and Central America, that it feeds on larvae of scarabid beetles, followed by ants and centipedes; earthworms were only found in approximately 8% of the investigated stomachs. As in the Soricidae, already cited above, Myrcha (1967) also gave a valuable account of the gastric anatomy in the Talpinae (Fig. 5.5). According to that author, the pars pylorica plus transition zone, expressed as percentage of the total gastric length, contribute less to the gastric form

22 Soricomorpha 

Fig. 5.9: Two drawings of the stomach of Talpa europea. Adapted from: Zhukova (2001) (A) and Pernkopf (1937) (B).

than in the Soricidae. In Condylura cristata, the star-nosed mole of North America only one fourth belongs to the “tubiform” section of the stomach, but in the European mole about one-third; the stomach of the Talpa europaea has a bag-like shape (Fig. 5.9) (Pernkopf, 1937; Zhukova, 2001).

22.2.4 Concluding remarks on gastric elongation in the Soricomorpha As early as 1868, Owen already wrote, “in many shrews the contracted pyloric part of the stomach is much prolonged” (page 427). According to Myrcha (1967), a very conspicuous growth in relation to body size in gastric volume and in the area of proper gastric glands can be observed.

 317

The pyloric portion undergoes a strong elongation, which is particularly well seen in the genus Sorex (Figs. 5.5 and 5.8), but was also found in the crocidurine Suncus varilla (Fig. 5.4). The growth in length of the pyloric region brings about a linear increase of the pyloric gland area and the transitional zone between the fundic and pyloric glandular regions. The pyloric glands generally produce mucous (Welsch, 2003), which lubricates the gastric surfaces and facilitates smooth transit of digesta containing exoskeletons of articulates. On the other hand, the linear and superficial enlargement of the transition zone of glands, which contain numerous parietal cells producing hydrochloric acid, is responsible for the action of active proteolytic enzymes (Myrcha, 1967). Saarikko (1989) mentions that it is the pyloric gland section of the stomach, where protein rich material is digested. Talpidae eat less food in relation to the weight of their body than the Soricidae (Myrcha, 1967). For this reason they need less digestive surface in their stomach, which can be bag-like, approximating a sphere with a relatively small surface. However, the zone with proper gastric glands is relatively extensive (Fig. 5.5). An increase in the area of the transition zone, which is characteristic of the Soricidae and especially of the species of the genus Sorex, cannot be observed in moles, Talpidae (Myrcha, 1967).

22.2.5 Blood vessels of the digestive tract in Soricomorpha The information available for the arterial supply or venous drainage of the colon of Soricomorpha is even more limited than in hedgehogs and their kin (Fig. 5.10). According to López-Fuster and Ventura (1997), an arterial ramus colicus from the truncus coeliacomesentericus can be found in Crocidura russula (A), the greater white-toothed shrew. In the same species and in Sorex sp. the formation of an

Fig. 5.10: Branches of the Aorta abdominalis in two soricomorph species. Adapted from López-Fuster and Ventura (1997).

318 

 V Laurasiatheria – 23 Pholidota

indirect branch via the A. mesenterica cranialis is possible (B). As there is no caecum (Mitchell, 1905), it is unclear whether the ramus colicus supplies colonic regions that can be homologised with either the ascending or transverse parts.

22.2.6 Colon of the Soricomorpha In Sorex minutus and S.  araneus, as well as in Crocidura horsfieldii (Horsfield’s shrew), all living on animal diet, the large intestine does not have a caecum or it is very short (Kurohmaru et al., 1982; Wilczynska, 1998). Despite this shortness, this segment forms the faecal pellets and absorbs water. This does not agree with the statement of Langenbeck (1996) that the large intestine is absent in the four shrew species he investigated (Crocidura russula, C.  suaveolens, Suncus murinus and S.  etruscus). Absence of the large intestine is improbable. In this case, the shrew would be unable to produce faeces or to absorb water. 22.2.6.1 Colonic villi in Soricomorpha Although the surface enlargement is smallest close to the anus, villi can be identified in the distal colon of Crocidura russula, C.  suaveolens, Suncus murinus and Suncus etruscus. In Suncus araneus, villi are responsible for small enlargements of the surface of the tunica mucosa in the large intestine (Jaroszewska and Wilczynska, 2006). In Suncus murinus, villi are not present in the large intestine during foetal development and early postnatal periods (Kiso et al., 1991), but circular folds and zigzag ridge-like folds contribute to the increase of the surface area in the large intestine. According to the same author goblet cells and absorptive cells differentiate just before birth. Thus, individual epithelial cells in the large intestine of Suncus murinus prepare for the onset of postnatal function during the last few days of gestation. However, absorptive cells do not show evidence of active postnatal endocytosis. Although villi were lacking in any stage (Kiso et al., 1991), many ridge-like structures were present on the mucosal surface of the large intestine.

23 Pholidota 23.1 General remarks Pholidota, also called pangolins or scaly anteaters, are independent from Xenarthra (Koenigswald, 1999), but

it will be indicated below that previous investigations created some confusion concerning the relationship between Pholidota and Xenarthra. According to Schlitter (2005), the order Pholidota consists of one family, the Manidae with one genus, Manis, and eight species (Gaudin et al., 2009). Four are from Asia and four from Africa (Schlitter, 2005). Two Asian pangolin species belong to the subgenus Manis (crassicaudata and pentadactyla) and two to the subgenus Paramanis (culionensis and javanica). In Africa, two species belong to the subgenus Smutsia (gigantea and temminckii), the African ground pangolin, and one each to Uromanis (tetradactyla), the African fourtoed pangolin and Phataginus (tricuspis), the African tree pangolin. Asian species are monophyletic and clearly separated from African species, which are also monophyletic (Gaudin and Wible, 1999). According to Rose et al. (2005), Pholidota have never been diverse or particularly common. “Fossil pangolins are too rare at present to allow a coherent picture of their phylogeny” (page 75, Koenigswald, 1999). The geologically oldest manid in Europe, Eomanis waldi, occurs in the middle Eocene in Messel near Darmstadt, Germany (Storch, 1978b); it is highly probable that Manidae originated in the Old World. Horny scales, which are typical for Pholidota (“scaly anteaters” or “Schuppentiere”) have been described in Eomanis waldi from Messel by Storch (1981) and Koenigswald et al. (1981). It has already been indicated above that there has been some confusion concerning the relationship between Pholidota and Xenarthra. The confusion arises because scaly anteaters (Pholidota) and Armadillos (Cingulata) are insectivorous, mostly myrmecophagous (ant- and termiteeating), lack teeth and have similar body shapes. They are adapted to the nutritional value of ants and termites. Although termite workers and soldiers tend to be high in ash and low in fat, alate ants (i.e. with wings) and termites, as well as other insect larvae and pupae, have much higher percentages of fat (Redford and Dorea, 1984) and thus represent a high-quality food. The classification of Manidae, created by the similar feeding types, has changed during the studies of this mammalian group. For example, Storch (1981) calls Eurotamandua joresi from the Middle Miocene of the “Grube Messel” an Old World Xenarthran and Gaudin and Branham (1998) speak of Eurotamandua as sister taxon to the Pilosa with “the most likely allocation…as a somewhat aberrant but true xenarthran closely allied to the Pilosa” (page 259/260). Eomanis waldi, as an early pangolin (Storch, 1978), is preserved in Messel alongside Eurotamandua (Savage and Long, 1986). Rose (2006) separates the order

23 Pholidota 

Pholidota clearly from the order Xenarthra and groups Eomanis and Eurotamandua close to the Pholidota. Szalay and Schrenk (1998) go even further: The European genus Eomanis krebsi cannot be classified within the Xenarthra; it is a juvenile specimen of Eurotamandua joresi. In their revised classification, Gaudin et al. (2009) propose that both Eurotamandua joresi and Euromanis krebsi belong to “uncertain” suborders, but are clearly members of the order Pholidota – and not Cingulata (Xenarthra).

 319

The African Manis (Smutsia) temminckii preys on only four ant and two termite species (Pietersen, 2013). They are selective in their prey choice and do not necessarily feed on the most abundant species. Ants comprise, according to Richer et al. (1997) and Swart (2005) the bulk of the ground pangolin’s diet, but termites are an important food source as well. The latter author writes that 96.7% of the available time was used feeding on ants, only 3.3% on termites, but monthly variation could also be observed.

23.2 Remarks on systematics, biogeography 23.3 Gastric anatomy of Pholidota and food in Pholidota There is no special relationship between the New World edentates and the Old World pangolins. Bugge (1979) compared the cephalic arterial supply of New World edentates and Old World pangolins, which differ markedly. A paper on Manis pentadactyla, the Chinese pangolin, by Pilliet (1891) contains many misinterpretations. According to that author, this is a species that eats leaves. By extrapolating from the ruminant stomach this publication is strongly biased. Manis pentadactyla is found in a wide area in South and East Asia from Nepal to Vietnam and Taiwan (Duckworth et al., 2008c). Paramanis (Manis) javanica, the Sunda pangolin, ranges in Southeast Asia from the Malay Peninsula to western Indonesia (Duckworth et al., 2008b) and has to be separated, as Feiler (1998) and Gaubert and Antunes (2005) write, from Manis culionensis, the Philippine pangolin. In Africa, pangolin species live in different habitats. Uromanis (Manis) tetradactyla, the long-tailed pangolin, is the most arboreal of the African species (Hoffmann, 2008b); Phataginus (Manis) tricuspis, the tree pangolin of West and Central Africa is semi-arboreal (Hoffmann, 2008c) and Smutsia (Manis) gigantea, the giant pangolin, is terrestrial according to Hoffmann (2008a), as well as the ground pangolin, Manis (Smutsia) temminckii (Pietersen, 2013). Kingdon (1971) deals with three of the African species, the giant, the ground and the tree pangolin. The Pholidota or pangolins are not just insectivores, but strictly myrmecophagous animals. According to Skinner and Chimimba (2005), they spend 96.7% of the time feeding on ants and only 3.3% on termites. The giant pangolin is able to dig forcefully and can open termitiaria, but the ground pangolin does not normally spend great energy in excavating. It is highly selective, feeding mainly of juvenile stages of ants and termites. Despite their differing habitats, all species feed on ants, termites and other invertebrates (Hoffmann, 2008a, b, c).

At the beginning of the 19th century, Cuvier (1805) investigated the Chinese pangolin species, Manis pentadactyla, and found an internal fold in the middle of the lesser curvature, which separates the gastric cavity into two compartments, which were not visible in all investigations, e.g. Fig. 5.11. When present, the proximal chamber, the fornix, into which the oesophagus opens, is thin-walled, the distal one, the pars pylorica, is thick-walled and narrows towards the pylorus. It is interesting that stomachs of Asian and African species of Pholidota show differences. In the African Phataginus (Manis) tricuspis and Uromanis (Manis) tetradactyla, which are depicted in Fig. 5.12, the external aspect of the stomach shows a fornix gastricus in a unilocular stomach, but as the picture of Pernkopf and Lehner (1937) shows, the stomach is divided into a pars cardiaca and a pars pylorica (Klinckowström, 1895).

Fig. 5.11: Illustrations of stomachs of two Asian species of Pholidota. Adapted from Oppel (1896), Pernkopf and Lehner (1937) and Krause and Leeson (1974).

320 

 V Laurasiatheria – 23 Pholidota

Fig. 5.13: Mucosal lining of the stomach of Manis javanica the Sunda pangolin. No cardiac glands could be demonstrated. Adapted from Klinckowström (1895). Fig. 5.12: Illustrations of stomachs of two African species of Pholidota. Adapted from Oppel (1896), Pernkopf and Lehner (1937) and Ofusori et al. (2008).

The pars cardiaca is lined with an intensively cornified squamous epithelium, as is mentioned by Ofusori et al. (2008) for Phataginus (Manis) tricuspis and depicted in the picture adapted from Oppel (1896). A sharp border or margin separates this from the pars pylorica, best seen in the picture from Pernkopf and Lehner (1937). The distal section of the cone-shaped pyloric part is lined with pyloric glands (Klinckowström, 1895), but according to this author proper gastric glands in Phataginus (Manis) tricuspis cover only an oval field in the area of the curvatura major. The stomach of Manis javanica is very different (Klinckowström, 1895, and Fig. 5.13): Proper gastric glands can be found in the large gastric gland, an invagination of the gastric mucosa. Weber (1892) mentions the remarkable formal similarity of the gastric glands of Manis javanica (Pholidota) and Trichechus manatus (Sirenia). Pyloric glands in Manis javanica can be found in three regions of the stomach (Weber, 1892, Oppel, 1896) (Fig. 5.13): 1. in the middle of the curvature minor, 2. directly caudad of the tubular opening of the abovementioned gastric gland, 3. close to the pylorus on the torus pyloricus and on the mucosal side opposite to that “trituration organ” (Weber, 1892, Oppel, 1896, Klinckowström, 1895). Krause and Leeson (1974) deal in some detail with the stomach of the Chinese pangolin, which is depicted in Fig. 5.11. “The stomach of Manis pentadactyla is fusiform

in shape, consists of a single chamber, and appears similar to numerous other mammalian forms. The wall of the proximal one-half of the stomach is thin…and apparently is devoid of glands. The entire stomach is lined by a keratinised stratified squamous epithelium. The distal one-third of the stomach exhibits several peculiarities. An area 3.5 cm in diameter is localised in the submucosa of the greater curvature” (marked in Fig. 5.11, upper right, with black arrowheads). “The large glandular masses are comprised of numerous gastric glands or tubules which show an abundance of parietal cells. Chief cells occur in greater numbers towards the base of each gland. Small mucous glands are found near the point where gastric glands drain into the lumen of the stomach. The lumen of the pylorus is extremely narrow. Projecting into the lumen from the mucosal lining are numerous hard spines. Orifices of small mucous glands are found near the proximal spines” (Krause and Leeson (1974, pages 2 to 4). It should be remarked that Nisa et al. (2005) studied the mucosal lining of Manis javanica under special consideration of endocrine gastric cells, but the function of these cells is not yet clear, therefore this paper is only of limited value. Cuvier (1805) and Weber (1892), as well as Flower (1872) and Heath (1992a) remark that the thick wall of this pyloric part in Manis sp. contains – near the middle of the great curvature – a large gland, which reminds Cuvier (1805) of a gland in the stomach of the beaver, Castor sp. A common excretory duct forms the terminal section of this gland (Owen, 1868; Flower, 1872). In the long-tailed pangolin or “phatagin”, Uromanis (Manis) tetradactyla, from West Africa a comparable gland could not be found by Cuvier (1805). No clear differentiation between African



21, 22, 23 Erinaceomorpha, Soricomorpha and Pholidota 

and Asian species is made by Rapp (1852), Owen (1868) and Flower (1872), but they state that the greater part of the gastric cavity is lined with a squamous, non-glandular epithelium, which is continuous with that of the oesophagus i.e., it is lined with a squamous tunica mucosa. The wall of the pars pylorica is generally very muscular and seems to function like the gizzards of birds. This “functional gizzard”, combined with a toothless mouth, is responsible for maceration and trituration of the insect food in the pars pylorica of the stomach before it enters duodenum. This has been mentioned for Smutsia (Manis) temminckii, the ground pangolin of South and East Africa by Heath (1992b) and for the Asian species Paramanis (Manis) javanicus (Evgenyeva and Dang Ngock Can, 2013), Manis pentadactyla (Heath, 1992a) and Manis crassicaudata (Heath, 1995). A hard spherical tissue mass at the pylorus grinds or macerates food (Heath, 1992a) in Manis pentadactyla and Evgenyeva and Dang Ngock Can (2013) speak of teeth at the pylorus of Manis javanicus. According to Pernkopf and Lehner (1937) the torus pyloricus on the side of the lesser curvature is lined with small pointed “teeth” and cornified squamous epithelium.

23.5 Arteries of the gut in Pholidota

23.4 Form of the small and large intestines in Pholidota

23.6 Concluding remarks on the colon of Erinaceomorpha, Soricomorpha and Pholidota

In the Pholidota, the hindgut colon is relatively short relative to the small intestine (Fig. 5.14), very wide and strongly marked with longitudinal striae and shows no trace of division into colon and rectum. According to Mitchell (1905), the pattern of the intestinal tract of Manis is of striking simplicity because of the loss of the caecum and the longitudinally folded hindgut.

Fig. 5.14: Venous drainage of the digestive tract in Manis tricuspis. Modified after Mitchell (1905).

 321

The present author did not find modern information on the arterial supply of the gastrointestinal tract of Pholidota. However, Mitchell (1905) presents an illustration on the digestive tract of the tree pangolin, Manis tricuspis, together with its venous drainage (Fig. 5.14). The portal system is simple, consisting of a single superior mesenteric vein curving round the jejunoileum and receiving numerous tributaries, a duodenal branch and many branches from the intestinal tract. There is a single rather large posterior mesenteric or rectal vein. Mitchell (1905) creates a problem in his drawing by leaving a considerable gap between the distalmost branch of the curving V.  mesenteria superior or cranialis and the area of supply of the V.  mesenterica inferior or caudalis (Fig. 5.14). Is this a zone of anastomosis or just the artificial product of dissection or of the drawing? One can only speculate that this zone “between” both mesenterial veins is homologous to the zone of the colon transversum in other mammals, including man. In addition, there is no trace of division into colon and rectum (Mitchell, 1905).

Before the colon in Erinaceomorpha, Soricomorpha and Pholidota is generally discussed, its specific functions that are found in many Eutheria, including man, should be recaptured: – Retention and storage of digesta – Absorption of nutrients as products of degradation further orally – Electrolyte absorption (Kerlin and Phillips, 1983). – Water absorption before faeces are voided – Mucus secretion as a protective agent (BustosFernandez et al., 1983) Retention of digesta in the colon does not seem to play an important role in the three orders. In the colon of Erinaceomorpha, Soricomorpha and Pholidota, special wall differentiations for digesta retention like taeniae and haustra are absent. Absorption of nutrients, electrolytes and water has to take place in the small intestine, where villi, covered with an absorptive epithelium with microvilli are differentiated. According to Kiso et al. (1991), the large intestine in Suncus murinus, the Asian house shrew, does not play a considerable role in the transepithelial transport of proteins. The colon is not only reduced in

322 

 V Laurasiatheria – 24 Chiroptera

function, but also shows considerable anatomical regression. This anatomical regression produces considerable problems, especially when a caecum is not differentiated and the colon cannot be clearly separated from the small intestine. It has already been mentioned that the mucosal lining cannot be used to differentiate a section of the digestive tract. The distribution of different types of mucosal lining in a species, e.g. squamous non-secreting and different types of secretory or absorptive epithelium, can be characteristic for a certain species, but the presence of a certain type of mucosa is not necessarily characteristic for a special section or region of the gut. This statement is also true for the colon in Erinaceomorpha, Soricomorpha and Pholidota. It is one of the aims of this compilation to obtain information on the morphological delineation between the small and large intestines in the Eutheria. The interrelationship between the gut sections and their supplying and draining blood vessels, especially the arteries, was taken into account. To a lesser extent the study of draining veins could be applied, but because veins are morphologically less conservative than arteries, they are less appropriate to identify gut sections. To differentiate sections of the colon, branches of the A. mesenterica superior (cranialis) and inferior (caudalis) have to be investigated. The most distal branches of the A. mesenterica superior or cranialis supply the proximal part of the colon, and the distal part of the colon and proximal section of the rectum is supplied by the A. mesenterica inferior or caudalis. However, it has been shown in the paragraphs on blood vessels supplying or draining the colon that this system is confused by a considerable variability in branching modes. Especially at the borders of gut sections a clear coordination between blood vessels and gut sections is not always possible. Generally, a blood vessel can be identified and named and the identity of the supplied or drained gut section can be determined. For example, the A.  colica sinistra as a branch of an A. mesenterica inferior (caudalis) supplies the proximal part of the rectum and the terminal section of the colon, which is homologous with the colon descendens. However, what can be said about the colon ascendens and transversum? The border between colon transversum and colon descendens cannot be clearly characterised because both are very short. The present author feels the vicious cycle when blood vessels are identified under consideration of the area they supply or drain and when the gut sections have to be named under consideration of their vascular supply. We have to accept that gut sections in Pholidota cannot be clearly delineated and are therefore morphologically “undetermined”. This is a clearly frustrating situation.

24 Chiroptera 24.1 Introductory remarks “Bats make up more than 20% of extant mammals” (Teeling et al., 2005, page 580); in sub-Saharan Africa they represent the most diverse group of mammals (Fenton, 2013). Approximately 71% of the total number of chiropteran species (1116 according to Wilson and Reeder, 2005) are insectivores, 23% frugivores, 5.3% are nectarivores pollen-feeders, 0.7% are carnivores and, very specialised, only 0.3% are sanguinivores, i.e. they feed on blood (Neuweiler, 1993). A statement of Kulzer (2003) that phylogenetic relationships within the order Chiroptera were still unresolved during the time when he wrote his text still seems to be valid. An estimate of the phylogenetic relationships among a large number of species of bats has been published by Jones et al. (2002). A relatively recent tree showing relationships among bat families has been published by Agnarsson et al. (2011). This tree has been pruned by the present author and adapted to the form depicted in Fig. 5.15. In this diagram, the family Pteropodidae (in the suborder Megachiroptera, Dobat and Peikert-Holle, 1985; Campbell and Kunz, 2006) represents a sister group to all other bats, combined as “Microchiroptera” (Forman, 1972; Kingdon, 1974a; Dobat and Peikert-Holle, 1985; Neuweiler, 1993; Schober, 1996). Gu et al. (2008) and Almeida et al. (2011) speak of “megabats” and “microbats”. The position of Pteropodidae as sistergroup to all other microchiropteran bats is not generally accepted. A study by Gu et al. (2008) indicates that two microchiropteran families, the Rhinolophidae and Hipposideridae, are phylogenetically more closely related to the Pteropodidae than to other microbats. A similar observation, grouping Pteropodidae and Rhinolophidae close together, was made by Teeling et al. (2005). The evolution of echolocation was certainly a key innovation (Dumont, 2007) in Microchiroptera. Pirlot (1977) states that insectivorous and nectarivorous Chiroptera exhibit hovering flight, which is probably related to their feeding mode. For example, Simmons and Geisler (1998) suggest that the earliest members of the microchiropteran lineage probably used vision for orientation and obstacle detection in their arboreal/aerial environment, and probably foraged by gleaning insects and perhaps some fruits and other edible items from foliage and from the ground. A thorough account of chiropteran food and feeding has been published by Kulzer (2005). He mentions that about 70% of all bat species are insectivorous. In a very detailed study, Ferrarezzi and Gimenez (1996) bring

24 Chiroptera 

 323

Fig. 5.15: Tree of chiropteran families and food characteristics. When a certain food class is taken, this is marked with one, two or three “X”, according to intensity. Information on food and on the tree is from Agnarsson et al. (2011), Ferrarezzi and Gimenez (1996) and Kingdon (1979). Stomach outlines adapted from: Robin (1881) (A); Schultz (1965) (B); Bhide (1980a) (C); Kamiya & Pirlot (1975) (D); Forman (1972) (E); Forman (1971b) (F); Roux and Glass (1970) (G); Perrin and Hughes (1992) (H).

phylogeny and feeding types in chiropterans together, by listing presence or absence, as well as intensity of different feeding types. A modified version of that table has been included in Fig. 5.15. In addition, outlines of stomachs, taken from eight different publications (Robin, 1881; Schultz, 1965; Rouk and Glass, 1970; Forman, 1971b, 1972; Kamiya and Pirlot, 1975; Bhide, 1980a; Perrin and Hughes, 1992) are also presented in the illustration. In their text, Ferrarezzi and Gimnenz (1996) propose that herbivory – not insectivory – is the symplesiomorphic feeding style for bats; as a consequence of this hypothesis, Pteropodidae as frugivorous, nectarivorous, pollenivorous and folivorous bats form a separate branch of the tree. Concentrating on insectivorous bats, Freeman (1981) presents an extensive list of the percentages of insect orders that contribute to chiropteran food. The same author also deals with the hardness of food, consisting either of beetles or moths, and its influence on the chewing apparatus, especially jaw modifications. Freeman (1988) suggests that describing foods in terms of their texture may be more important in oral design than a characterisation as frugivore, insectivore or carnivore.

In Fig. 5.15, a remarkable differentiation of the food in the family Phyllostomidae is indicated, showing preponderance of insectivory, as well as some frugivory, nectarivory, pollenivory, folivory and even sanguinivory. Feeding specialisation is accompanied by morphological, physiological and behavioural adaptations (Datzmann et al., 2010). Because of this wide food range an additional diagram for some species of Phyllostomidae has been compiled from a diagram that was originally published by Datzmann et al. (2010) (Fig. 5.16). Generalised food characteristics, which are listed on the right side of the diagram, are also taken from the same authors. The outlines of the gastric forms are taken from six different publications (Huxley, 1865; Rouk and Glass, 1970; Forman, 1971b, 1973; Mennone et al., 1986).

24.2 Systematics, phylogeny, zoogeography of Chiroptera The origin of the order Chiroptera and the evolution of mammalian flight is poorly understood (Storch, 1999).

324 

 V Laurasiatheria – 24 Chiroptera

Fig. 5.16: Tree of some species of Phyllostomidae and food characteristics. Numbers at branching points indicate million years before present. Letters F, N, O represent timepoints for origins of frugivory, nectarivory and omnivory, respectively. This information is from Datzman et al. (2010). Gastric outlines adapted from: Huxley (1865) (A); Forman (1973) (B); Forman (1972) (C); Forman (1971b) (D); Mennone et al. (1986) (E); Rouk and Glass (1970) (F).

In their craniological studies, Wible and Novacek (1988) came to the conclusion that Megachiroptera and Microchiroptera evolved monophyletically within the order Chiroptera. Laryngeal echolocation for orientation and food acquisition evolved in the common ancestor of bats, but was subsequently lost in Megachiroptera (Springer et al., 2001). According to these authors, molecular dating suggests that the last common ancestor of bats lived 52 to 54 million years ago around the Palaeocene-Eocene boundary. The fossil record also extends back to the Early Eocene in Microchiroptera and to the Early Oligocene in Megachiroptera. Storch (1999) argued that the order started to evolve well before its appearance in the fossil record. According to Marshall (1983), bats arose, presumably monophyletically, in the early Tertiary, the Megachiroptera soon diverging from the Microchiroptera. Megachiroptera today feed upon floral resources, fruit and leaves from a great number of plant genera (Marshall, 1983). They may effect both pollination and seed-dispersal and are thus useful to flowering plants (“Chiropterophily”, dealt with in detail in Dobat and Peikert-Holle, 1985). By the Cretaceous-Tertiary boundary the major groups of modern angiosperms were present, some of these probably being pollinated nocturnally by large insects and non-flying mammals and others with seeds dispersed by terrestrial vertebrates (Marshall, 1983). Early bats were perhaps initially attracted to such flowers and fruit by the

insects found around them, later finding the plants themselves nutritious. Molecular data suggest, according to Eick et al. (2005), that Chiroptera may go back in time into the Late Cretaceous (approximately 136 to 65 MYBP, Murawaski, 1972). This was the era during which angiosperms diversified and became dominant resulting in radiation of pollinating insects, which are major prey items of bats. Cutworm moths, Noctuidae, are the largest family of Lepidoptera (Eick et al., 2005). They are characterised by hearing organs that supposedly evolved in direct response to bat predation. The evolution of aerial hunting of insects is the key innovation of microchiropterans. Simmons and Geisler (1998) believe that this foraging strategy is primitive for these chiropterans. There are indications that flying, echolocating, insectivorous bats may already have been present in the Late Cretaceous, about 75 million years ago (Eick et al., 2005). Because of the age of Microchiroptera the range of forms is greater than in any other suborder of mammals (Kingdon, 1974a). Also geographically extension of the range of bats happened. For example, diversification occurred within several lineages of microchiropterans (Emballonuroidea, Noctilionoidea and Vespertilionoidea, Lim, 2009) after colonisation of North and South America during the Tertiary. Storch (1999) subdivides the Microchiroptera into two infraorders, the Yinochiroptera and the Yangochiroptera.

24 Chiroptera 

The Yinochiroptera comprise the following microchioropteran families: Rhinopomatidae, Craseonycteridae, Nyceridae, Megadermatidae, Rhinolophidae and Hipposideridae. In the Yangochiroptera, we find Emballunuridae (this position according to Springer et al., 2001), Noctilionidae, Mormoopidae, Phyllostomatidae, Natalidae, Furipteridae, Thyroteridae, Myzpodidae, Vespertilionidae, Mystacinidae and Molossidae (Storch, 1999). According to Gu et al. (2008), the yinochiropteran Rhinolophidae and Hipposideridae are more closely related phylogenetically to the Pteropodidae, which are Megachiroptera. The present study does not mainly deal with eutherian systematics, but it should be mentioned here that bats can also be grouped differently. For example, Simmons (2005b) writes that within the Microchiroptera three clades emerged: 1. Emballonuridae, 2. Yinochiroptera (= Rhinolophidae + Hipposideridae + Megadermatidae + Nycteridae + Rhinopomatidae + Craseonycteridae), and 3. Yangochiroptera (Mystacinidae + Noctilionidae + Mormoopidae + Phyllostomidae + Myzopodidae + Thyropteridae + Puripteridae + Natalidae + Molos­ sidae + Antrozoidae + Vespertilionidae; Simmons and Geisler, 1998, 2002).

24.3 Food of Chiroptera in general As has already been mentioned above, a very wide range of food material is eaten by the Chiroptera, an order with about 1000 species (990 according to Neuweiler, 1993, and 1116 according to Wilson and Reeder, 2005). The importance of insectivory in chiropterans is related to a remarkable diversity of foraging methods in this order. For example, Simmons and Geisler (1998) propose that foraging behaviour in the chiropteran lineage evolved in a series of steps: The first one is characterised by gleaning food objects during nights from a perch (already mentioned above) by detecting prey by vision and/or listening for prey-generated sounds. Second, stationary prey is collected using echolocation and vision for orientation. The third type of food hunting uses echo location in prey detection and tracking for flying insects and vision and hearing for stationary prey. Finally, prey can be detected and tracked exclusively by using echolocation. Extrapolated from the morphological and physiological differentiations related with the digestion of cellulose, one might assume that chitin can be degraded by insectivorous mammals. Chitin – after cellulose – is the polysaccharide with the second widest range of

 325

distribution on earth. However, in bats there is no convincing proof that chitin is digested. The excellently preserved chitin pieces in their faeces speak against digestion of chitin (Neuweiler, 1993). In contradiction to this, Webb et al. (1993) write that 59% (!) of the chitin is absorbed during digestion. Fruits are generally considered to be poor sources of dietary protein (Steller, 1986). In this type of food, nitrogen is less available than in foliage or in invertebrate animals. Many frugivorous and nectarivorous bats therefore supplement their diet by consuming insects. Pollen, which is rich in protein, is another possible dietary source of nitrogen. The abundance of pollen, however, fluctuates seasonally and is annually unpredictable (Steller, 1986). An interesting functional parallelism between frugivory and folivory was mentioned by Lowry (1989): The black fruit bat, Pteropus alecto, chews leaves to a bolus, swallowing the liquid fraction, and expels the pellet of fibrous residue from its mouth. The extract so ingested contains about 51% of the crude protein of the eaten leaf material. This feeding behaviour occurs widely among Old World Pteropodidae. The strategy represents one extreme of the ways in which leaf material can be handled in the herbivore digestive tract; rather than dealing with leaf fibre, which is thus avoided. Fruit bats feed on fruit of different types (Qumsiyeh, 1996). In the Egyptian fruit bat (Rousettus aegyptiacus, Pteropodidae), the rate of uptake of glucose and fructose from this type of food is extremely rapid (Keegan, 1977) and especially the concentration of fructose in the systemic circulation is high. This causes an osmotic drag, which the small intestine has to resist. It will be seen later that there is a “virtual absence” of a large bowel in bats (Keegan, 1977), which has the functional value of an effective mechanism preventing excessive fluid loss. Another example of the difficulties to adapt to a food rich in sugars is demonstrated by investigation in the nectarivorous Pallas’s long-nosed bat, Glossophaga soricina (Phyllostomidae), by Herrera and Mancina (2008). When the sugar concentration increases in the nectar, the animals consume lower volumes. On the other hand, bats feeding on dilute nectar have to process excess water and are limited by their excretory capacity and the amount of water that can be eliminated through evaporation. Also, the blood licking vampire bats (Phyllostomidae, Subfamily Desmodontinae) 80% of the food, namely water, has to be absorbed from a specialised section of the digestive tract, a gastric blindsac (Neuweiler, 1993), before the cellular constituents of the blood can be digested

326 

 V Laurasiatheria – 24 Chiroptera

enzymatically. The nutritive value of the cellular constituents of blood is very stable and varies little. The above lines indicated that the food of the Chiroptera generally offers nutrients that are easily absorbed and are subjected to swift enzymatic degradation before being transferred into the animal’s metabolism. This process affords only a short time period with the consequence of a rapid transit through the digestive tract. For example, within 12 to 34 minutes were needed for fruits to pass the tract in Pteropus alecto, the black flying fox, and Pteropus poliocephalus, the grey-headed flying fox (Tedman and Hall, 1985a). In insectivorous species, the passage time can be 35 to 170 minutes. Because of the high passage rate, increased activities of digestive enzymes have to be expected (Neuweiler, 1993). In the frugivorous Microchiropterans, the digestion is different from superficial assumptions comparing them with herbivores. Neuweiler (1993) states that they are not strictly fruit eating, because these animals, such as flying foxes, drink the juice and spit out the fibrous material. Thus they ingest about 33% of the energy that is supplied by fruits. Plant-visiting bats, from both the New and Old World tropics do include leaf extracts as a regular or episodic part of their diets (Kunz and Ingalls, 1994). In doing so, frugivorous bats have circumvented the need for an enlarged or specialised gut that might otherwise have led to excessive wing load and high energy expenditure during flight. By ingesting only the protein-rich, soluble contents of leaves, and expelling the indigestible fibrous portions, frugivorous bats must be considered partially folivorous in the broadest sense. Frugivorous bats appear to be pre-adapted for folivory, given that their dentition and gut morphology are specialised for extracting and digesting a largely liquid diet (Kunz and Ingalls, 1994). An example for this is Rousettus sp. (Pteropodidae), which feeds not only on a great variety of fruit, nectar or blossoms (Kulzer, 1979), but also on leaves and other kinds of plant material. After feeding, their stomachs are found to be full of foliage. Kunz and Diaz (1995) investigated leaf-eating by neotropical bat Artibeus jamaicensis (Phyllostomidae). Leaf-eating has been reported for at least 17 species of Old-World Megachiroptera and four species of New World Microchiroptera. Leaves can be relatively high in protein (>19% dry-matter content) and low in fat (~1%). The observations that A. jamaicensis selects and chews leaves high in protein and rejects protein-poor, fibrous pellets, support the hypothesis that these bats extract liquid fractions which contain a reliable source of dietary protein. Folivory, once thought to be rare among plant-visiting bats, may in fact be quite common and widespread, especially among species that feed largely on fruits which are low in protein (Kunz and Diaz, 1995); they drink the fruit sap.

Neuweiler (1993) mentions that about 700 bat species (approximately 71% of the total number of species) live mainly on insects. Staliński (1994) investigated the insectivorous Myotis myotis, the mouse-eared myotis (Vespertilionidae) where an extremely rapid digestion could be demonstrated. About 23% are frugivorous and about 5% of all chiropterans live on nectar and pollen. Pteropodidae of the Old World tropics, which eat this type of food, have an ecological equivalent in the Phyllostomatidae of the New World tropics (Kulzer, 1982). Only 7 species of the Chiroptera are carnivorous (Neuweiler, 1993). Norberg and Fenton (1988) remarked that the tendency to feed regularly on small terrestrial vertebrates occurs in species in three families, the Nycteridae, Megadermatidae and Phyllostomidae. According to these authors, there are occasional records of a vespertilionids eating small terrestrial vertebrates. No single morphological or behavioural trait identifies the bats which regularly include small terrestrial vertebrates in their diets. Only 0.3% of all chiropterans lick blood (Desmodus rotundus, the common vampire bat, Diaemus youngi, white-winged vampire bat, and Diphylla ecaudata, hairy-legged vampire bat). The latter is specialised for feeding on the blood of birds (Eisenberg, 1989). The Chiroptera extend their range by widening their spectrum of food characters. For example, members of a large clade of Neotropical bats, the Phyllostomidae, with 160 species in 55 genera (Wilson and Reeder, 2005) feed on blood, insects, vertebrates, nectar, pollen, and fruits. Dietary evolution in phyllostomids appears somewhat more complex than previously thought (Wetterer et al., 2000). According to these authors, each feeding specialisation evolved once, but reversals did occur, such as loss of nectar-drinking or pollen-feeding. On the other hand, some specialisations may have evolved more than once (e.g. carnivory, which might further differentiate into piscivory, insectivory and even sanguinivory). According to Ferrarezzi and Gimenez (1996), the insectivorous condition represents one of the possible starting points of food types in the Chiroptera. Alternatively, the same authors propose herbivorous ancestry of the bats. From this herbivorous basic condition the insectivorous feeding habit might have been acquired in the Microchiroptera. The question of the evolution of sanguinivory is particularly interesting. Ferrarezzi and Gimenez (1996) speculate that sanguinivory originated through ancestral habits of preying on birds in trees. “All bats are restricted to nutritionally concentrated and easily absorbed foods with a high energy return” (Kingdon, 1974a, page 113). By this concentration on higher quality feeds bats circumvent the need for an enlarged and specialised gut that might otherwise have led to

24 Chiroptera 

high energy expenditure during flight. Leaves provide a high yield of protein per unit foraging effort (Kunz and Ingalls, 1994) and drinking of leaf sap can be interpreted as ingestion of “leaf extracts”. On the other hand, Rousettus sp., a species belonging to the Megachiroptera (Pteropodidae), feeds not only on a great variety of fruit, nectar or blossoms, but also on leaves (Kulzer, 1979), so that their stomach is found full of foliage. Megaloglossus woermanni, Woermann’s long-tongued fruit bat, another species of the Pteropodidae, practices to nectar-licking (Kulzer, 1982). Very similar adaptations to nectarivory can be found in the phyllostomid African long-tongued fruitbat and Glossophaga soricina (Microchiroptera) of South America. In the New World, the leaf-nosed bats or Phyllostomidae show a high richness in species and are characterised by a remarkable trophic diversity. Bonato et al. (2004) believe that ancestral feeding might have been insectivory with occasional inclusion of vertebrate prey. Many species of the Phyllostomidae are frugivorous (Estrada-Villegas et al., 2010). A species like Glossophaga commissarisi (Commissaris’ longtongued bat) is generally nectar-feeding, but facultatively frugivorous. Fruits appear to be the more energy efficient food items (Kelm et al., 2008). However, nectar appears to be the preferred food and fruits are ingested only during periods of low nectar availability and feeding experiments reveal a clear preference of Glossophaga commissarisi for nectar over fruits. An interesting aspect in relation to leaf-eating was presented by Kunz and Diaz (1995): Artibeus jamaicensis (Jamaican fruit-eating bat) eats leaves because most fruits are low in protein and consumption of the soluble fraction of protein-rich leaves could satisfy the daily protein needs. In other species of the Phyllostomidae, a widening of the food niche could be observed. For example, Nogueira et al. (2005) characterise Chiroderma sp., the big-eyed bat, as a seed-eater, having adapted from frugivory to granivory. On the other hand, Chrotopterus auritus, the woolly false vampire bat, is a carnivorous species, eating rodents, small birds and coleopterans (beetles); fruits were only occasionally consumed (Witt and Fabían, 2010). Finally, according to Arata et al. (1967) Carollia perspicillata (Seba’s short-tailed bat) extends its range from insectivory to carnivory and probably also practices cannibalism. For the Molossidae, or free-tailed bats, Freeman (1979, 1981) gives additional examples of extension of food ranges. Originally, insectivores that hunt “soft-bodied” insects, for example, butterflies and moths, can change to eat hard-shelled Coleoptera. To be able to chew this very special food material, these molossids have to change their skull morphology by developing “thick jaws” (Freeman, 1979) and differentiation of a well-developed cranial crest.

 327

24.4 General remarks on the gastric anatomy of Chiroptera In the digestive system of chiropterans, the “chief modifications are to be seen in the stomach” (Koopmann, 1994, page 8). The description of gastric anatomy in bats will be subdivided in the following and described for chiropteran families. It will have to be asked whether generalised statements are really valid. For example, according to Forman (1971a), frugivorous bats possess an elongate portion (relative to stomach size) between the cardia and the pyloric sphincter; this condition is shared in all frugivorous species and could be related to increased food intake. However, such statements which are based on a wide range of chiropteran species, certainly represent an improvement compared with ecological accounts that do not mention the digestive tract at all, although it represents one of the morphological differentiations that make a life-ensuring relationship between environment and species possible. For example, in a monograph on longeared bats, Swift (1998) makes some remarks on digestive efficiency, on the diet and a short note on teeth, but the rest of the digestive tract is not even mentioned! It should already be mentioned here, at the beginning of the discussion of chiropteran gastric anatomy, that confusing terms are often used in the descriptions without clear definitions. Rouk (1973) is an author who clearly defines the terms “cardiac caecum” and “cardiac vestibule”, but this is more the exception in different studies on gastric anatomy. Rouk (1973) on page 6 of his dissertation gives the following definition: Cardiac caecum represents the blind pouch or diverticulum that is part of the stomach and most distant from the pylorus. On the other hand, the cardiac vestibule is a dilatation at the end of the oesophagus. In the following, “cardiac caecum” will be replaced by a term which is also used in human (TA, 1998) anatomy, the “fornix gastricus”. Both cardiac vestibule and vestibulum cardiacum will be used. Certainly, one of the most detailed studies on the gastrointestinal tract of chiropterans has been published by Schultz (1965). Sections from page 349 of this eminent paper have been translated from German by the present author and will be cited extensively in due course. In the Chiroptera, we find very different forms of the gastric region, but the gut is rather simple and uniform. The primitive gastric shape, formed by a simple dilatation of the digestive tube, as well as by bending of the pyloric end from the longitudinal into transverse orientation, the oesophagus and the pylorus approach each other and form the endpoints of the so-called lesser and greater curvatures. Differentiations of this general shape

328 

 V Laurasiatheria – 24 Chiroptera

are primarily formed by a strong “outpocketing” of the fornix gastricus. The saddle between the blindsac and the oesophagus can be differentiated to a deep incisura cardiaca and can form, in continuation and dilatation of the oesophagus a special gastric region, the vestibulum cardiacum. Finally, the pars pylorica can be intensively lengthened.

24.5 Chiroptera, Pteropodidae 24.5.1 General remarks on Pteropodidae The Old World fruit bats or flying foxes of the family Pteropodidae have an extremely wide range of distribution in the tropics of the Old World (Almeida et al., 2011). From western through central into eastern Africa pteropodid species range (Kingdon, 1974a; Langevin and Barkley, 1990; Owen-Ashley and Wilson, 1998), extending into southeastern Africa (Boulay and Robbins, 1989) and even into subtropical South Africa (Kwiecinski and Griffiths, 1999; Acharya, 1992), as well as into the northern section of the Arabian Peninsula and Israel (Gray et al., 1999) and Rousettus aegyptiacus “appears to continue to expand its range in the Holy Land” (Qumsiyeh, 1996, page 79). Pteropus livingstonii is endemic to two of the Comoro Islands (Smith and Leslie, 2006). The Indian subcontinent, as well as the Southeast Asian peninsula is settled (e.g. by Cynopterus sphinx, Storz and Kunz, 1999), but species also extend the pteropodid range into the Malay Peninsula and the Indonesian Islands (Kunz and Jones, 2000; Campbell and Kunz, 2006; Hodgkinson and Kunz, 2006), as well as into the Philippines (Jones and Kunz, 2000). The northern and eastern margins of Australia are also inhabited by flying foxes (Strahan, 1983), as well as Pacific islands, like New Guinea and the Solomon Islands (Flannery, 1991), and the Fiji and Samoa Groups, extending as far east as the Cook Islands (Miller and Wilson, 1997; Banack, 2001). The food of Pteropodidae is generally of plant origin: fruits, flowers and their products, like nectar and pollen; leaves are also eaten in various proportions by different species. On the Micronesian Caroline Islands endemic members of the genus Pteropus exist. Ripe breadfruit (Artocarpus altilis) is apparently one of the most preferred food items of P. pelagicus (an obsolete name, which has been resurrected according to Buden et al., 2013). According to these authors, this species, as well as P. insularis, also take other fruit, such as bananas, coconuts, papayas, figures and pandanus fruits.

24.5.2 Gastric anatomy of Pteropodidae Because of the large number of pteropodid species (186 according to Wilson and Reeder, 2005), only a small selection, for which anatomical information is available, can be considered in the following. Overviews for the genus Pteropus, have been made available during more than 130 years (Fig. 5.17) by Robin (1881), Fischer (1909), Schultz (1965), Kamiya and Pirlot (1975) and Zhukova (2001). The proportions differ considerably between the different drawings. Comparing Pteropus alecto, the black flying fox, with P.  poliocephalus (grey-headed flying fox) Tedman and Hall (1985a) found little variation between the gastrointestinal tracts of the two species. Differentiations that are characteristic for this fruit bat genus, can clearly be differentiated (Kamiya and Pirlot, 1975) as cardiac, fundus (better fornix), corpus and pyloric areas. First, the oesophagus opens into a dilated and tube-like region, which belongs to the stomach. Robin (1881) calls this section the “cardiac portion” of the stomach and Forman (1990) speaks of a cardiac vestibule (vestibulum cardiacum). Second, the fornix gastricus forms a voluminous, expanded diverticle (Tedman and Hall, 1985a) and, third, an also expanded corpus gastricum, which narrows into an elongated tubifom pars pylorica, which can be confused with the small intestine and is folded back on itself (Robin, 1881). However, the compact arrowheads in the five

Fig. 5.17: Stomachs of three species of the megachiropean genus Pteropus. The compact arrowheads identify the pylorus. Adapted from five different authors.



24 Chiroptera, Pteropodidae 

 329

drawings (Fig. 5.17) show the position of the pyloric sphincter, which lies far down the tubular gastric section. Both cardiac and fornical regions have a relatively thick gastric mucosa and abundant parietal cells (Tedman and Hall, 1985a). It has been speculated by Zhukova (2001), who worked on Pteropus tonganus, the Pacific flying fox, that abundant parietal cells might be related with digestion of plant proteins, a speculation which is not really convincing.

24.5.2.1 Pteropus sp. Grosser (1901) published an informative illustration of the stomach of Pteropus vampyrus (edulis) (Fig. 5.18) together with the arterial supply. This shows very clearly that most of the tube-like section, which gives the impression of belonging to the small intestine (duodenum), is, in fact, supplied by the Aa.  gastrica dextra and gastroepiploica dextra. From this drawing by Grosser (1901) the present author produced a schematic diagram of the arterial supply of the stomach (Fig. 5.19). The truncus coeliacus produces a branch, which is called A. splenica, running along the facies visceralis of the stomach and enters the cranial pole of the spleen and can be followed along the hilus, producing Aa.  gastricae breves entering the wall of the greater gastric curvature. The A.  gastrica sinistra follows the lesser curvature and gives off branches to both sides of the stomach (parietal and visceral). The third branch of the truncus coeliacus, the A.  hepatica, gives off an A. gastroduodenalis, which divides into two branches, the A. gastrica dextra (which anastomoses with the A. gastrica sinistra) and the A. gastroepiploica dextra. To obtain an idea about the shape of pteropodid stomachs, comparing Pteropus sp. with other genera, an

Fig. 5.18: Visceral aspect of the stomach in the large flying fox, Pteropus vampyrus. Adapted from Grosser (1901).

Fig. 5.19: Arterial branches of the Aorta abdominalis in Pteropus vampyrus. Adapted from Grosser (1901).

additional diagram was drawn, based on a phylogenetic tree published by Almeida et al. (2011) (Fig. 5.20), as well as on gastric outlines from the literature. In two cases, the tree is unresolved, for Harpyionycteris and Pteropus, as well as for Eonycteris, Rousettus and a branch leading to Megaloglossus and Epomophorus. The range of food eaten by species belonging to the genus Pteropus is exclusively vegetarian, but nevertheless of different physical consistency. For example, in five species dealt with in the “Mammalian Species” series of the American Society of Mammalogist, there is one species, Pteropus samoensis, the Samoan flying fox, which is classified by Banack (2001) as a “generalist feeder”, mainly eating fruits and leaves, but also flower parts. Predominantly frugivorous are Pteropus hypomelanus, the variable flying fox (Jones and Kunz, 2000), as well as Pt. livingstonii (Comoro flying fox) (Smith and Leslie, 2006). On the other hand, Pt. tonganus, the Pacific flying fox, concentrates on pollen and drinks nectar (Miller and Wilson, 1997), similar to the large flying fox, Pt. vampyrus, which eats mainly flowers and drinks nectar (Kunz and Jones, 2000). The general gastric shape in the genus Pteropus is characterised by a variable shape of the fornix gastricus (Fig. 5.17) and a narrow and long corpus gastricum plus pars pylorica. The vestibulum cardiacum at the terminal oesophagus can be of considerable volume. 24.5.2.2 Harpyionycteris sp. The outlines that can be found in Fig. 5.20 give different information, for example, the direction of flow of digesta through the stomach (arrows) the position of the pylorus (bold arrowhead) and, in the case of Scotonycteris and Eonycteris, the position of the distal end of the ductus

330 

 V Laurasiatheria – 24 Chiroptera

Fig. 5.20: Stomachs of seven genera of Pteropodidae. Arrowheads indicate the pylorus and the stars in Scotonycteris and Eonycteris the ductus choledochus. The tree is adapted from Almeida et al. (2011). Gastric outlines: Schultz (1970) (A); Schultz (1965) (B); Robin (1881) (C); Zhukova (2001) (D); Dobat and Peikert-Holle (1985) [empty and filled] (E); Makanya et al. (2001) (F).

choledochus (star). The illustration of the stomach of Harpyionycteris whiteheadi, the harpy fruit bat, is based on a picture by Schultz (1970) that was mirrored and redrawn by the present author. Schultz (1970) writes that dental features indicate that this species does not exclusively live on fruits as other Megachiroptera. According to Ong et al. (2008), this species might depend on fruits. Eisentraut (1950) writes that H. harpyia (a species not mentioned anywhere else according to the present author’s knowledge) eats predominantly beetles with hard elythra. The phylogenetic affinities of the genus Harpyionycteris remain obscure, as Tate (1951) writes, but the genus

undoubtedly branches off from the pteropodid stem in very ancient times. This is in accordance with a tree representing the consensus of six most parsimonious trees published by Almeida et al. (2011), on which the tree in Fig. 5.20 is based. In a tree on phylogenetic relationships of Pteropodidae, Jones et al. (2002) demonstrate that Harpyionycteris sp. is positioned relatively close to Pteropus sp., which is also shown in the tree based on Almeida et al. (2011). In two other studies of megachiropteran phylogeny by Giannini and Simmons (2003) and Guan et al. (2006), Harpyionycteris sp. is not even mentioned. The ambiguity of food characterisation and phylogenetic



position is reflected in comments made by Schultz (1970), that the digestive tract of Harpyionycteris sp. exhibits partly characteristics of fruit-eating species in relation to shape and structure of the stomach and the length of the intestine. He does not expand on these interesting aspects and further anatomical studies are needed. 24.5.2.3 Scotonycteris sp. The stomach of Scotonycteris zenkeri, Zenker’s fruit bat, a frugivore (Mickleburgh et al., 2008a, b) from western Africa (Kuhn, 1961), has been depicted by Schultz (1965) and can also be found in Fig. 5.20. That author also gives a description of the simple form (“einfache Form”) of the stomach of S.  ophiodon (Pohle’s fruit bat), which, according to Mickleburgh et al. (2010) is a frugivorous specialist. The fornix gastricus of S. ophiodon is not prominent and the pars pylorica slightly lengthened, but does not form loops. In his illustration, Schultz (1965) indicates the position of the ductus choledochus (marked by a star) and it is natural that the pylorus (tip of the arrowhead) lies orad of this duct. 24.5.2.4 Eonycteris sp. A similar situation can also be demonstrated in the stomach of Eonycteris spelaea: (lesser dawn bat) that was originally depicted by Robin (1881). This plant-feeding species has, as Forman (1990) demonstrated, a cone-shaped vestibulum cardiacum, which is positioned parallel with the tubular pars pylorica and is situated between the oesophagus proper and lesser curvature of the stomach. The vestibule is widest at its union with the stomach at the lesser curvature. The fornix gastricus terminates in a distinct apex. It has a craniad direction and together with a long pars pylorica produces a C-shape. The pyloric tube tapers to a narrow junction with the duodenum (Forman, 1990). Remarks on mucosal lining of the stomach can be found further down, when Megaloglossus woermann is discussed. 24.5.2.5 Rousettus sp. Rousettus aegyptiacus, the Egyptian rousette, adds flowers and leaves (Kwiecinski and Griffiths, 1999) to its food, but prefer fruits (Skinner and Chimimba, 2005). The pars cardiaca, as Schultz (1965) calls the vestibulum cardiacum, is not very voluminous in this species. Especially the illustration, which was originally published by Zhukova (2001), shows this very clearly. It seems justified to assume that ingestion of fruit saps does not afford extensive gastric storing volumes, so that the vestibulum cardiacum can be small. Megachiropterans, in contrast to Microchiroptera, have wide zygomata and greater jaw musculature, indicating

24 Chiroptera, Pteropodidae 

 331

more generalised diets (Freeman, 1995). These bats probably did not evolve from some insectivorous ancestor. Megachiropteran nectarivory probably evolved from frugivory. Pollen can be taken both in the process of nectar-feeding as well as independently and is considered as the main source of protein in obligate nectarivores (Muñoz-Romo et al., 2005). It seems unlikely that nectar and tiny grains of pollen affect hard structures like teeth. Pollen is broken down chemically in bats by saliva and digestive juices in the gastrointestinal tract. Although pollen grain exine coats are highly resistant to degradation by digestive enzymes, as Herrera and Martínez del Rio (1998) found in studies of New World microchoropterans, pollen can be an important source of proteins, vitamins, and minerals. According to Muñoz-Romo et al. (2005), Phyllostomidae of the genus Glossophaga digested 64.2 and 71.3% of all the pollens fed. 24.5.2.6 Megaloglossus sp. Nectar is a very special diet of mammals (Kelm et al., 2008). Floral nectar constitutes a sugar-rich and highly digestible, but protein- and fibre-deficient, food source. Dietary constraints, such as a temporary scarcity of nectar, may sometimes require the uptake of alternative food items. In the IUCN Red List of Threatened Species, Mickleburgh et al. (2008b) writes that the pteropodid Woermann’s longtongued fruit bat, Megaloglossus woermanni, is “the obligate nectarivorous bat species in Africa”, which might perhaps eat some pollen (Kingdon, 1974a). Although pollen is of unique importance to flower-visiting bats as a nitrogen source, the carbohydrate fraction of the diet can be found in nectar and the anatomical modifications seen in the tongue of Old World pteropodid macroglossines, allow the bat to ingest much nectar in a minimum of time (Howell and Hodgkin, 1976). Nectarivorous bats ingest very dilute natural diets. Sucrose hydrolysis is not the limiting factor in nectar processing, but this is limited by the burden of excess water in the dilute solution. According to Herrera and Mancina (2008), the intake rates of sugar and energy decrease in diluted sucrose solutions. To accommodate the high energetic requirements of nectar-feeders, large amounts of nectar have to be consumed during feeding periods (Gonzalez-Terrazas et al., 2012). Gastric storing volumes can be represented either by the fornix gastricus or by the vestibulum cardiacum. Forman (1990) found that stomachs with large fornical regions have small cardiac vestibules, and vice versa. This relationship applies to nectar- and pollen-feeding species. Blindsacs of the fornix gastricus are well developed in most pteropodids. In Fig. 5.20, two illustrations (panel E), published by Dobat and Peikert-Holle (1985) (originally from Brown, 1962), show the empty and filled stomach of Megaloglossus

332 

 V Laurasiatheria – 24 Chiroptera

woermanni. The fornix gastricus, confusingly called “caecum” by Forman (1990), is relatively small in Megaloglossus and is cranially directed. In Eonycteris, it lies parallel to the pars pylorica, so that the whole stomach has a C-shape. The cardiac vestibule, the gastric fornix and the lengthened pars pylorica of the pteropodid stomach might not necessarily play a considerable role in digesta retention. In a study on pollen digestion, which they undertook in New World microchiroptera, Herrera and Martínez del Rio (1998) found that food was retained in the stomach for a “short period”. Forman (1990) gives an account of the gastric mucosal lining in four species of pteropodids subfamily Macroglossinae. Transition from oesophageal stratified squamous to gastric mucosa is abrupt in all genera and borders on a very narrow zone of cardiac glands. Fundic (oxynctic or proper gastric) glands occupy approximately 70% of the internal gastric surface in Megaloglossus, as well as in the genus Eonycteris. In the Pteropodid, genera Macroglossus and Melonycteris even as much as 73–76% of the stomach is lined with proper gastric glands (Forman, 1990), including the fornix gastricus, the cardiac vestibule, the corpus gastricum, and varied portions of the distal tubular stomach (pars pylorica). The narrow terminal pyloric “tube” contains mucous-producing pyloric glands and is devoid of parietal cells. According to Forman (1990), an asymmetrical pyloric sphincter (valvular flap on the greater curvature larger than that of the lesser) is present at the gastroduodenal junction in all pteropodid genera. 24.5.2.7 Epomophorus sp. The last branch shown in the tree in Fig. 5.20 leads to Epomophorus sp. The illustration (G) given in this diagram is adapted from a photo of the stomach of Epomophorus wahlbergi, Wahlberg’s epauletted fruit bat, and was originally published by Makanya et al. (2001). In a paper that is characterised by a chaotic, non-professional terminology, the authors write on page 75: “The stomach was the carnivorous type…”, probably indicating that the organ is unilocular. It is roughly U-shaped, has a voluminous corpus gastricum and a long pars pylorica. The pylorus, marked again by a bold arrowhead, is characterised by a prominent groove externally and by a big fold luminally (Makanya et al., 2001). The gastric mucosa has longitudinal folds that extend down to the pylorus and end abruptly at the very prominent transverse fold that marks the pyloric sphincter. 24.5.2.8 Concluding remarks on pteropodid stomach anatomy Reference is made here to Fig. 5.20. A U-shaped stomach with cardia and pylorus lying close together is characteristic for many species of the Pteropodidae, which also

tend to have a lengthened pars pylorica. The length of the fornix gastricus relative to the total gastric length from cardia to pylorus is highly variable. The same is also true for the volume of the vestibulum cardiacum. However, one has to ask how reliable the original gastric outlines are, on which the present compilation is based. For example, no vestibule is indicated in the drawings of empty and filled stomachs of Megaloglossus woermanni from Dobat and Peikert-Holle (1985). On the other hand, Forman (1990) shows this volume increase very clearly in this species. The information given on the gastric mucosal lining shows that the organ is a simple glandular structure.

24.5.2.9 Gastric digestion in Pteropodidae Eidolon helvum, the African straw-coloured fruit bat of the eastern part of Africa, is frugivorous (Okon, 1977). According to this author, the aboral 10 mm of the oesophagus has developed gastric glands; this statement, most probably, refers to the vestibulum cardiacum. The stomach is unexpectedly rich in proper gastric gland (zymogenic) and parietal cells. Active digestion of proteins appears to start from the oesophagus (vestibulum cardiacum?) (Ogunbiyi and Okon, 1976) and continues along the gastric canal by the action of both pepsin and trypsin. These authors noted that the oesophagus, which is usually believed to be concerned mainly with passing of food by peristalsis, was found to contain the highest activity of amylase. In addition, the oesophagus (vestibulum?) and the stomach had the highest activity of pepsin. This enzyme is usually found in the mammalian stomach where the HCl present provides a medium of low pH (~2) required for peptic action. The high activity in the oesophagus and stomach suggests the consumption of protein-rich diets by Eidolon helvum. This is unexpected considering that this bat feeds predominantly on sweet succulent fruits (Ogunbiyi and Okon, 1976). Another pteropodid bat, Cynopterus sphinx, the greater short-nosed fruit bat, is a frugivore living on the Indian subcontinent and Southeast Asia. It adds leaves to its diet and cellulose and xylan degrading bacteria were found in the intestine (Alwin Prem Anand et al., 2012). On the other hand, these authors used an insectivorous microchiropteran bat, Hipposideros fulvus, a leaf-noses species as a control and found no cellulose and xylan degrading bacteria in the digestive tract of this animal.

24.5.3 Small intestine of the Pteropodidae, form and function Villi are a characteristic feature of the small intestine in representatives of the frugivorous Pteropodidae (Tedman



and Hall, 1985a). Dealing with function – and, especially, with the morphology of the intestine – the investigators have to overcome the problem that there is no obvious demarcation between the small and large intestines in the fruit bats, as Nelson (1989) writes. However, Makanya et al. (1995, 1997) mention that the boundary between both gut sections in frugivores “can be taken as the beginning of macroscopically visible longitudinal rugae in the colon” (page 2416 in Makanya, 1997). Tedman and Hall (1985a) believe that anatomical characteristics of the gastrointestinal tract in Pteropodidae allow fruit bats to process large quantities of food rapidly. The anatomy of the absorptive structure in the digestive tract of bats is important to the quantitative aspect of digesta processing (Keegan, 1977). However, the mucosal lining cannot be used to differentiate sections of the gut from each other. The tunica mucosa “floats” on the tela submucosa and is not bound to theoretical limits between organ sections. Selim and El Nahas (2015) compared villous density and cellular composition of the tunica mucosa in the small intestine of the frugivorous Rousettus aegyptiacus (Egyptian rousette, Pteropodidae), as well as of the insectivorous Taphozous nudiventris (naked-rumped tomb bat, Emballonuridae), they found differences, but any clear functional interpretation could not be given. However, earlier investigations on form and function have supplied interesting insights: Small birds and bats, according to Caviedes-Vidal et al. (2007) have significantly shorter small intestines and a small intestinal “smooth bore tube” surface area (page 19132) than similarly sized nonflying mammals. The reduction in intestinal volume amounts to more than 50%, which is advantageous because the energetic costs of flight would increase with the load carried. In Pteropodidae, the load that has to be carried is reduced by a shorter retention time in the gut: Food can pass through it in 30 minutes or less, as Nelson (1989) states. A compensation for the short time of digesta transit might be a large absorptive area in the small intestine. This creates a central dilemma (Caviedes-Vidal et al., 2007), which may be compensated by application of an enhanced pathway for intestinal absorption of water-soluble nutrients such as glucose and amino acids. One way to compensate for rapid digesta transit in pteropodids might be that fruit bats have greater intestinal surface areas than insect-eaters (Makanya et al., 1995, 1997). Additionally, paracellular absorption can be found in bats according to Caviedes-Vidal et al. (2008). More than 70% of their total glucose absorption – much more than in non-flying mammals – can be absorbed paracellularly. Makanya et al. (1995, 1997) showed that absorptive adaptations occur at several levels of structural organisation. These

24 Chiroptera, Microchiroptera 

 333

include changes in intestinal length and diameter, villous enlargement and microvillous amplification. Fruit bats eat large quantities of fruit each day in order to meet the critical levels of the nutrients that are deficient in their diets. The large absorptive surface area may be important in maximising the absorption of digesta. For example, in Pteropus poliocephalus, the grey-headed flying fox, 65.6 microvilli per µm2 can be found in the small intestine, as Tedman and Hall (1985b) write; microvilli are unusually high (mean: 3.13 µm [1.36 µm in humans]). According to Keegan and Mödinger (1979), microvilli in the small intestine of Rousettus aegyptiacus are also very long and slender and result in absorption surface – per unit tube area – that is more than three times greater than in the rat. Mean microvillus length was 3.6 µm with a diameter of 0.099 µm in Rousettus, but 1.14 µm and 0.14 µm in the laboratory rat. Due to the microvilli the increase of the mucosal surface was a 57 fold increase in the bat, but only 18 fold in the rat (Keegan and Mödinger, 1979). Makanya et al. (1997), who studied Epomophorus wahlbergi, Wahlberg’s epauletted fruit bat, supplied interesting information on different levels of structural differentiations: For example, in the proximal small intestine the amplification factor of mucosal villi is 10.5 times, at the distal end only 3.14 times. On the other hand, the surface increase through microvilli is 50.4 times proximally and 38.0 time in the distal small intestine. In another pteropodid species, Lisonyceteris angolensis, the Angolan soft-furred fruit bat, the results concerning surface magnification are very similar to those determined in Epomophorus wahlbegi. Villous amplification factor on the proximal small intestine: 9.6, distal: 3.86; microvillous amplification factor proximal: 56.1 and distal: 33.5 (Makanya et al., 1997).

24.5.4 Large intestine in the Pteropodidae A rectum cannot be distinguished and the colonic mucosa is restricted to a short segment featuring prominent longitudinal folds. Some remarks on the caecum and colon of the fruit bats will be included in the section “Concluding remarks on the digestive tract of Chiroptera” (Section 24.16.).

24.6 Chiroptera, Microchiroptera 24.6.1 Gastric anatomy 24.6.1.1 Gastric anatomy of the Megadermatidae Schultz (1965) presents illustrations and descriptions of stomachs in Megaderma. The species M.  spasma and M. lyra have similar stomachs with a slightly lengthened pars pylorica, but Eisentraut (1950) depicts a globular

334 

 V Laurasiatheria – 24 Chiroptera

gastric form for M. lyra (he writes Lyroderma lyra). These forms are insectivorous and carnivorous (Fig. 5.15). In the figure, the stomach of a megadermatid species, Lavia frons, the yellow-winged bat, is depicted. It is a retort-like organ with a short pars pylorica. This bat of central and eastern Africa is an insectivore. It eats hard and soft-bodied insects (Kingdon, 1974a). 24.6.1.2 Gastric anatomy of the Rhinopomatidae Schultz (1965) describes the shape of the stomach of Rhinopoma hardwickii (lesser mouse-tailed bat) as a simple globular form, as should be typical for insectivorous bats. Bhide (1980a) gives a slightly different description of the stomach of Rhinopoma microphyllum (the author writes R.  kinneari), greater mouse-tailed bat. The illustration marked “C” shows a stomach that is 0.9 cm long and 0.5 cm broad at the corpus gastricum (the author writes “fundus”), which is the broadest region. A prominent fornix gastricus is not developed. The pylorus is sharply bent craniad so that the gastrooesophageal (cardia) and gastroduodenal junctions (pylorus) lie relatively close to each other. The mucosal folds lie along the long axis of the stomach in a zigzag pattern. 24.6.1.3 Gastric anatomy of the Hipposideridae This family has 81 species in 9 genera and Hipposideros with 67 species (Simmons, 2005a) is the largest genus within this family (Gu et al., 2008). Kamiya and Pirlot (1975) published an illustration of the stomach of the insectivorous Hipposideros bicolor, the bicoloured leafnosed bat. The authors show a unilocular organ, but indicate a slight incisure across the greater curvature in the corpus gastricum. This is the hipposiderid stomach shown in Fig. 5.15. For Hipposideros caffer, Sundevall’s leafed-nosed bat, Schultz (1965) depicts and briefly characterises a simple unilocular stomach. It is obvious in the illustrations of Kamiya and Pirlot (1975) and Schultz (1965) that the pars pylorica is not a lengthened tube. 24.6.1.4 Gastric anatomy of the Rhinolophidae In their study on the phylogenetics of three bat families, Gu et al. (2008) found a close relationship between Hipposideridae and the Rhinolophidae. In their description of the insectivorous Rhinolophus cornutus (little Japanese horseshoe bat), Kamiya and Pirlot (1975) concluded that the sophisticated sensory systems and in the flight patterns represent the “real” specialisation of these bats, but not the stomach and digestive tract (Fig. 5.15). Robin (1881) and Zhukova (2001) describe the stomach of Rhinolophus ferrumequinum (greater horseshoe bat) as a rounded unilocular organ with a prominent fornix

gastricus. The cardia lies halfway between fornix gastricus and pylorus. Zhukova (2001) adds that a vestibulum cardiacum is not differentiated. Zhukova (2001) speculates that this type of the stomach represent an ancestral state in this insectivorous species. Kolb (1954) studied the stomach of Rhinolophus hipposideros, the lesser horseshoe bat. Cranially the gastric tunica mucosa and tunica muscularis are thicker than further aborally. Longitudinal folds extend into the gastric lumen; in these insectivorous bats they are united by a few anastomoses (Robin, 1881). As the “lean” interval between periods of food uptake lies between 12 and 18 hours in Rhinolophus hipposideros, and because the food has a high content of voluminous and practically indigestible chitin, the stomach is very extensible (Kolb, 1954). In Rhinolophus ferrumequinum, Scillitani et al. (2005) found a change in the distribution of mucosal chief and parietal cells in the corpus gastricum. This suggests that the composition of gastric juice could change from the oral to the aboral area. The lack of chief cells in the aboral corpus, which was also described by Kamiya and Pirlot (1975), is coincident with a low percentage of pepsinogen in the gastric juice of that region. On the other hand, the increase in the number of parietal cells in the aboral direction suggests a maximum percent content of HCl in the aboral part of the corpus gastricum.

24.6.1.5 Concluding remarks on the gastric form of Megadermatidae, Rhinopomatidae, Hipposideridae and Rhinolophidae When one considers the gastric forms of those insectivorous microchiropterans that have as yet been briefly discussed (Fig. 5.21), one is inclined to repeat and generalise the statement, made by Makanya et al. (2001) in relation to Megadermatidae, that all four families have a stomach of the “carnivorous type”. This term, used here in the purely descriptive sense, means a unilocular organ with purely glandular mucosa. The diagrams make no indication of the functional aspect that storing volume of digesta can vary considerably. Pairs of illustrations that are comparable to those presented by Dobat and Peikert-Holle (1985) for Megaloglossus woermanni (Fig. 5.20 E) and show a filled and an empty stomach, are not available for these microchiropteran families. 24.6.2 Gastric anatomy of the Noctilionidae and Mormoopidae Illustrations of the gastric shape of Noctilio leporinus (Forman, 1972) (greater bulldog bat) have been published by Forman (1972) (who writes “N. labialis”), and Zhukova (2001) and for Mormoops megalophylla (Peters’ ghost-



24 Chiroptera, Microchiroptera 

 335

Fig. 5.21: Outlines of gastric forms in four chiropteran families. Adapted from: Schultz (1965) (A); Kamiya and Pirlot (1975) (B); Robin (1881) (C); Zhukova (2001) (D).

faced bat) by Forman (1971b). These illustrations have been redrawn in Fig. 5.22. Systematic relationships among various closely related families of North American bats (here Mormoopidae and Noctilionidae) have been reviewed by Smith (1972). This author stated, on the basis of several features of external and internal morphology, that the Mormoopidae are closely related with Noctilionidae and, perhaps to a lesser degree, with Phyllostomatidae. An examination of gastric structure in these two groups by Forman (1971b) provides support for the proposal by Smith (1972). In the papers of Forman (1971b, 1972), the gastric anatomy of noctilionid and mormoopid chiropteran species is discussed. The author mentions that fish-eating bats of the family Noctilionidae are similar in gross and histological features to Mormoopidae. In both families, elongation of the pars pylorica cannot be found. Stomachs of Noctilio leporinus and Mormoops megalophylla also resemble one another in having a short and pointed fornix gastricus, a

very narrow ring of cardiac glands and a relatively short terminal pars pylorica with limited distributions of pyloric glands (Forman, 1971b). Histologically, the slightly coiled gastric glands are long and narrow. The number of proper gastric gland cells or zymogenic cells is extensive. A large number of Brunner’s glands occur at the gastroduodenal junction of Noctilio leporinus. Also the glands of Brunner at the gastro-oesophageal junction in Mormoops are extremely abundant. This resembles the condition not only in the fish-eating Noctilio, but also in carnivorous species. Non-carnivorous bats generally possess relatively fewer and less well-developed Brunner’s glands (Forman, 1971b).

24.6.3 Gastric anatomy of the Thyropteridae Thyroptera discifera (Peters’ disc-winged bat) has a stomach that has a corpus gastricum and pars pylorica

336 

 V Laurasiatheria – 24 Chiroptera stomach” sensu Makanya et al. (2001). Only in the Thyropteridae (Fig. 5.22) there might be a storing room for the chitinous exoskeletons in the food in the prominent fornix gastricus.

24.7 Chiroptera, Phyllostomidae 24.7.1 General remarks

Fig. 5.22: Outlines of gastric forms in three families of Microchiroptera. Adapted from different authors.

of approximately the same diameter as the small intestine (Schultz, 1965), so that it is practically difficult to be differentiated from the gut (Fig. 5.22). The cranial blind end of the gastric tube it represented by a prominent fornix gastricus. Because of the strong development of the blindsac of the fornix Schultz (1965) speaks of a blindsacstomach (“Blindsackmagen”). Noctilionidae, Mormoopidae and Thyropteridae (Fig. 5.15) have stomachs that are simple (i.e. glandular) and unilocular. The pars pylorica is not remarkably lengthened and a vestibulum cardiacum is not differentiated. As all three families are insectivorous, they can “afford” to differentiate just a “carnivore

Phyllostomidae or New World leaf-nosed bats form some of the most speciose mammalian assemblages known (Rex et al., 2010), with 160 species in 55 genera (Simmons, 2005a). These authors differentiate the following seven subfamilies: Phyllostominae, Brachyphillinae, Phyllonycterinae, Glossophaginae, Carolliinae, Stenodermatinae, Desmodontinae. “No family of mammals has undergone a greater adaptive radiation than phyllostomatid bats.… From an ancestral diet of insectivory, descendants today can still be insectivorous”, but they are also consumers of vertebrate prey (carnivorous) or are nectarivorous and pollinivorous, frugivorous, and sanguinivorous (Freeman, 2000, page 317). Species of the phyllostomid subfamily Glossophaginae are nectarivores. “Much of the diversification within this group occurred by lengthening the snout or rostrum and reducing the size of the teeth on the palate in conjunction with evolution of a specialized nectarfeeding tongue” (Freeman, 2000, page 326). Phyllostomidae encompass a dietary spectrum that ranges across several trophic levels, and many show morphological specialisations for their dietary behaviour (Rex et al., 2010). These authors analysed the diet of 67 phyllostomid bat species from the Neotropics. Most species complemented their primary diet opportunistically with nutrients from many different food sources. Phyllostomid species may have specialised on distinct diets during their radiation without sacrificing their capability to exploit a variety of food types (Rex et al., 2010).

24.7.2 Feeding types within the Phyllostomidae Dietary evolution in phyllostomids is, as Wetterer et al. (2000) write, quite complex. The major dietary guilds (omnivory, insectivory, frugivory, pollinivory together with nectarivory, as well as sanguivory) evolved within Phyllostomidae. Reversals might confuse the situation, and some specialisations may have evolved more than once (e.g. carnivory). Bats of the family Phyllostomidae include species that specialise on insects, blood, small vertebrates, fruits, nectar, and pollen (Datzmann et al., 2010).



Wetterer et al. (2000) presented a phylogenetic tree, which in many places is unresolved, but based on detailed information on feeding types, that was originally published by Ferrarezzi and Gimenez (1996). Separate trees, dealing with insectivory, frugivory, nectarivory and carnivory are also given by Wetterer et al. (2000). Datzmann et al., 2010 published interesting phylogenetic trees under consideration of the feeding styles. As this source shows clearly resolved trees, Fig. 5.16 is based on that evolutionary tree. An excellent account of the possible differentiation of feeding styles in the Chiroptera has been given by Gillette (1975). The cosmopolitan suborder Microchiroptera contains insectivorous, herbivorous, piscivorous, carnivorous and blood-sapping forms. Their diversity of feeding habits started with initially insectivorous bats. Later it came to an evolutionary loss of insectivory and adoption of alternate strategies, such as feeding on fruit, flowers, pollen or nectar, on the one hand, or aquatic prey, such as fish, which was originally taken in what Gillette (1975) calls “food source duality”, namely insects together with other prey on water surfaces, or, as a nutritional extreme, bloodlapping, which might have started with another example of food source duality, namely ingestion of ectoparasites together with ingestion of vertebrate blood. In these situations, the shift away from ancestral insectivorous diet is connected with utilisation of dual food sources. Baker (1973) emphasises that there is no clearly recognisable and taxonomically significant boundary between fruit-eating and nectarivorous bats. This can possibly be generalised and should be kept in mind. There is considerable overlap of different types of material eaten by Phyllostomidae. For example, the Glossophaginae, represented by Anoura, Glossophaga and Leptonycteris, are characterised by Baker (1973) as insectivorous and nectarivorous. In Phyllostominae (Chrotopterus, Phyllostomus), insectivory, nectarivory and frugivory can be observed. “Omnivory” is a term which is sometimes applied, but does not give much information. The subfamilies Carollinae (Carollia) and Stenodermatinae (Centurio, Uroderma) are primarily frugivorous. The vampires (Desmodontidae) are sanguinivorous. Freeman (1995) assumes that nectarivory, frugivory and carnivory probably all evolved independently from an insectivorous microchiropteran ancestor. The ingestion of pollen and nectar by bats is closely related and cannot easily be differentiated. Pollen is taken both in the process of nectar-feeding as well as independently and can be the main source of protein in nectarivores (Freeman, 1995). Pollen is broken down chemically in bats by saliva and digestive juices. It is an important

24 Chiroptera, Phyllostomidae 

 337

nitrogen source in flower-visiting bats (Howell and Hodgkin, 1976), the carbohydrate fraction of the diet is represented by nectar. In the chiropteran stomach, little degradation of pollen grains occurs and this food is retained for a relatively short period in the stomach. Although the exine coat of pollen grains is highly resistant to degradation by digestive enzymes, pollen can be an important source of proteins, vitamins and minerals (Herrera and Martínez del Rio, 1998). Nectarivory and pollenivory cannot be clearly separated from each other and a strict separation of a feeding style does not make sense: According to Howell (1979) and Kelm et al. (2008), many “nectar-feeding” bats may be totally omnivorous and eat fruits and insects during times of flower scarcity. Temporary scarcity of nectar or increased protein demands (lactation) may sometimes require the uptake of alternative food items. On the other hand, it has to be taken into account that some Phyllostomidae are largely dependent on nectar resources, as the subfamily Glossophaginae. Gonzalez-Terrazas et al. (2012) mention that these nectarfeeders have an extremely high energetic requirement and consume large amounts of nectar every night. Tropical nectar-feeding bats of the subfamily Glossophaginae have one of the most specialised diets among mammals, as floral nectar constitutes a sugar-rich and highly digestible, but is also deficient in protein and fibre (Kelm et al., 2008). To conclude, sanguinivorous bats of the subfamily Desmodontinae live on a very special type of food: Only 44% of the (human) blood is represented by cellular material (Kunsch, 1997). Despite their highly specialised food the daily energy consumption in “blood-lapping” (Gillette, 1975) chiropterans is similar to that in frugivorous (Kovtun and Zhukova, 1994, speak of “herbivorous”) chiropterans.

24.7.3 Gastric anatomy of the Phyllostominae Outlines of stomachs of members of this phyllostomid subfamily have been compiled from different authors and can be found in Fig. 5.23. Gastric outlines from four genera are depicted and it is obvious that the shapes differ. Eisentraut (1950), who studied Phyllostomus hastatus, the greater spear-nosed bat, remarks that the gastric wall can be very extensible, so that the gastric shape can vary considerably according to the state of filling. This seems to apply in the illustration that was originally supplied by Eisentraut (1950). It is known that bats can ingest surprisingly large volume of food within short time. Contrary to the observations on food characteristics as published by Datzmann et al. (2010) and compiled in Fig. 5.16, Phyllostomus hastatus is – according to Eisentraut (1950) – a carnivore and its

338 

 V Laurasiatheria – 24 Chiroptera in descriptions of bat gastric anatomy, is extremely wide in the latter two species.

24.7.4 Gastric anatomy of the Carollinae

Fig. 5.23: Gastric outlines in Phyllostomidae: Phyllostominae and Carollinae. Adapted from different authors.

bag-like stomach does not show special differentiations. In Phyllostomidae, the gastric wall can be strongly dilated, so that the volume of the filled stomach widens the empty organ and is many times larger than that of the empty organ. Because of this it is difficult to determine exact volumes in preserved material. In another species of the genus, Phyllostomus discolor, the pale spear-nosed bat, the fornix gastricus is long (Forman, 1972, 1973) and slightly dilated at its apex (Fig. 5.23). The relatively short pars pylorica tapers abruptly and the vestibulum cardiacum is extremely short and not prominent. An incisura cardiaca is present, but shallow and only few mucosal folds can be differentiated. Macrotus californicus, the Californian leaf-nosed bat, has a remarkable gastric shape (Park and Hall, 1951), the stomach is pear-shaped and strongly constricted at the pylorus (Fig. 5.23). Micronycteris megalotis, little bigeared bat, and Chrotopterus auritus, woolly false vampire bat, both originally depicted by Forman (1973), show a unilocular gastric shape without cardiac vestibule, and a moderately developed pars pylorica. The transitional mucosal zone between the proper gastric gland zone and the pyloric mucosa, which is often intensively discussed

Outlines of stomachs of the genus Carollia are depicted together with the Phyllostominae in Fig. 5.23. Some remarks on the gastric anatomy are rather rudimentary. For example, Schultz (1965) writes that the stomach of Carollia perspicillata, Seba’s short-tailed bat, has a fornix gastricus and a pars pylorica. Robin (1881) goes into more detail when describing the stomach of Carollia brevicauda, the silky short-tailed bat: It is elongated and cylindrical, the oesophagus opens in the middle of the length of the stomach. The pars pylorica, which follows the large blindsac of the fornix gastricus, is narrowed and bent, so that the duodenum approaches the oesophagus. Park and Hall (1951) also mention that cardia and pylorus lie close together and they describe this organ of the frugivorous Carollia brevicauda as “triangular”. The tendency of gastric extension by lengthening and formation of voluminous gastric dilatations in the fornix gastricus can be found in frugivorous Phyllostomatidae, such as Carollia brevicauda (Eisentraut, 1950). Very clearly this can be seen in an illustration by Robin (1881) in the stomach of Corallia brevicauda where a vestibulum cardiacum is differentiated. Forman (1972) gives a detailed description of the stomach of Carollia perspicillata. It is club-shaped and decidedly asymmetrical with moderately well-developed vestibulum cardiacum, which enters the lesser curvature at an angle, thus creating a marked incisura cardiaca. The author describes the gastro-oesophageal junction as “abrupt”, but the tunica muscularis externa as continuous with tunica muscularis of oesophagus. The pars pylorica is tubular and markedly elongate. The fornix gastricus of Carollia sp. is saccular, appearing as dilated circular bulb which lies lateral to the cardiac vestibule. The fornix gastricus is never truly separated from the corpus gastricum by folds formed either by musculature or by gastric mucosa.

24.7.5 Gastric anatomy of the Stenodermatinae The highly frugivorous phyllostomid species Artibeus jamaicensis includes both insects and the liquid fraction of protein-rich leaves in their diet, at least seasonally, if not on a regular basis. It meets almost all of its protein requirements from plants (Herrera M. et al., 2001). Fruits are commonly considered to be nutritionally poor with most succulent fruits offering R; (b) Passage rate: E > R;

27 Perissodactyla, Rhinocerotidae 

 385

(c) Intake rate: E > R; (d) Nutrient extraction: E > R; (e) Rate of digestion: E < R. Using microbial fermentation for alloenzymatic digestion, as the host’s own digestive enzymes are not sufficient to make sufficient amounts of nutrients available, incubation of the digestive tract with microbes has to be accomplished. This also holds true in perissodactyla, which are large-intestine fermenters. Crowell-Davis and Caudel (1989) studied this process in equine foals: When given choice between maternal faeces and material from another mare, equine foals sniff on both materials equally, but practice coprophagy only of maternal faeces.

27.5 Rhinocerotidae The rhinocerotoids, sister taxon to the tapiroids within the infraorder Ceratomorpha, developed higher crowned teeth and adopted various strategies for dealing with more fibrous vegetation, including increased body size (Colbert and Schoch, 1998). In terms of diversity and longevity, the Rhinocerotidae represent a really successful group during the Oligocene radiation (Radinsky, 1966). The late Oligocene was the time when Rhinoceros unicornis, the Indian rhinoceros, separated from the two-horned Asian and African rhinos (Steiner and Ryder, 2011). Tougard et al. (2001) estimated that the paleontological emergence of the genus Dicerorhinus has to be dated in the Lower Miocene (between 23 and 16 Myr, Carroll, 1988). Today, there exist four rhinocerotid genera and five species (Grubb, 2005); they are treated in monographic style by Meister and Owen-Smith (1997): Ceratotherium simum; Van Strien (1997): Dicerorhinus sumatrensis; Adcock and Emslie (1997): Diceros bicornis; Schenkel (1997): Rhinoceros sondaicus; Laurie (1997): Rhinoceros unicornis. The monophyly of the five recent rhinoceros species (Tougard et al., 2001) is well established from morphological and paleontological data. Strong support is provided for the split between Diceros and Ceratothorium 17.1 ± 2.5 year before present; the divergence of Rhinoceros is dated about 11.7 ± 1.9 million years ago (Tougard et al., 2001). Although not a single rhinocerotid species can presently be found in North America, Rhinocerotidae from that continent represented one of the largest, longest, and most complete records of a mammalian family (Prothero, 2005) as well as one of the most successful groups of mammals in North America (Prothero, 1998). According to this author, the most striking thing about the pattern of rhinocerotid evolution is that of stasis: Most species appear suddenly and then are unchanged through most of

386 

 V Laurasiatheria – 27 Perissodactyla

their history. During the Miocene Rhinocerotidae occurred in enormous herds, especially in the High Plains of North America. Ecologically they were very diverse. “There were large hippolike grazers (Teleoceras, Brachypotherium, and Peraceras superciliosum); prehensile-lipped browsers; four independent examples of dwarfing (Peraceras hessei, Teleoceras meridianum, and still undescribed species of Teleoceras and Diceratherium); pig-sized herding rhinos (Menoceras arikarense); and many other less specialized kinds. Rhinocerotids occupied the large-bodied herbivorous niches in North America from the early Oligocene to the end of the Miocene” (page 595, Prothero, 1998). In their Old World evolution, rhinoceroses changed, as Kahlke and Lacombat (2008) describe it, from cursorily mixed feeders of central Asian origin to heavy, highly specialised grazers in the Plio-Pleistocene tundra of Central Europe. 27.5.1 Food of the Rhinocerotidae The reader should refer to Tab. 5.4. Ecologically, most Miocene (26 to 7 MYBP) rhinoceroses were brush or leaf eaters; later they adapted to hard and dry brush vegetation (Heissig, 1999b). Much later, between 44,000 and 24,000, rhinos eat considerable amounts of forbs, and Willerslev et al. (2014) show for some rhino species that they eat considerably more graminoids than horses. In the following, a short account of the food of rhinoceroses will be given in the systematic differentiation supplied by Groves (1983, 1997): There are five recent species of four genera in two tribes in the family Rhinocerotidae. In tribe Dicerotini, two species are grouped, namely, Diceros bicornis (black rhinoceros) and Ceratotherium simum (white rhino); in tribe Rhinocerotini, the genus Rhinoceros consists of two species (sondaicus, Javan rhino, and unicornis, Indian rhino), as well as of Dicerorhinus sumatrensis, the Sumatran rhino. The black rhino, Diceros bicornis, is predominantly a browser with woody dwarf shrubs, small trees and forbs providing the bulk of the diet (Hillman-Smith and Groves, 1994; Kaiser and Kahlke, 2005). According to Hall-Martin et al. (1982), this species selects against herbs, as well as against grass and sedges. Adcock and Emslie (1997) write that this species eats from small bushes, including Acacia species, but rarely eats plant parts that grow higher than 2 m. In a study comparing digestion of the black rhinoceros with the horse, Clauss et al. (2006) observed that these rhinos achieved only relatively low digestion coefficients when compared with the horse, a species of similar gastrointestinal morphology. The warning statement of Foose (1982) is corroborated that extrapolation from morphology to physiological details between species of different families should be avoided.

In contrast to the black rhino, a strict browser, the white rhino, Ceratotherium simum, is entirely graminivorous (Groves, 1972; Kaiser and Kahlke, 2005) or a “pure grazer” (Meister and Owen-Smith, 1997). The two handbooks of Mills and Hes (1997) and of Skinner and Chimimba (2005) on South African mammals make more detailed comparisons of both species possible. However, the stenophagous dependence of Ceratotherium simum on grass does not exclude seasonal variability of grassland usage, as has been described by Shrader and Perrin (2006). Schenkel (1997) characterises the food of the Javan rhinoceros, Rhinoceros sondaicus, as that of a true browser; young trees, bushes, shrubs and lianas represent the items that are eaten by that South East Asian species. R. sondaicus, is classified by Kaiser and Kahlke (2005) as browser. Schenkel (1997) mentions that the food of this species is composed of more than 100 plant species. Rhinoceros unicornis, Indian rhino, as described by Laurie (1997) is very diverse, but long grasses of the genus Saccharum represent the main food item. Additionally, the author observed 180 plant species as food items in Nepal. Food of R. unicornis, characterised by Laurie et al. (1983), consists of 70 to 89% of grass, but food composition shows seasonal changes. Sarma et al. (2012) indicates that Indian rhinos prefer wet grassland in all seasons of the year. R.  unicornis was also studied by Steinheim et al. (2005) in Nepal. These authors found an effective fermentation of cellulose in this species, it feeds on grass, which represents 63%, while browse contributes 28% to the food, this species has to be called a grazer; according to Laurie et al. (1983) grasses even make up between 70 and 89% of the diet in Nepal. It also eats fruits, leaves, branches sedges and ferns, aquatic plants and agricultural crops as additional material. Laurie et al. (1983) observed high seasonal variability. The Sumatran rhinoceros, Dicerorhinus sumatrensis, eats, as Hubback (1939) and van Strien (1997) remark, a diversity of plants, leaves and twigs, but also herbs and succulent leaves, but grass and sedges growing along streams are never eaten. According to Groves and Kurt (1972), the food of this species consists of fruit, leaves, twigs, bark and saplings (Kaiser, 2003); Dicerorhinus sumatrensis is a true browser (Kaiser and Kahlke, 2005). From the data presented by the literature and cited above, one can classify Rhinocerotidae into two feeding groups, two species that can be called grazers, namely Ceratotherium simum (white rhino, totally graminivorous), and Rhinoceros unicornis (Indian rhinoceros) with a high percentage of grass in its food. On the other hand, Diceros bicornis (Black rhino), Rhinoceros sondaicus (Javan rhinoceros) and Dicerorhinus sumatrensis (Sumatran rhino) are browsers.



27 Perissodactyla, Rhinocerotidae 

 387

27.5.2 Anatomy of the stomach of Rhinocerotidae Thomas (1801) describes the stomach of Rhinoceros unicornis, Indian rhino. In external appearance, it is very similar to the equine stomach, as is shown in illustrations that were originally published by Owen (1862) (Fig. 5.82). It is unilocular and composite, i.e. it is partly covered with squamous epithelium on a non-glandular tunica mucosa (Burne, 1905; Bhattacharya and Chakraborty, 1993). The mucosa in the fornix and proximal corpus is covered by a white, thick and non-glandular tunica mucosa, which is covered with a squamous epithelium with fine rugae (Fig. 5.82 B). A well-defined border separates this mucosal section from a much thicker pars glandularis. On the cardiac side of the pars pylorica the stomach has its smallest circumference. The total straight length of the stomach is 122 cm in a male and 81 cm in a female, but the length of the lesser curvature is only 53 cm (Owen, 1862). The stomach of the Sumatran rhinoceros, Dicerorhinus sumatrensis, is different from that of the Indian rhino, as Garrod (1873) writes. There is no constriction between cardiac and pyloric portion of the organ, but there is a peculiar diverticulum in the region of the fornix gastricus (Fig. 5.83), which is lined by non-glandular mucosa and covered by a squamous epithelium. Near the lesser

Fig. 5.82: External (A) and internal (B) aspects of the stomach of Rhinoceros unicornis. Adapted from Owen (1862).

Fig. 5.83: Opened stomach of the Sumatran rhinoceros, Dicerorhinus sumatrensis. In the area with question mark, cardiac glands are not clearly delimited. Adapted from Garrod (1873).

curvature Cave and Aumonier (1963) mention a “cobblestone” surface, which is cornified and papillated (Fig. 5.84). “The “cobblestone” area of the Dicerorhinus (the author writes “Didermocerus”) sumatrensis stomach must

Fig. 5.84: Mucosal lining of the stomach of Dicerorhinus sumatrensis. Adapted from Cave and Aumonier (1963).

388 

 V Laurasiatheria – 27 Perissodactyla

be regarded as a specialised prolongation of the cardiac mucosal field” (page 35). The authors speculate that “milling” of gastric contents occurs in this mucosal zone. The length of the above-mentioned diverticulum is 28 cm and its diameter is 13 cm. The mucosa in the fundus and parts of the corpus, as well as of the fornical diverticulum, is much plicated and looks white and opaque. The pars pylorica and the corpus gastricum are covered with a thick glandular mucosa. Between both types of mucosa – glandular and non-glandular – the bordering line is abrupt and of a type of the margo plicatus of the equine stomach. For two other rhinocerotid species a composite stomach is described. Clemens and Maloiy (1982) mention for the black rhino of Africa, Diceros bicornis, that the cranial one-half to twothirds of the stomach are lined with stratified squamous tissue. On the other hand, a smooth white squamous epithelium occupies about one-third of the total gastric area of Rhinoceros sondaicus, the Javan rhinoceros (Garrod, 1877). In the gastric zone lined with a non-glandular mucosa, Clemens and Maloiy (1983) found for the browsing black rhinoceros, Diceros bicornis, that the apparent digestibility of cell wall material, cellulose and hemicellulose is higher than in the glandular part of the organ (“caudal stomach”). On the other hand, energy uptake from the pars proventricularis (“cranial stomach”) is considerably lower than in the pars glandularis (“caudal stomach”).

continents in the early Eocene. Subsequently, separate families differentiated in North America, Europe, and Asia” (Carroll, 1988, page 530). The earliest species of the family Tapiridae, Prototapir, “appeared in the early Oligocene in Europe and the Middle Oligocene in North America” (page 531). Holanda and Ferrero (2013) stated that the genus Tapirus immigrated to South America and represents a lineage that has diversified in South America. Fluctuations in tapir distribution were observed by Holanda et al. (2012): In the Pleistocene Tapirus terrestris could be found in southern Brazil. The South American tapir species T.  terrestris and T.  pinchaque are closely related, but the tapir of Central America, T. bairdii, and the Asian species, T.  indicus, diverged from them earlier (Ashley et al., 1996). According to Holanda et al. (2011), the distribution of fossil tapirs was similar to the actual distribution.

27.6.1 Food of the Tapiridae

27.6 Tapiridae

In relation to tapir food, reference should be made to Tab. 5.4. The four well-established species of tapirs, Tapirus bairdii (Baird’s tapir), T. pinchaque (Mountain tapir), T. terrestris (South American or Brazilian or lowland tapir), T. indicus (Malaysian tapir), eat similar types of food: twigs, fruits and leaves (IUCN, 2011). They disperse seeds of the fruits they have eaten (Olmos, 1997). Tapirs and their kin might specialise in consuming a high amount of foliage with a moderate amount of fibre content (Schoch, 1989).

The tapiroid and the rhinocerotid lineage, together forming the suborder Ceratomorpha (McKenna and Bell, 1997), diverged, as Colbert and Schoch (1998) write, from a common ancestor in the early Eocene (~50 Mya.). The two superfamilies Tapiroidea and Rhinocertoidea are considered as sister-groups by Janis (1984). The recent family Tapiridae has only four established (Grubb, 2005a), as well as a recently described fifth species (Cozzuol et al., 2011, 2013), Tapirus kabomani, the little brown tapir or “tapir negrito” (Antelo Aguilar, 2014) from western Amazonia. Cerqueira (1982) presents information about the three, already wellestablished species from Ibero-America, represented by Tapirus pinchaque, the Andean form, which occupied the rising Andes during the Pliocene/Pleistocene; T.  terrestris lives in the lowlands east of the Andes; and T. bairdi might have differentiated from the latter. T. terrestris and T. bairdi seem to occupy similar habitats. Tapirus indicus, from Sumatra, Malaysia, Thailand and Myanmar (IUCN, 2011) prefers secondary lowland forest in flat and damp areas, as Novarino et al. (2005) describe. All species of the family belong to the genus Tapirus. Ancestors of tapirs “were common in all northern

27.6.1.1 Food of Tapirus bairdii The Central American or Baird’s tapir, Tapirus bairdii, is a completely herbivorous species, eating a wide variety of leaves, twigs, flowers and bark (Naranjo, 2001), as well as “fleshy fruits” (Eisenberg and Redford, 1999; García et al., 2006, Fig. 5.85). In the state of Quintana Roo, Mexico, the majority of plants eaten are herbs and bushes (Pérez Cortez and Matus Pérez, 2010). According to García et al. (2006), this species is able to eat approximately 15 kg of vegetation per day, consisting of leaves, stems and a small amount of fruits (Naranjo Piñera, 1995). Depending on the availability of food items, tapirs can shift their foraging strategy among habitat types and seasons (Naranjo, 2009), but also habitat differences between geographical regions can be responsible for variable food composition (Henry et al., 2000). The most noticeable changes in proportions of food items ingested by Baird’s tapir throughout the year are those related to fruit consumption. This type of food is related with seed dispersal, described by O’Farrill et al. (2006). However, according to Naranjo Piñera and Cruz Aldán (1998), the contribution of fruits



27 Perissodactyla, Tapiridae 

 389

highest. Even ferns were eaten by this species in considerable amount (Lizcano and Cavelier, 2004). In addition to being a foliage browser, Tapirus pinchaque, eats a broad array of plants, shoots, berries (Padilla et al., 2010). These authors write that T. pinchaque is a foliage browser, eating bromeliads, ferns, grass herbs, shrubs and trees. Practically, it does not eat fruits. In a comparison of data for T. pinchaque (Lizcano and Cavelier, 2004) and for T. bairdii (Naranjo-Piñera, 1995), the relative values for twigs, leaves and fruits in Fig. 5.85 are similar, but not identical. However, seed dispersal was also found in this species: Seeds from over 50 species still germinated in tapir dung (Eisenberg and Redford, 1999).

Fig. 5.85: Percentages of twigs, leaves and fruits in two American tapir species. Adapted from Naranjo (1995) and Lizcano and Cavalier (2004).

to the food of Tapirus bairdii is minimal; in all cases less than 10%. In Tapirus bairdii in montane cloud forest of Costa Rica, an analysis of faeces by Tobler (2002) showed that fibres were the largest component (40–55%) of the ingested food, followed by leaves (10–30%) and twigs (15%). Similar results from Chiapas, Mexico, have been published by Lira Torres et al. (2004). Bamboo was found in all samples and probably accounts for the high proportion of fibres. Twenty-seven plant species were identified in that study to be eaten by tapirs. This species takes only small amounts of a given plant species during one feeding bout. About 98 plant species of 50 families were counted as food items by Naranjo (2009), but, depending on the availability, tapirs can shift their foraging strategy. Nevertheless, Naranjo Piñera and Cruz Aldán (1998) state that the tapirs try to maintain a stable diet. In Chiapas, Mexico, the diet of Tapirus bairdii was composed by fibre (50.6%) leaves (45.5%), and fruit (3.9%) (Lira Torres et al., 2004). Fruits usually constitute a smaller proportion of Central American tapir food than leaves and other fibre sources. Despite these small proportions found in tapir faeces, it is very likely that fruits provide important amounts of calories consumed by this ungulate. Baird’s tapir is an important disperser of seeds, a fact documented by O’Farrill et al. (2006).

27.6.1.2 Food of Tapirus pinchaque In Equador, the mountain tapir, Tapirus pinchaque lives in moist forests (Stummer, 1971). According to Eisenberg and Redford (1999), the mountain tapir feeds on at least 264 species of vascular plants, species of Lupinus ranked

27.6.1.3 Food of Tapirus terrestris The lowland tapir is a browser and frugivore (Eisenberg and Redford (1999), but it also grazes; its bulk food generally consists of green shoots and also includes fruits, leaves, twigs, grasses, aquatic and agricultural plants (Padilla and Dowler, 1994). It disperses plant seeds, but not in a very high degree (Galetti et al., 2001). It browses selectively, eating 88 out of 256 available plant species in southern Venezuela (Salas and Fuller, 1996). Padilla and Dowler (1994) call T.  terrestris “generalized browsers and grazers” (page 4), which is, at least in the Argentinian province Misiones, strictly nocturnal all year round (Cruz et al., 2014). Data supplied by Richard and Juliá (2000) on the same species are more detailed: 45% of the food they eat are derived from trees, 13% from bushes, 19% from herbage, 6% cm herbaceous vines and as much as 16% from Gramineae. Salas and Fuller (1996) clearly showed that Tapirus terrestris is a selector, eating just 88 out of 256 available plants. Browsing on terminal buds, the animals ingest a more nutritive and less fibrous food, which might possibly contain toxic substances. By selectively eating only small amounts of material from many different plant species, tapirs might use several different detoxification pathways and eventually ingest larger amounts of food (Hibert et al., 2011). In their study, Galetti et al. (2001) came to the conclusion that Tapirus terrestris is a frugivorous species, but Talamoni and Assis (2009) write that this species is mainly browsing on leaves and stems, but in the dry season a higher percentage seeds was eaten. According to Chalukian et al. (2013), lowland tapirs switch their diet from frugivory to herbivory when fruits are scarce. T. terrestris of the Amazon basin consumes the more abundant low-quality forage because this gives them the advantage of reducing search efforts. Diets of the lowland tapir were examined by Bodmer (1990) in northeastern Peru to investigate the relationship among high-quality fruit, lower-quality browse and searching behaviour.

390 

 V Laurasiatheria – 27 Perissodactyla

Lowland tapir consumed an average of 33% fruit, which is relatively high for a large non-ruminant ungulate. Bodmer (1990) states that lowland tapirs can consume greater proportions of fruit than other large non-ruminant ungulates; they exploit a nutritious fruits that occur in large patches and that meets the energy demands of their large body size. Investigations that were also undertaken by Bodmer (1991), demonstrated that Tapirus terrestris eats fruit, leaves and fibrous material at about equal levels (i.e. about a third of total food ingestion); animal material was not taken. The lowland tapir always eats much fibrous material, in the peak fruiting season it amounts to about 40%, but in the low fruiting period up to 80% fibrous food (Henry et al., 2000) is ingested. 27.6.1.4 Food of Tapirus indicus Very little information from the literature is available on the food of the fourth species, Tapirus indicus, the Malay tapir. It is generally classified as browser, eating 115 species of tree leaves, shrubs and wines, herbs and mosses, but only five species are included in its most preferred food (Mohamed and Traeholt, 2010). 27.6.1.5 Summing up the information on tapir food To conclude the presentation of average feeds of the four recent tapir species, the following statements are, most probably, premature and only more detailed ecological studies, done under comparative aspects, may bring light into the presently obscure situation: From the remarks in the literature one obtains the impression that the Central American tapir, T.  bairdii, is the species that ingests the highest amount of fibrous material, approximately half of total food (50.6% according to Lira Torres et al., 2004), but in the South American lowland tapir, T. terrestris, fibrous material amounts to only about a third of the total ingested food (Bodmer, 1991). The other two species seem to be intermediate, the mountain tapir (T. pinchaque) eats a “broad array” of plant material (Padilla et al., 2010) and the Malay tapir (T.  indicus) is characterised as browser (Mohamed and Traeholt, 2010). Modern functional investigations on the digestive process in tapirs have just recently started to be published. Two examples can be given: Güiris et al. (2007) isolated bacteria from faeces of Tapirus bairdii; however, on the nutritional significance of these microorganisms little is said. Clauss et al. (2009e) studied the absorption of minerals in tapir species, as compared to the horse. For example, absorptive efficiency for Ca, which is higher in tapirs than in horses, might have evolved to ensure that high dietary Ca concentrations do not lead to the binding of dietary P in the intestine, making it unavailable for hindgut microbes.

27.6.2 Anatomy of the stomach of Tapiridae Literature on the gastric anatomy in tapirs is very limited and in most cases from the 19th century. “Tapirs are the only group among the perissodactyls for which no recent description of the gastrointestinal tract (GIT) exists” (Hagen et al., 2015, page 171). The description of the situation in Tapirus terrestris (the author writes T. americanus), given by Owen (1868), has been almost verbally cited by the description of the South American tapir that was given by Padilla and Dowler (1994). Owen (1868) writes on page 458: In the Tapir a “thick epithelium is continued for the extent of 3 inches (=7.7 cm) to the left of the cardia, and for that of 7 inches (=18.0 cm) to the right, toward the pylorus: the rest of the stomach has a compact villous surface with a few narrow well-defined rugae: the gastromucous membrane increases in thickness, through lengthening of the gastric tubules, as it nears the pylorus. The stomach of the Sumatran Tapir presents a similar disposition and proportion of the cuticular lining”. On the other hand, Beddard (1889) mentions for the same species, Tapirus terrestris, that the oesophageal epithelium extends over the cardia into the stomach in an area with a width between 3.8 and 6.4 cm. The author mentions that the squamous nonglandular area is wider in the Rhinoceros and the horse. For the Malayan tapir, T. indicus, Newton Parker (1882) published an illustration of the stomach (Fig. 5.86). It is obvious that this species has a unilocular stomach, but a fold separates corpus gastricum from pars pylorica on the lesser curvature. According to Newton Parker (1882), the length along the greater curvature is 61 cm. The oesophageal non-glandular lining is different from the situation described for T. terrestris by Owen (1868) and Padilla and Dowler (1994). In T.  indicus, it extends into the stomach for only about 2.5 cm and does not line the total fundus gastricus and proximal corpus gastricum, where this

Fig. 5.86: Opened stomach of Tapirus indicus. Adapted from Newton Parker (1882).



27 Perissodactyla, Equidae – recent horses, donkeys and asses 

epithelium extends much further into the stomach. Newton Parker (1882) writes: “The epithelial lining of the oesophagus extends into the stomach for about an inch (= 2.54 cm) all round from the cardia. In this, it differs from T.  americanus (= T.  terrestris), in which the oesophageal epithelium extends much further over the interior of the stomach more like the arrangement in the rhinoceros and horse. The greater part of the mucous membrane is very smooth, but for a region extending round the cardiac portion of the greater curvature it is considerably ridged. There are also a few slight ridges in the pyloric end. The muscular coat thickens considerably towards the pylorus; and there is a well-marked circular pyloric valve” (page 770). There are only extremely limited remarks on the digestive physiology in Tapiridae. Lang-Deuerling (2008) cites data by Foose (1982) on mean retention time. It seems to be shorter in tapirs (36–45 hours) than in rhinos (51–66 hours).

27.7 The small intestine of Perissodactyla Detailed studies on the anatomy of the perissodactyl small intestine have only been published for the domestic horse and are compiled in textbooks of veterinary anatomy (Zietzschmann et al., 1943; Nickel et al., 1967, 1973; Getty, 1975). At the pylorus of the horse, Equus caballus, the small intestine with an average length of 22 m and a capacity of 40 to 50 L (Getty, 1975), begins. The first duodenal section is the pars cranialis duodeni and lies on the right side of the median plane of the abdomen. First, it forms a widened section, the ampulla duodeni. The ductus choledochus and the ductus pancreaticus, as well as a ductus pancreaticus accessories, enter the gut about 12.5 to 15 cm from the pylorus (Getty, 1975). The duodenum then approaches the facies visceralis of the liver (Nickel et al., 1967, 1973). “The length of the small intestine of the horse varies from animal to animal, but it also varies within the same animal in relation to digestive activity. The duodenum has an average length of 1 m, the jejunum 25 m, and the ileum about 0.7 m. The variation in length is associated with great variation in the lumen of the small intestine, which may have a diameter of up to 5–7 cm or may be almost obliterated by contraction of the intestinal wall” (Nickel et al., 1973, page 185). The duodenum terminates at the lesser curvature of the caecum (Getty, 1975). With the exception of the duodenum the small intestine of the horse has an extensive and movable mesentery (Zietzschmann et al., 1943). “The duodenum has a relatively fixed position, being attached to adjacent organs by the short hepatoduodenal ligament and its continuation, the mesoduodenum” (Nickel et al., 1973, page 186).

 391

A flexura cranialis or flexura prima represents the beginning of the duodenum descendens. The duodenum descendens proceeds caudodorsally to the flexura caudalis or flexura secunda. The convexity of the duodenum descendens shows right (Getty, 1975). Starting at the flexura caudalis, the pars ascendens duodeni passes from right to left and runs cranially and ends at the ostium ileocaecocolicum, as Nickel et al. (1967, 1973) write. The jejunum of Equus caballus has a broad mesojejunum, is movable and represents the longest section of the small intestine. Approximately 1 m before it opens into the capud caeci, the ileal wall becomes thicker. A ligamentum ileocaecale connects ileum and caecum, which lie parallel and very close to each other (Zietzschmann et al., 1943). In the whole small intestinal wall of the horse, lymphonoduli solitarii can be found, but Zietzschmann et al. (1943) mention that lymphonoduli aggregati or Peyer’s plaques, can only be found close to the antimesenterial side; the largest of these can be found close to the end of the ileum. The mucosal glands are of tubuloalveolar type and in the tela submucosa Brunner’s glands or glandulae duodenales are differentiated (Getty, 1975). The arterial supply of the small intestine is accomplished via branches of the A. mesenterica cranialis and the A. coeliaca. Venous drainage transports blood back to the heart via the V. portae. (Zietzschmann et al., 1943)

27.8 Caecum of the Equidae – recent horses, donkeys and asses In Fig. 5.87, the caecum of the domestic horse is shown, which is adapted from an originally published illustration by Zietzschmann (1925). According to Ellenberger (1879), the caecum has a conical body and a dilatation at its base, which lies dorsally in the abdomen and which he calls “saccus ventriculiformis” or “saccus ventriculosus” because of its “stomach-like shape” (“magenähnliche Gestalt”). The same structure is called “caput caeci” by Nickel et al. (1967, 1973). The dorsal dilatation of the equine caecum can also be called “basis caeci” (Schaller, 1992) or “ampulla caecalis coli” according to Kostanecki (1926). The caput caeci is formed, as Krüger (1929) mentions, out of the colon ascendens. However, the isthmus at the ostium caecocolicum lies between the caput caeci and the colon ascendens. The apex caeci lies medially on the sternum. The body and apex lie deeper (i.e. ventral) in the abdomen then the caput caeci, which lies close to the lumbal vertebral column (Campbell et al., 1984). Krüger (1929) states that this “primitive” dorsoventral position of the caecum is in contrast to the situation in most other mammals and in man.

392 

 V Laurasiatheria – 27 Perissodactyla

Fig. 5.87: Large intestine of the horse. Scientific terminology after Schaller (1992). Modified after: Zietzschmann (1925) (A) and Argenzio et al. (1974) (B).

There are four taeniae on the caecal body, a dorsal and a ventral one, as well as a lateral and a medial one. The length of the caecum is approximately 72 to 95 cm long (Ellenberger, 1879), when the taeniae are cut, the length of the caecum lies between 100 and 150 cm. At the ostium caecocolicum circular muscle bundles form a sphincter-like narrowing. The mucosa in the region forms a semicircular fold, which is mobile and not permanent. Meyer (1980) mentions that the opening between the caecum and the colon is relatively small; it might contribute to the separation of fine from coarse particles. Clauss et al. (2008a) describe and depict the isthmus between the caput caeci and the colon ascendens for Equus burchelli, Burchell’s zebra. They even propose a correction of an illustration supplied by Stevens and Hume (1995), which does not show this isthmus-like constriction. As the caecum and ileum have the same direction, the ileum meets the large intestine perpendicularly, as is depicted in one of the illustrations that were originally published by Kostanecki (1926) (Fig. 5.88). It protrudes as a small cone, the papilla ilealis, into the capud caeci, which, according to Schumann (1907), is reminiscent in shape of the portio vaginalis uteri. The arrangement of the muscle layers at the equine ileocaecal junction has already been illustrated by Kotzé (1988) (Fig. 5.89). The mucosa around the

ileal opening forms many transverse and longitudinal folds. The same author also writes that the tunica muscularis is strongly thickened – sphincter-like – close to the opening. Not only the large intestine of the domestic horse has been depicted. For example, an illustration of the gastrointestinal tract of Burchell’s zebra, Equus burchellii (Fig. 5.90), has been published by Mitchell (1905, 1916), but this illustration is rather schematic.

Fig. 5.88: Medial aspect of the equine caecum. Adapted from Kostanecki (1926) and terminology according to Schaller (1992).



27 Perissodactyla, Equidae – recent horses, donkeys and asses 

Fig. 5.89: Opening of the ileum onto the caecum of the horse. Adapted from Kotzé (198). a) Ileum, Tunica muscularis, Stratum circulare, b) Ileum, Tunica muscularis, Stratum longitudinale, c) Papilla ilei, Tunica muscularis, Stratum longitudinale, d) Taenia dorsalis caeci, e) Caecum, Tunica muscularis, Stratum circulare.

In the horse, the corpus caeci has longitudinal muscular bands on the dorsal, ventral, lateral, and medial surfaces. The taenia caeci ventralis is hidden in Fig. 5.88, which was originally published by Kostanecki (1926). The taeniae have been described by Nickel et al. (1967, 1973): between these four muscular bands lie four rows of sacculations or “haustrations”, as well as deep grooves that form the “semilunar folds”, separating neighbouring haustrations. The dorsal and medial taeniae end at the

 393

apex caeci; the taenia caeci ventralis usually joins the taenia caeci medialis near the apex or “fades out”. The caecal vessels and lymph nodes of the equine caecum accompany the medial and lateral taeniae, as was depicted by Sieber (1903) (Fig. 5.91). Dart et al. (1991) write that the major vascular supply to the caecal apex appears to be through the medial caecal artery (ramus caecalis medialis). Both the lateral and medial caecal arteries give rise to arterial meshworks (retia) around the respective veins (Sellers et al., 1982, compare with Fig. 5.92 for the colon after Snyder et al., 1989). Vessels from these retia supply the caecal tissue and lymph nodes and continue to form an extensive submucosal plexus. This plexus supplies both the mucosa, and the tunica muscularis and serosa. Vessels within the longitudinal and circular muscle layers of the muscularis externa run parallel to the muscle fibres and consequently, perpendicular to each other. “Drainage was facilitated by more sparsely distributed venules that united with venules from adjacent areas and descended to the submucosal veins.” (Dart et al., 1991, page 1545).

27.8.1 Functional remarks on the equine caecum The seminal paper by Janis (1976) focussed on the comparison between forestomach fermenters, especially ruminants, on one side, and hindgut fermenters, on the other. This latter group has to be further differentiated. Not only gastric regions or the caecum can represent fermentation sites in large mammalian herbivores, but the proximal colon is often the principal lumen where microbial fermentation takes place. For example, Bergner and Ketz (1969) clearly differentiate between caecal digestion and colonic digestion. According to Stevens and Hume (1995), the caecum “is the principal site of microbial fermentation

Fig. 5.90: Small and large intestines in three species of the Perissodactyla. Adapted from Mitchell (1905).

394 

 V Laurasiatheria – 27 Perissodactyla

Fig. 5.91: Vascular supply of the wall of the equine colon ascendens. Adapted from Snyder et al. (1989).

Fig. 5.92: Arterial supply of the equine caecum. Adapted from Sieber (1903).

in most small herbivores” (page 227). Species belonging to the genus Equus are certainly not “small herbivores” and this makes a functional evaluation of their caecum even more interesting. In addition, “knowledge about digestive functions in the domestic horse will be extrapolated to characterize these processes in large nonruminant ungulates” (Foose, 1982, page 40) and a comparative evaluation of the caecum as part of the large intestine is worthwhile. Much information about the large intestine represents the situation after death of the animal. However, one has to keep in mind that the large intestinal volume can vary considerably during lifetime (Meyer, 1980). For example,

after feeding bouts in horses, the volume of the caecum can be three to four times as voluminous as that before food ingestion. Because of this functional variability, measurements of caecal capacity – and colon capacity – are given from different sources of the literature: Zietzschmann et al. (1943) mentions a volume of the caecum of 51 L and of the total colon of 118 L. Bergner and Ketz (1969) give average volumes for the caecum (33 L) and for the colon (90 to 100 L). According to Nickel et al. (1967, 1973), the volume of the caecum ranges from 16 to 68 L (mean 33 L). The volume of the colon ascendens alone already lies between 55 and 130 L (mean 80 L). In “Shetland-type ponies”, Simmons and Ford (1990) determined the capacities of the caecum and large colon after death as 7.0 ± 0.8 and 17.7 ± 3.7 L, respectively. Putman (1986) investigated grazing of large herbivores in temperate ecosystems of the New Forest in Southern England. Ponies showed pronounced adaptation to changing trophic situations. When the favoured grasslands were exhausted, they drifted towards exploitation of other forages during autumn and winter when food became scarce. The large intestine of the horse enables, if necessary, to cope with alternative feedstuffs even if of low digestibility. This means that the caecum, which is set off from the direct connection between influx via the ileum and the efflux via rectum and anus, has to contribute to this adaptation to varying trophic situations. Horses as hindgut fermenters “can exploit extremes of the fibrosity gradient better than any of the ruminants can” (Foose, 1982, page 50). Equines are able to live under different trophic conditions, a fact that reflects the width of their ecological spectrum (Kaiser, 2003). Mean home range sizes for equine subpopulations can differ considerably, as Smuts (1975) described for Equus burchellii. In the caecum of the horse, the decomposition of cellulose with the help of bacteria takes place. According to Kern et al. (1973), 19.7% of different types of equine caecal bacteria were proteolytic; cellulolytic bacteria numbers per gram ingesta were similar in the caecum of the horse and in the rumen of steers. In addition, protozoa transforms the proteins of plant and bacterial origin into protozoal protein, which is easier to digest than the original proteinaceous material (Bonhomme-Florentin, 1974); the caecum of the horse can be inoculated with protozoa by ingestion of fresh faeces. The same author (BonhommeFlorentin, 1985) also showed that the ciliates of the equine caecum do not only attach to plant fragments, but also to bacteria, which provide higher-quality microbial protein. Moore and Dehority (1993) draw a differentiated picture: Removal of protozoa from the hindgut of the pony (defaunation) resulted in only a slight decrease (P < 0.1) in the overall digestibility of dry matter and had



27 Perissodactyla, Equidae – recent horses, donkeys and asses 

no effect on cellulose digestion. Thus, it seems that protozoa are not essential for fermentation of ingesta in the caecum and colon of the equine and bacteria and fungi simply take over this role in cellulose digestion when the protozoa are removed. The caecum acts as a reservoir for the partially digested feeds (Howell and Cupps, 1950) and its activity is coordinated to move digesta on to the large intestine. Although the contraction length at both ends of the caecum is practically identical (7.6 to 7.8 seconds), it is to be expected that the frequency of contractions at the apex caeci is higher than at the basis caeci. 49.7 ± 10.3 contractions per hour can be found at the apex, but only 36.9 ± 6.8 at the caecal base (Howell and Cupps, 1950). A long stasis of digesta that reached the blind apex is avoided by more frequent contractions. Ross et al. (1990) describe motility patterns in the horse caecum, either running from the caecal apex to its base or vice versa. There are also contractions of the “progressive pattern”, beginning at the caecal apex, conducted through the caecal base and caecocolic orifice and into the right ventral colon. Argenzio, Lowe et al. (1974) remark that the equine caecum does not show prolonged retention of fluid markers and does not have the ability to selectively reject particulate markers. Both fluid and particulate markers rapidly leave the caecum and pass into the large colon segments where markers are retained. During most of the period between meals the caecum absorbs large quantities of water. Ruckebusch et al. (1981) give a detailed description of the motility of the caecum of the horse. It contracts locally at the level of the haustrations, i.e. between the taeniae, and also produces powerful propagating contractions. These latter movements either refill or empty this part of the gut. The contractions are propagated with a speed of 12 cm/sec in the direction from base to apex and 10 cm/sec from apex to base. The comparative examination by Ruckebusch et al. (1981) of the variations of caecal volume and caecal motility and caecocolic transit shows that small variations of the volume correspond with considerable increase of the motility. In donkeys, Equus asinus, Lechner-Doll et al. (1992) observed an increase of cellulose in the food from 30% during the rainy season to approximately 40% in the dry season. Reduced digestibility of food is compensated by increased food uptake and increased voiding of faeces. In the dry season, donkeys excrete significantly more faeces (60% more in relation to body mass) than in the rainy season. Thus, Equus asinus compensates for a lower-quality diet by increasing its intake rate (Izraely et al., 1989a). The same authors write: “By lowering its requirement for metabolic energy, the donkey is able to balance its energy metabolism under the adverse nutritional conditions, but only if sufficient forage is available” (Izraely et al., 1989b, page 97).

 395

The contractions of haustrations in the equine caecum, as well as powerful propagating contractions, as described by Ruckebusch et al. (1981), have already been mentioned above. Langer and Takács (2004) discussed the effects of anatomical differences between those types of digestive tracts where the longitudinal musculature is reduced to small bands, the taeniae, and those where such differentiations are not found. They were able to show that tubes with two or three or, at most, four taeniae reduce their internal volume more efficiently than all others, namely, with a relatively small reduction of the surface area and contraction of the tunica muscularis at a relatively low rate. Folds formed with relatively little contraction of the musculature, i.e. little change in the surface area, represent an effective means of retention and thus of flow regulation. A caecum with taeniae and semilunar folds is an adaptation for this effective type of regulation of digesta transit. In the area close to the apex caeci of the horse, only two taeniae are differentiated: the taenia caeci dorsalis and medialis end at the apex caeci; the taenia caeci ventralis usually joins the taenia caeci medialis near the apex or “fades out” (Nickel et al., 1967, 1973). Hence, the equine caecum is an area of the digestive tract that seems to be able to regulate digesta transit. Slow transport through a gut section makes microbial fermentation and absorption of fermentation products possible. Engelhardt et al. (1983) demonstrated for the pony a higher concentration of VFA in the caecum than in the rectum or in faeces; absorption from the caecum is higher (5.9 L/day/0.75 kg) than in the proximal colon (2.0 L/day/0.75 kg) and the distal colon (0.4 L/day/0.75 kg). The investigations of Argenzio, Southworth et al. (1974) come to a different conclusion. According to them, the mucosa of the caecum and colon transport products of microbial fermentation, VFAs, at substantial and approximately equivalent rates. In both sections of the large intestine, regulation of digesta transit through the caecum of the horse is advantageous because it supplies time for epithelial transport. Morphologically, Wille and Zahner (1997) describe another adaptation to absorption: Immediately under the mucosal surface of the caecum and the proximal colon the arterial side opens via capillaries into submucosal veins, this differentiation is connected with the transmucosal transport (absorption). In five-sixths of the horses and in donkeys lymphatic follicles, forming plaques, were accumulated at the apex of the caecum (May, 1906). This can be interpreted as an adaptation to increased antigen concentration in the blind end area of the caecum. Although absorption of microbial products has been proven, material that should be preserved leaves the digestive tract via rectum and anus. When faeces are eaten again (coprophagy), some easy to digest material is saved. Another important function of coprophagy is inoculation

396 

 V Laurasiatheria – 27 Perissodactyla

of the tract with bacteria, fungi and protozoa. Coprophagy represents a sensitive period for equine foals to learn about food plants (Marinier and Alexander, 1995) during the first 6 weeks of life when the rate of coprophagy is highest. Although Lyons et al. (1996) had repeatedly observed a stallconfined orphan horse foal eating its own faeces, it does not seem that coprophagy is satisfying hunger per se and the donor of the faces is almost always the mother, as has also been shown by Crowell-Davis and Caudle (1989).

27.8.2 General remarks on the caecum of the Tapiridae Compared with the Equidae data for the Tapiridae are much less available and gaps of knowledge are still obvious. Tapirs are the most conservative of living perissodactyls (Carroll, 1988); Heissig (1999a) even speaks of the oldest family of perissodactyls with extant representatives, but since the beginning of the Miocene only few genera and species survived. However, Hooker (2005) lists the first Equidae in the early Eocene and the earliest Tapiridae somewhat later in the middle Eocene. Rose (2006) differentiates this picture a little bit more: According to him, Hyracotherium (=Eohippus) lived during the North American and European early Eocene and Heptodon was the first definitive tapiroid from the early Eocene of North America. The oldest true tapir, Protapirus, lived during the middle Eocene in North America. Olmos (1997) assumes that tapirs have already been eating leaves, twigs, and fruits from the world’s forests perhaps for a very long time. It has already been mentioned above that five tapir species live to the present day as a relic group of tropical forest browsers (Palmqvist et al., 2008; Cozzuol et al., 2013). One of the five surviving species, Tapirus pinchaque, the mountain tapir of Columbia, Equador and northernmost Peru, has been characterised by Padilla et al. (2010) as a rare mammal, which is considered the smallest and least specialised of the four species of Tapirus. Palaeontologists made interesting discoveries on Arctic Ellesmere Island (Eberle, 2005): A fossil tapir, Thuliadanta mayri, has been found in early Eocene remains. The animal must have been able to withstand the dark arctic night because the present-day island was already north of the Arctic Circle at the Eocene. On the other hand, the fossil-bearing strata of Ellesmere Island imply a lush environment with high biomass production that was probably adequate to maintain the fossil tapir over the winter months. The warning statement of Foose (1982) should be repeated that knowledge about digestive functions in the domestic horse is often extrapolated to characterise digestive processes in large nonruminant ungulates. For example, Clauss, Wilkins et al. (2009) investigated the feeding regime

in tapirs kept in zoos of the United Kingdom. They eat highly digestible feeds, such as commercial produce and pellets, as they are normally given to horses, which produced unnaturally soft faeces. The authors promote increased roughage intake in captive populations to attempt to reduce obesity and to produce more solid “normal” dung.

27.8.2.1 Anatomy of the caecum of the Tapiridae Detailed descriptions of the caecal anatomy of tapir species are not available. Lang-Deuerling (2008) compiled measurements of the sections, including the caecum, of the gastrointestinal tract of Tapirus terrestris and T. indicus from older sources of the literature. For Tapirus pinchaque, the smallest and least specialised tapir, a foliage browser, Padilla et al. (2010) characterise the caecum as short and with a large lumen and Home (1821a) mentions that the caecum of the Malayan tapir, Tapirus indicus, short. There is a description and illustration of the caecum of the lowland tapir, Tapirus terrestris, published by Beddard (1889). The illustration can be found in Fig. 5.93 A. The ileum in this species is attached to the caecum by a large mesenterial fold, which is free of blood vessels and is identified in the illustration as plica ileocaecalis. This structure has also been depicted by Huntington

Fig. 5.93: Four pictures of tapir caeca. Adapted from: Beddard (1889) (A); Huntington (1903) (B); Mitchell (1905) (C); Clauss, Zürich, personal communication (D).



27 Perissodactyla, Equidae – recent horses, donkeys and asses 

Fig. 5.94: Photo of the caecum of Tapirus terrestris. Adapted from Clauss, Zürich, personal communication.

(1903) (panel B). According to Beddard (1889), there are four taeniae running “from end to end” of the caecum; they have approximately equal distance from each other. One taenia, the taenia caeci dorsalis, is covered by the ileocaecal mesentery. The caecum of the lowland tapir has also been depicted by Mitchell (1905) (Fig. 5.93 C), but this illustration is not very informative. On the other hand, it shows the capud caeci in the transition zone between the caecum and the colon ascendens. This dilatation cannot be clearly identified in the photo, which was supplied to the present author by Clauss (personal communication) and is therefore marked with question marks in panel D. In a photo of the caecum of a specimen of Tapirus terrestris, the South American tapir (Fig. 5.94), that was also supplied by Prof. Clauss, the capud caeci is clearly delineated. The following colon ascendens starts with a tight constriction before the extended diameter of that colonic section follows. Another interesting aspect is the apex caeci, which forms a conical diverticle. According to Prof. Clauss, this structure has a lumen. Two taeniae can be seen that end on the apex caeci; other taeniae are not visible on the photograph. With the help of the scale given on the original photograph, the approximate volume of the caecum amounted to about 26 L in this specimen of Tapirus terrestris.

27.8.3 General remarks on the caecum of Rhinocerotidae Tapiridae became separated from Rhinoceritidae 47 MYBP. The Asian rhinoceroses can be distinguished from the African species since about 29 MYBP and the two genera, Ceratiotherium, the white rhino, and Diceros, black rhino,

 397

of Africa separated from each other 17 MYBP (Tougard et al., 2001). During the course of the Pliocene and Pleistocene, about 2.5 MYBP, fossil woolly rhinoceroses of the genus Coelodonta evolved, as Kahlke and Lacombat (2008) described, from central Asian mixed feeders to highly specialised grazers inhabiting huge belts of tundra-steppe environments during cool to cold periods. According to these authors, woolly rhinoceroses were characteristic elements of the progressively evolving Middle and Late Pleistocene cold-adapted mammalian faunas of Eurasia. After that they died out. The recent rhinocerotid species are also under severe survival stress. For example, the northern form of the black rhinoceros, Diceros bicornis, survives only in very low numbers, or even became extinct, in central Africa (Emslie and Adcock, 1997). When the present lines were written, Nov. 10, 2011, Deutschlandfunk and BBC News informed their listeners that the western subspecies (longipes) of the black rhino had been declared extinct. To characterise the frightening situation of rhinoceros conservation in Indonesia, Foose (1982) writes with “brutal directness” on page 132 of his impressive thesis: “It will be interesting to determine more about the diets of the other two browsing rhinos in the wild, the Javan and the Sumatran, if they survive long enough for us to investigate them.”

27.8.4 Anatomy of the caecum of the Rhinocerotidae From different references illustrations showing the outer shapes of the caeca in four species of Rhinocerotidae are compiled in Fig. 5.95. The small arrows indicate the net flow of digesta from the ileum into the caecum and from there into the ascending and following sections of the colon. It is remarkable that a caecal diverticle in the black rhinoceros, Diceros bicornis, a browser, could also be found during a dissection done by M. Clauss et al. (personal communication) (Fig. 5.96). This structure could not be found in the caecum of the grazing white rhino Ceratotherium simum. However, in none of the rhinoceros caeca, depicted in Fig. 5.95, a caecal diverticle was discernible. Whether such a diverticle could represent a “safe haven” for microbes, is not known. It is also not clear whether the prominence of haustra, as can be seen in C.  simum (Fig. 5.96), is related with an increased uptake of grass, as compared with a much “smoother” external appearance of the colon of the browsing Diceros bicornis. In Fig. 5.95 B and D, two different mesenterial folds (plicae) are identified, and in panel C, the caecocolic transit zone is hidden by the colon descendens, which has been depicted in the original woodcut published by Garrod (1873). In the three depicted browsing species,

398 

 V Laurasiatheria – 27 Perissodactyla

Fig. 5.95: Drawings of caeca of four rhinocerotid species. Pi: Plica ileocaecalis; Pc: Plica caecocolica. In panel C a section of the colon descendens (marked with two stars) covers the caecocolic region. From different authors.

Fig. 5.96: Caecum and colon of two species of Rhinocerotidae with different feeding styles. Sections of the large intestine: C: caecum; 1: colon ventrale; 2: flexura pelvina; 3: colon dorsale; 4: colon traversum; 5: colon descendens. Adapted from photos by Prof. M Clauss, Zürich, personal communication.

the caecal shape can be described as “piriform”, a term applied by Garrod (1873) in the Sumatran rhino (Dicerorhinus sumatrensis), a browsing species. The caecum of the grazing Indian rhino, Rhinoceros unicornis, the caecum looks relatively simple – especially in panel E – although Mitchell (1905) himself, the original author of this illustration, characterizes this caecum as “short, but very wide and capacious” (page 478). It will become obvious in the following lines that detailed information on comparative caecal anatomy in rhinoceroses is very limited. Some basic measurements are available for all five species. De Bouveignes (1953), fide Hillman-Smith and Groves (1994) and Castell (2005) published some information on the black rhino, Diceros bicornis. Data on the Javan rhinoceros, Rhinoceros sondaicus, were supplied by Garrod (1877) and Beddard and Treves (1890). In the third browsing rhino, Dicerorhinus sumatrensis (Sumatran rhino) the situation is only slightly better than in the previous species, data originate from Home (1821), Garrod (1873) and Groves and Kurt (1972). For the grazing white rhino, Ceratotherium simum, relatively recent, though rather superficial, information was made available by Endo et al. (1999) and Kiefer (2002) supplied information in a remarkable comparative and experimental study. Finally, anatomical data on the caecum of the Indian rhino, Rhinoceros unicornis, originate from “historical” references (Thomas, 1801; Owen, 1862, 1868; Mitchell, 1905). In Ceratotherium simum (grazer), the caecum was investigated by Endo et al. (1999). According to these authors, the ostium ileocaecale was encircled by a welldeveloped muscle, the M. sphincter ilei. This means that there is a constricture between the small and large intestines that regulates ileocaecal digesta transport. Dorsal of and near to the ostium ileocaecale the opening to the colon, the ostium caecocolicum, was observed. No comments on any structures regulating caecocolonic digesta flow was mentioned by Endo et al. (1999). It has already been mentioned before (Langer and Takács, 2004) that taeniae, together with the semilunar folds that are “anchored” on the muscular bands, are an adaptations to the need of intermittent regulation of digesta flow. Because of this it is of interest to identify and characterise taeniae as differentiations of the wall of the digestive tract. For the browsing Javan rhino, Rhinoceros sondaicus Beddard and Treves (1890) comment on the taeniae: Three of them can be found on the caecum of this species. One taenia begins at the ileocaecal junction and runs to the apex, a second one starts at the capud caeci and runs under the fixation of the mesentery. The third band starts from the superior of the two taeniae and proceeds to the colon ascendens, the outgoing

27 Perissodactyla 

limb of the colic loop. All three taeniae meet at the apex caeci. On the piriform caecum of the browsing Dicerorhinus sumatrensis Garrod (1873) found three taeniae. This author also makes a remark on the intraabdominal position of the caecum. According to him, it is median in position, and the colon lies orad of it. The limited information and incomplete descriptions is made obvious when papers of prominent historical anatomical studies are screened for information. For example, from the excellent illustration of the caecum of Rhinoceros unicornis, a browser, which was published by Owen (1862), it does not become clear how many taeniae are differentiated on caecum and colon (Fig. 5.95; in the text only two are mentioned). Functional investigations by Clauss et al. (2005) were able to distinguish between browsers and grazers: “Black rhinoceroses, which are large hindgut-fermenting browsers, retained ingesta for a shorter period relative to their body size than…grazing rhinoceros species” (page 367). Comparing the two African rhinoceros species, Steuer et al. (2010) were able to demonstrate the grazing white rhino, Ceratotherium simum, retain particles selectively, which corresponds with the higher digestion of fibre as can be found in the browsing black rhino, Diceros bicornis. It should be kept in mind that a completely different mechanism might also contribute to higher values for digestibility; they are probably the result of intensive selection during food uptake (Clemens and Maloiy, 1983). The importance of anatomical differentiations is still unclear in connection with many functional aspects. For example, Clauss et al. (2007a) indicate in a comparative study dealing with the black rhino (Diceros bicornis) and the domestic horse (Equus caballus), that differences in mineral absorption occur even between species of similar digestive anatomy. The observed difference in faecal particle size between black and white rhinos might be due to the considerable difference in body weight, as Steuer et al. (2010) assume. These authors do not mention possible morphological variations within the digestive tract, such as differences between caecal morphology. It is disappointing for the anatomist, that only very generalised, and therefore imprecise, information is given, even when more specific information could be obtained. For example, in a protozoological study on the ciliates of the white rhino, Ceratotherium simum, Ito et al. (2006) mention the “large intestine” as habitat of protozoan species; a differentiation between the caecum and the colon is not made. In both sections of the large intestine the problem arises how this section is incubated by microbes. In a remark not dealing with an identified rhino species, Meister (1997) mentions that calves eat maternal faeces to incubate their digestive tract with fermentatively active microbes.

 399

27.9 Topography and morphology of the colon in Perissodactyla Only the anatomy of the digestive tract of the domestic horse, Equus caballus, has been described in great detail, whereas publications on tapirs and rhinoceroses deal with this subject just in passing, presenting only very rudimentary information. However, the illustrations published by Mitchell (1905) for Equus burchellii (Burchell’s zebra, E. granti in Mitchell, 1905), Tapirus terrestris (South American tapir, T.  americanus in Mitchell, 1905) and Rhinoceros unicornis (Indian rhinoceros) present very similar structures so that it can be hoped that Equidae, Tapiridae and Rhinocerotidae have similar large intestines (Fig. 5.90). In a description of the species Equus burchellii, Grubb (1981) states that by analogy with horses, zebras show a similar digestive physiology. Comparable “extrapolations” are also made for the other two perissodactyl families (Janis, 1976). Because of lack of detailed descriptions, the following will be based on the textbooks by Zitzschmann (1925), Nickel et al. (1973) and Schaller (1992), dealing with the domestic horse. The colon of the horse is characterised by its extraordinary capacity. As can be seen in Tab. 5.2, which contains information from Nickel et al. (1967, 1973) and Schaller (1992), it is a complex section of the gastrointestinal tract. The ascending colon has an average capacity of about 90 L (range 55–130 L, Nickel et al., 1973). In Shetlandtype ponies, as they were investigated by Simmons and Ford (1990), the mean volumes of the ventral and dorsal section of the ascending colon were practically identical, representing about 36% of the contents of the total “fermentation vat” consisting of the caecum and colon ascendens. The colon is a long, U-shaped loop consisting of two parallel limbs and a terminal flexure (Fig. 5.87). The beginning and the end of the ascending colon are attached to the mesenteric root by the mesocolon ascendens, which passes distally between the two limbs of the loop and connects them. The lines of fixation of the ascending mesocolon to both “branches” of the colonic loop are marked in the illustration by broken lines. Bertone et al. (1987), mentions that the connecting tissue and blood vessels that can be found between dorsal and ventral right colons. This long U-shaped loop is folded once on itself, forming a double horseshoe-shaped loop. The colon ventrale dextrum (Fig. 5.87 A) begins at the caecocolic orifice, which opens into the basis caeci as a narrow and muscular structure, the collum coli, which is at first directed dorsocaudally. This is followed by a dilated part of the colon ventrale dextrum, which lies in the “lesser curvature” of the caecum. The colon ventrale dextrum is connected with the caecum by means of the caecocolic fold. The transversely directed flexura

400 

 V Laurasiatheria – 27 Perissodactyla

diaphragmatica ventralis, also called sternal flexure (Schaller, 1992), lies in the xiphoid region of the floor of the abdominal cavity and continues on the left side as the left ventral colon. The colon ventrale sinistrum passes caudally and slightly dorsally following the curvature of the abdominal floor and ends in front of the pelvic inlet at the flexura pelvina. From the colon dorsale sinistrum it passes cranially along the left abdominal wall and lies dorsal and slightly lateral to the left ventral colon. Upon reaching the diaphragm, it forms the flexura diaphragmatica dorsalis, which lies dorsal to the flexura diaphragmatica ventralis and is in contact with the diaphragm, the liver, and the stomach. The colon dorsale dextrum passes caudodorsally on the right and makes considerable contact with the diaphragm and with the right lobe of the liver, remaining within the intrathoracic part of the abdominal cavity (Nickel et al., 1973). It has a widened diameter and has been described as a “stomachlike dilatation” of the colon. On the right side of the root of the mesentery, the colon dorsale dextrum decreases sharply in diameter, turns medially, and is continued by the narrow transverse colon, which is short, and – as in the other species – is located just cranial to the cranial mesenteric artery. It is connected with the radix mesenterii by a short mesocolon transversum (not depicted in Fig. 5.87). The colon descendens, which begins to the left of the A. mesenterica cranialis and the radix mesenterii, in which this vessel lies, has a smaller diameter than the ascending colon and is therefore also known as the small colon or colon tenue. When compared to the descending colon of the other domestic mammals, it is long, reaching a length of 2.5–4 m. It is suspended by the descending mesocolon (not depicted), which allows it to form large loops. Most of the small colon occupies, together with the jejunum, the space caudal to the stomach and dorsal to the left parts of the great colon, extending also into the pelvic cavity (Nickel et al., 1973). These authors also present measurements of the diameter of different sections of the equine colon. The flow of ingesta may be impeded by dilatations and reductions of the diameter of the colon, which can result at the following colonic sections: At the narrow caecocolic orifice and the initial collum coli, at the flexura pelvina and at the funnel-shaped narrowing where the ascending colon joins the colon transversum. Except for the constricted part at the collum coli, the diameter of the right and left ventral colon is very wide, measuring about 25–30 cm. A short distance proximal to the flexura pelvina, the diameter of the colon narrows to about 6–10 cm and retains this width to about the middle of the colon dorsale sinistrum. From here onward the diameter increases steadily and reaches its maximum of 30–50 cm at the colon dorsale dextrum. The

zones of dilatation of the colon ascendens have a “reservoir function” in addition to the caecum (Sellers et al., 1982) and according to Moore and Dehority (1993) the total and the cellulolytic fungal concentrations were even approximately 10-fold higher in the colon than in the caecum. The diameter of the equine colon decreases considerably, for example, at the beginning of the transverse colon. The diameter of the descending colon is about 7–10 cm. and remains the same throughout its length. In a diagrammatic drawing that was highly simplified for the purposes of a physiological study, Argenzio, Lowe et al. (1974) indicate two of the functionally important zones of reduction of cross-sectional diameter in the colon of the horse (Fig. 5.87 B). They make no indication of the collum coli, nor do they clearly separate the colon transversum from the colon descendens, which they call “small colon”. The following rectum is not depicted by these authors. The descending colon is followed by the rectum, which in the horse is 20–30 cm long. First, it is suspended by the mesorectum, but in the retroperitoneal part of the pelvic cavity, it is completely surrounded by connective tissue and forms a capacious and muscular ampulla recti. According to Schumann (1907), three narrowed zones of the equine colon the stratum circulare of the muscular coat is thicker than general, namely, at the transit zone from caecum to colon (the collum coli), at the flexura pelvina and aborad of the end of the colon ascendens, i.e. in the short colon transversum. It is deeply frustrating that the only real comparative study that deals with all three recent families of odd-toed ungulates is more than 100 years old. Mitchell (1905) depicted and briefly described the digestive tract of Equus burchellii, Tapirus terrestris and Rhinoceros unicornis. These illustrations are compiled in modified forms in Fig. 5.90. In all three species (and in all species of the respective families?), a hairpin-shaped course of the colon ascendens can be seen. In addition, the colon transversum, which lies orad of the A.  mesenterica cranialis, is very short. In none of the species, the following colon descendens does not form such a hairpin-shaped course. Compared with the small intestine the colon of perissodactyls is not a very long structure. This also becomes obvious in a comparison presented in Fig. 5.97, which shows the relative lengths of small intestine, caecum and colon, but also the functionally more relevant volume – determined by Nickel et al., 1973; Getty (1975) for the domestic horse and for the black rhino (Diceros bicornis) by Clemens and Maloiy (1982). In this triangular diagram, it can be seen that relative lengths and relative contents (volume) show a very different distribution. The length of the small intestine is relatively large in both species, but concerning volume the colon clearly predominates. The colon of odd-toed ungulates represents a storing

27 Perissodactyla 

Fig. 5.97: Relative lengths (open symbols) and relative contents (closed triangles) in three sections of the gut of the horse (Equus caballus, Ec), the American tapir (Tapirus terrsetris, Tt) and the black rhino (Diceros bicornis, Db). Adapted for Ec from Nickel et al. (1973) and Getty (1975); for Tt from Padilla and Dowler (1994) and for Db from Clemens and Maloiy (1982).

volume, obviously of greater importance than the caecum, whereas the small intestine is a region of intensive contact between the digesta and he intestinal wall with effective secretion of digestive enzymes and absorption of nutrients. It is remarkable that the relative lengths of

 401

small intestine, caecum and colon is surprisingly similar in all three perissodactyla families, the colon represents approximately 20% of the whole post-gastric gut, the small intestine has a relative length of about 70% and the caecum a length that attributes about 10% to the gut. Many odd-toed ungulates have a colon that is 4 m long, or even longer (columns in Fig. 5.98), only in Tapirus terrestris was a shorter colon mentioned by Padilla and Dowler (1994) and by Owen (1868). However, there seems to be a difference in the length of the colon relative to CR length (diamonds in the illustration), as taken from Grzimek (1987). The domestic horse has a colon that is about three times as long as its CR length (diamonds). (Meyer, 1980, speaks of a volume of 2 to 9 L/100 kg live mass.) The Indian tapir (T.  indicus) has a colon that is about three times as long as the CR length, but in the American tapir (T.  terrestris), the total colon is only slightly longer that the CR length. Because of their “stretched” body form, the colon of the Rhinocerotidae is short relative to CR length (generally 40%). The pyloric mucosa lies along the lesser curvature of the corpus, as well as in the “funnel” of the pars pylorica (Maxwell and Stewart, 1995). In sucking piglets, Cranwell (1995) determined the surface areas of different mucosal lining as percentage of total internal surface: pars oesophagea: 5.6%; pars cardiaca: 30.1%; proper gastric gland area: 44.4%; pars pylorica: 20.0%. Mucus and proteases were produced in all three glandular sections of the porcine stomach, lipase was produced in piglets by the cardiac gland zone, hydrochloric acid, somatostatin, serotonin and histamine by the proper gastric glands and somatostatin and gastrin by the pyloric gland zone (Cranwell, 1995). The three glandular mucosal types, as well as the squamous non-glandular mucosa, absorb and transport volatile fatty acids at substantial rates (Argenzio and Southworth, 1974); the highest rate of absorption is determined in the cardiac gland area, followed by the stratified squamous zone. In pigs, an anaerobic

microflora has been found attached to this squamous non-glandular mucosal surface (Maxwell and Stewart, 1995). Endocrine cells in pig stomach have been studied by Capella and Solcia (1972). The functional interpretation of the different types of cells remains obscure in many cases – a conclusive picture cannot be drawn. In a detailed study, Schmidt (1939) deals with the lamina muscularis mucosae, which adapts to different organ widths by change of muscular position. According to Schmidt (1939), it is incorrect to speak of circular and longitudinal fibres in the lamina muscularis mucosae. The musculature is perforated by arteriae and venae in the middle of these areae to regulate blood circulation in the mucosa. On the other hand, the surface of the tunica mucosa is subdivided into different sections, so-called areae gastricae, which are separated from each other by shallow grooves.

28.3.3.2 Gastric mesenteries (Babirusa follows separately) In the representatives of the Suidae (Sus scrofa, Phacochoerus aethiopicus, Potamochoerus porcus), the mesogastrium ventrale follows the lesser curvature and forms the lesser omentum (Langer, 1973, Fig. 5.122). The dorsal mesogastrium is very similarly orientated in Sus, Phacochoerus, and Potamochoerus. The mesogastrium dorsale starts near the oesophagus and forms the gastrophrenic ligament that fixes the stomach to the diaphragm. From there the line of fixation of the dorsal mesogastrium runs dorsal of the ventricular diverticulum and passes over to the greater curvature. Along the greater curvature the mesogastrium dorsale can be followed down to the pylorus and the proximal few

Fig. 5.122: Lines of fixation of both mesogastria to the porcine stomach. Adapted from Langer (1988).



28 Cetartiodactyla, Artiodactyls, Suidae 

 437

centimetres of the duodenum where it forms the mesoduodenum. This course of the line of fixation of the dorsal mesogastrium means that it characterises a long stretch of the greater curvature. Many authors write that gastric rotation takes place and shifts the line of fixation of the dorsal mesogastrium to the stomach. However, Andersen (1929) does not believe that gastric rotation is responsible for the relation between both the omenta and the stomach. Originally the peritoneal duplicature differentiating in the greater omentum is fixed to the stomach primordium of Sus scrofa in a broad zone. According to Andersen (1929), the peritoneal cavity forms a slit-like excavation or diverticle of the peritoneal cavity. This peritoneal excavation lies between the anlage of the stomach and the anlage of the greater omentum. It deepens towards the left by active growth of the peritoneal epithelium. The cavity widens towards the left of the gastric primordium. The ventrocaudal part of the omentum extends by growth and thus forms the pouch-like extension of the omentum majus.

28.3.3.3 Architecture of the tunica muscularis (Babirusa follows separately) The three muscular layers that are characteristic for most mammalian stomachs (Langer, 1974a, c, 1988), have already been discussed by Forssell (1912, 1913), Torgersen (1942, 1968), Pernkopf (1930), G. Müller (1962), Farthmann (1973). Despite the small diverticulum ventriculi, the stomach of Sus scrofa has to be called a unilocular organ. The wall of the stomach is able to dilate or contract to be able to adapt to variable quantities of food. It should be generally advantageous to cover the whole content of the organ with a double muscle layer, even when the stomach is maximally filled. For this purpose the tunica muscularis is formed by at least two muscular layers that form a net-like wall (Pernkopf, 1930; Langer 1973; Farthmann, 1973). The highly schematic illustration in Fig. 5.123 does not always show this clearly. In many cases, both layers cross each other almost perpendicularly (Langer, 1973). In the fornix gastricus, all three layers of the tunica muscularis contribute to the formation of the muscular wall, but only the longitudinal and the fornical oblique fibres are indicated in Fig. 5.123 A. In the fornix, the circular and the longitudinal layers may deviate considerably from their “regular” course. In close relation to the oblique fibres, a ventricular groove is formed at the lesser curvature (Fig. 5.121 B) (Bauer, 1923; Black, 1970; Black and Sharkey, 1970). The superficial layer can be followed from the external muscular sheet of the oesophagus along both gastric sides, the facies visceralis and parietalis. Fibres can be

Fig. 5.123: Muscular architecture of the stomach of Sus, seen from the facies parietalis (A) and visceralis (B). Adapted from Langer (1988).

seen between the cardia and the tip of the diverticulum ventriculi (Weissflog, 1903; Zietschmann et al., 1943), from where they can be followed along the greater curvature down to the pylorus and duodenum (Langer, 1973). On the facies parietalis and visceralis of the stomach, the external longitudinal muscular layer is considerably reduced in density, so that these gastric regions are partly free from musculature of the superficial layer (Weissflog, 1903; Zietschmann et al., 1943; Schwarze and Schröder, 1962; Langer, 1973). In the pyloric region of the pig, this layer is not formed along the lesser curvature (Weissflog, 1903; Zietschmann et al., 1943/1974; Langer, 1973). Weissflog (1903) assumes that this lack of longitudinal musculature on the lesser curvature of the pars pylorica might be related to the very thick internal muscular sheet that forms the torus pyloricus. It seems appropriate to consider the circular and internal oblique layers together because they are functionally and anatomically closely related. The deep layer can be differentiated into two sheets, the internal oblique and circular fibres that replace each other in different regions

438 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

of the stomach. In Sus, Phacochoerus and Potamochoerus, a U-shaped loop of internal oblique fibres “rides” on the cardiac incisure (Langer, 1973) and forms the two muscular lips of the sulcus ventriculi (Fig. 5.121 B). This arrangement of the oblique fibres is called the cardiac loop or ansa cardiaca (NAV, 1983) and is to be found in many mammals (Bauer, 1923; Pernkopf, 1930; Langer, 1974a, c). The smooth muscles of the ansa cardiaca run parallel to the lesser curvature down to the border between the corpus and the pars pylorica and then as the lower segmental loop (Torgersen, 1942, 1968, “untere Segmentschlinge” of Forssell, 1912, 1913) towards the greater curvature of where they cross to the opposite side to join the corresponding loop there. The fibres that lie in this loop partly branch out in the direction of the greater curvature and are called fibrae obliquae corporis by Langer (1973, 1974a, c). This is the region where internal oblique fibres and circular musculature cannot be differentiated from each other. Only on the lesser curvature circular fibres can clearly be discerned (Bauer, 1923), running perpendicular to the longitudinal axis of the stomach (Fig. 5.121 B). The deep muscular layer of the diverticulum ventriculi in the pig and its relatives has a complicated architecture. Weissflog (1903) and Pernkopf (1930) as well as Zietschmann et al. (1943/74) give detailed descriptions and illustrations of the spiral fold that runs in counterclockwise direction from the fornix into the diverticulum (Fig. 5.121 A). The internal fibres run obliquely over the spiral into the diverticulum. Pernkopf (1930) states that this is the spiral muscular architecture of the blindsac. In the pyloric region, on the lesser curvature of the stomach of Sus and its relatives, a pyloric torus is formed by circular muscles

(Weissflog, 1903; Zietschmann et al., 1943/1974; Langer, 1973). The circular fibres converge towards the pyloric torus. Opposite to the torus on the greater curvature lies a semicircular sphincteric muscle that is in continuity with the fibres forming the muscular cushion of the torus pyloricus. Thus, a functional unit that may close the stomach is formed.

28.3.3.4 Blood vessels of the porcine stomach (Babirusa follows separately) In the domestic pig, detailed investigations on the vascular architecture of the gastric wall have been supplied by Martin (1922), Zietschmann et al. (1943), Schwarze and Schröder (1964), Engelmann (1971), Trixl (1973), Chatelain (1973), NAV (1983), Nickel et al. (1976, 1981), Zamora and Reddy (1981). Although the stomach of the pig is unilocular and therefore externally similar to that in man, the arterial supply of the organ is different. In Sus scrofa, most of the above investigators found only two branches of the A. coeliaca (NAV, 1983) or truncus coeliacus (NA, 1989); the two branches are the A. linealis and the A. hepatica (NAV, 1983) (Fig. 5.124). The Aa.  gastrica sin., diverticuli, and gastroepiploica sin. are all branches of the A. lienalis. An additional diverticular artery, a branch of the A. gastroepiploica sin., supplies the diverticulum. It can be generalised from the above observations that the supply of the ventricular diverticulum is very variable. The A.  gastrica dextra and A. hepatica propria are branches of the A. hepatica (communis). A branch of the A.  hepatica supplies the lesser curvature on the right-hand side via the A. gastrica dextra and on the greater curvature via the A.  gastroepiploica

Fig. 5.124: Arterial supply of the porcine stomach, seen from the facies visceralis. Adapted from Langer (1973).



28 Cetartiodactyla, Artiodactyls, Suidae 

 439

or gastroomentalis, which in Sus is itself a branch of the A.  gastroduodenalis. A.  gastrica sinistra seems to be an artery that shows variable origin. It is not only a branch of the A. lienalis, but Schwarze and Schröder (1964) showed that it can have its origin from the A.  hepatica. Another variety of the branching system, presented by Schwarze and Schröder (1964), consists of three arteries given off by the coeliac artery: in addition to the Aa.  lienalis and A.  hepatica, the A.  gastrica caudalis can branch directly from the A. coeliaca. The venous drainage of the stomach of the pig is discussed in Hapke (1957), as well as in the textbooks of veterinary anatomy that were mentioned above. 28.3.3.5 Ontogenetic development of the porcine stomach In his book, Langer (1988) presents a graph (Fig. 5.125) that shows data originally collected by Kvasnitskii, which were published by Kidder and Manners (1978). From the time of birth body weight increases (A), but the volume of the stomach, related to body weight (BW), increases up to as much as about 10 L/100 kg BW from birth to approximately 70 days (B). From there on relative gastric volume gradually decreases again during the first year of life. Afterwards, there is a slight increase. The peak value around 70 days of postnatal life lies within the natural weaning period in Sus scrofa. Pigs are naturally weaned between 60 and 90 days (Haltenorth, 1963), between 75 and 105 days according to Heptner and Naumov (1966) or between 42 and 70 days (Kolb, 1971, 1974). The high relative volume of the stomach of the pig of about 70 days postnatal age (Fig. 5.125), expressed in litres per 100 kg BW, could be the result of the stimulating effect of particulate food matter. For unit body weight storing volume for ingesta in the stomach should increase when the quality of food decreases. This is the case when intake of high-quality milk decreases and feeding of lower-quality particulate food increases. The considerable increase in body weight after weaning more than compensates this “overshooting” effect and the relative gastric volume decreases. In livestock handling, domestic pig weaning is intensively manipulated. Artificial weaning of domestic animals might already occur around 42 days after parturition (Helfferich and Gütte, 1972; Park, 2006). Meyer and Kamphues (1990) even mention that the earliest possible period after which piglets can be artificially weaned lies in the second postnatal week. This is possible because the wild piglet already begins to dig in the ground and to take up small quantities of non-milk food from an age of 14 to 21 days onward (Heptner and Naumov, 1966; Linderoth, 2005) or 21 to 28 days (Kolb, 1974).

Fig. 5.125: Body weight (A) and gastric volume in relation to body weight (B) in pigs of increasing postnatal age. Adapted from Kidder and Manners (1978).

Acidity of gastric juice approached that found in adult pigs when pigs reached 56 to 70 days of age (Manners, 1976). There is a very long period during which the contents of the stomach of the young pig are insufficiently acid to inhibit bacterial multiplication and activation of pepsinogen. Because of the secretory activity of parietal cells in adults, pH falls rapidly to the region of pH 2, at least in the outer layers of gastric contents (Kidder and Manners, 1978). In these mature pigs, the high acidity of the contents of the stomach will prevent rapid bacterial multiplication. It was already stated above, that the diverticulum ventriculi on the fundus gastricus is a special differentiation of the porcine stomach. This structure begins to be differentiated before birth. Krüger (1929) indicated that a porcine diverticular blindsac is differentiated prenatally near the cranial end of the stomach anlage in an embryo of about 1.5 cm CR length (CR length). Andersen (1929) was able to show that an embryo with a CR length of 1.6 cm does not only have a fundus anlage, but also a primordial diverticulum. The diverticle acquires a spherical shape when the pig embryo has a CR length of 2.4 cm; it is bent caudally and increases in length with further growth: In embryos with CR length of 2.8 cm the diverticulum is 0.35 mm long,

440 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

at a CR length of 3.5 cm the diverticulum was recorded to be 0.75 mm long. There is a growth tendency of the diverticulum in dorsal direction (2.8 cm CR length) and later also in caudal direction (3.5 cm CR length). However, in the period between 3.5 and 17.5 cm CR length, the relative size of the diverticle even decreases (Andersen, 1929). 28.3.3.6 Short functional remarks on the porcine stomach It can be expected that functional processes in the unilocular and diverticulated stomach are practically identical to conditions in other unilocular stomachs. However, investigations on the digestion of the pig, which is the only species of the Suidae in which digestive physiology has been intensively studied, showed that carbohydrate fermentation in the pars oesophagea of the stomach is limited (Kidder and Manners, 1978). It has already been mentioned above that an anaerobic microflora has been found attached to the squamous non-glandular epithelial surface of pigs (Maxwell and Stewart, 1995). In this species, bacterial fermentation – which is lower than in the ruminoreticulum of ruminants, but higher than in the stomachs of rabbit, horse, and rat (Kidder and Manners, 1978) – could be found in the “uppermost layer of the stomach”. It can be speculated from the descriptions of the above-cited authors that this “uppermost layer” might be the fornix ventriculi. It is still unclear whether the diverticulum of the porcine stomach is of any functional significance. In the domestic pig, the region of the fornix gastricus is partly lined with cardiac mucosa (Fig. 5.120), which produces alkaline mucins. Because of a pH of about 7.8 of the cardiac secretion (Höller, 1970), microbial activity in the fornical region is possible to a limited extent, but in the proper gastric gland region a highly acid secretion is produced, together with proteolytic enzymes. Investigations of gastric microbial fermentation in Suidae, also including species other than Sus scrofa, are needed. The limited information that is available, is still ambiguous, often leaving open questions. For example, in the stomach of Phacochoerus aethiopicus, collected in different regions of South Africa, Booyse and Dehority (2012) found a wide range of pH values between 3.25 and 6.06. Another unresolved question has been presented by Jensen et al. (1997). They found gastric lipase in the cardiac mucosa of the porcine stomach, but the significance of gastric lipase for the digestion of fat in pigs remains to be elucidated. Warthogs digest fiber more efficiently than other wild suids (Clauss et al., 2008b). In a comparative study, these authors deal, in addition to the wild boar (Sus scrofa), with warthogs (Phacochoerus sp.), red river

hogs (Potamochoerus porcus), Visayan warty pigs (Sus cebifrons), babirusa (Babyrousa babyrussa), and peccaries (Tayassuidae). Most of these species appear to share a similar relationship of protein digestion with domestic pigs, regardless of differences in digestive anatomy.

28.3.4 Small intestine of Suidae According to Nickel et al. (1967, 1973), the pars cranialis of the duodenum in the pig, Sus scrofa, starts at the right side of the abdomen at the pylorus and ascends caudodorsally along the facies visceralis of the liver and forms an S-shaped loop (ansa sigmoidea, Schaller, 1992), ending on the flexura cranialis or prima. Two to five centimetres distal to the pylorus the ductus choledochus opens into the duodenum. The ductus pancreaticus accessories of the porcine pancreas (Nickel et al., 1973; Crevier, 1993) enters the duodenum 12 to 20 cm distal to the papilla of the ductus choledochus. From here the duodenum descendens passes caudally to the right kidney and is suspended by a broad (6 to 8 cm) mesoduodenum. The pars transversa duodeni is short and crosses the caudal edge of the mesentery. On the flexura caudalis or secunda the ascending part of the duodenum crosses to the left of the median plane, proceeds craniodorsally to a sharp bend to the right, the flexura duodenojejunalis. The pars ascendens duodeni is connected with the colon descendens by a plica duodeno-colica (Crevier, 1993). 28.3.4.1 Glands and villi in the small intestine of Suidae Tuch and Amtsberg (1973) differentiated types of villi intestinales in the pig, Sus scrofa. The villi can be found in the duodenum on longitudinal folds. Their flattened surface lies transverse to the longitudinal axis of the duodenum. Smith (1988) states that absorption, especially of proteins, appears to be possible only in enterocytes that were produced prenatally. Based on investigations of corrosion casts, Bellamy et al. (1973) demonstrated blood circuit around Lieberkühn’s crypts or glandulae intestinales.

28.3.5 Topography and morphology of the colon in Suidae Compared with other herbivorous ungulates (Fig. 5.126), the porcine large intestine – colon plus caecum – is relatively voluminous (Horst, 1956; Bayley, 1978). In the large intestine, the microbes can be found in different sites. The anatomy of the colon of the Suidae is well documented and the following descriptions of these even-hooved



28 Cetartiodactyla, Artiodactyls, Suidae 

 441

Fig. 5.126: Relative volumes of stomach and small and large intestines in domestic ungulate species. Raw data are from Horst (1956) and Bayley (1978).

mammals will be based on the publications of veterinary anatomists, especially Nickel et al. (1973). Arising from the caecum, which is slightly wider than the colon ascendens, this latter section in the pig is greatly elongated and characteristically rolled up on itself, forming a conical mass of spiralling coils, the ansa spiralis (Schaller, 1992), which is suspended by the ascending mesocolon and has the general shape of a beehive (Engelmann, 1971; Koch and Berg, 1990; Crevier, 1993) (Fig. 5.127). The ansa spiralis coli is formed by 3 1/2 centripetal turns, which bring it to the apex of the cone. The centripetal turns form the outside of the cone, are relatively wide, compared with the following 4 ½ centrifugal coils and are characterised by two taeniae, as well as two rows of haustra (Koch and Berg, 1990). At the apex or flexura centralis (Schaller, 1992), the ascending colon reverses direction and returns in a counterclockwise direction as centrifugal coils inside the centripetal turns to the base of the cone (Fig. 5.127). The centrifugal turns have no taeniae (Nickel et al., 1973). Engelmann (1971) mentions that there might be variations in the numbers of coils, namely 3 centripetal and 4 centrifugal coils, but the present author doubts, whether this difference is of any morphological or functional importance. After emerging from the base of the cone the last of the centrifugal turns passes cranially and forms the transverse colon, which passes from right to left cranial to the root of the mesentery. The descending colon, suspended by a short, fat-filled descending mesocolon, lies close to the median plane and passes in a straight line to the pelvic inlet. The rectum is embedded in fat. Before it ends

Fig. 5.127: Schematic representation of the ansa spiralis coli of the pig with 3½ centripetal and 4½ centrifugal coils. Adapted from Engelmann (1971).

at the anal canal it widens to form a distinct ampulla recti. (Nickel et al., 1973). 28.3.5.1 Blood vessels of the colon in Suidae Blood vessels that supply or drain the colon have been described in detail for pig by textbooks of veterinary anatomy, such as Zietzschmann et al. (1943/1974), Schwarze and Schröder (1964), Getty (1975), Schummer et al. (1976, 1981), Koch and Berg (1985), Dyce et al. (1987), as well as in detailed doctoral dissertations for by Hapke (1957) and Engelmann (1971) for the pig. The following is cited from Schummer et al. (1981). The A. ileocolica arises from the A.  mesenterica cranialis. It is caudoventrally directed and runs to the caecocolic junction. The colic ramus of the pig runs in wide spiral coils in the colonic cone, giving off numerous vessels to the centripetal coils. The right colic artery in pigs runs towards the distal part of the ascending colon and within the narrow spirals of the colonic cone. It gives off branches to the inner centrifugal coils of the ascending colon. The artery to the distal loop anastomoses with the A colica media, which is the artery that supplies the colon transversum. The A. mesenterica caudalis is the third unpaired visceral branch to arise from the ventral wall of the abdominal aorta. In the pig, its origin is located at the level of the 5th lumbar vertebrae (Schummer et al., 1976, 1981). The

442 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

A. colica sinistra is a branch of the A. mesenterica caudalis. It supplies the greater part of the colon descendens with short vessels and anastomoses with the A. colica media. The visceral veins that drain the colon into the V.  portae have also been described by Schummer et al. (1981). The V. mesenterica cranialis obtains blood from the V. ileocolica draining ileum and colon. The V. mesenterica caudalis is the terminal branch of the portal vein, to which it transports its blood. Zahner and Wille (1996) concentrate on the intramural vascular system of the porcine large intestine. In the following, their English summary will be cited: According to them, the Aa.  et Vv.  breves et longae leave from the mesenteric vessels and approach the colon via the mesentery. The short vessels enter the deeper layers of the wall, whereas the Aa. et Vv. longae, by taking a variable subserous course, reach the submucosa after penetrating the muscular layers. The tela submucosa contains an arterial and a venous vascular plexus. A deep vascular plexus supplies and drains the inner circular muscles, whereas a superficial plexus is adjacent to the lamina muscularis mucosae. The vascularisation of the mucosa is derived from the (superficial) submucosal plexus. The arteries that ascend the tunica mucosa form networks around each crypts of Lieberkühn. Close to the lumen, a polygonal subepithelial capillary system is formed. These subepithelial capillaries are furnished with a fenestrated endothelium, but the capillaries surrounding Liebekühn’s crypts mainly show a continuous endothelium. 28.3.6 The caecum of the Suidae, general remarks According to Savage (1983), the microbial symbionts of the caecum and colon can be found – free in the lumen – they can be associated intimately with particles of digesta (e.g. plant material consumed by the animal) or – associated with the epithelium The mucosal surface of the caecum can vary considerably, but Wille (1975) was not able to elucidate whether the differences encountered were related to different functional conditions. He also mentions the possibility that some of the differences are characteristic for certain species of mammalian hosts. 28.3.6.1 Anatomy of the porcine caecum Before starting more detailed anatomical considerations, reference should again be made to Fig. 5.126, which is based on data supplied by Horst (1956) and Bayley (1978).

Comparison of relative volumes of the stomach and small and large intestines in four domestic ungulate species shows the relatively large small and large intestines in the pig. However, the relative importance of the caecum in relation to the total post-gastric digestive tract has been shown by Bergner and Ketz (1969): The caecum is relatively small in the pig; it represents about 8% of the total gut volume (small intestine 47%, large intestine 45%). The main function of the caecum is intensive mixing of the digesta with the microbial flora. This intensive mixing makes alloenzymatic digestion in the colon possible. Mosenthin (1998) writes that bacteria in the caecum and proximal colon have a much higher growth rate than those located more distally because of the availability of substrates that present the basis of bacterial growth in the large intestine. According to Fonty and Gouet (1989), the microflora in porcine gastrointestinal tract shows some adaptation to different types of food. Although it is difficult to make a strict estimate, bacterial proteins constitute the major part of the faecal protein in pigs (Laplace et al., 1985). In contrast to bacteria, protozoa were not intensively investigated (Van der Heyde, 1973). In a comparative study dealing with different suids, Macdonald (2000) remarks that the structure of the intestinal tract is largely the same in different species of pigs (domestic and wild hog, pygmy hog, common warthog, bushpig, babirusa). However, in species other than Sus scrofa, specific information on the caecum is extremely limited. In a comparative investigation of the porcine and tayassuine digestive tracts, the same author (Macdonald, 1991c) only informs about the “short and wide” caecum of the bush pig (Potamochoerus porcus). The terminal portion of the ileum in the domestic pig ends in the large intestine at the level of the ileocaecocolic junction, forming and acute angle with the caecum (Prado et al., 1997). With the exception of the incoming ileum there is no demarcation between the caecum and the colon (Crevier, 1993). According to Sisson (1975), the ileum joins the caecum obliquely and projects considerably into the latter. Crevier (1993) mentions a caecal length of the domestic pig between 30 and 40 cm and gives a diameter of 8–10 cm. The caecum of Sus scrofa has a capacity of 1.5–2.2 L, is characterised by a cylindrical form with a slightly tapering blind end (Nickel et al., 1967, 1973). One of the branches of the porcine A. mesenterica cranialis is the A. ileocaecocolica (Schiltsky, 1966). After this has given off the R. colicus, the A. caecalis runs towards the apex caeci. The caecum and the ileum are supplied via an arterial meshwork (Chatelain, 1973), which is formed by the A.  ileocaecalis (Fig. 5.128). The caecal artery itself is represented by a bundle of vessels which sends secondary branches over the lateral and medial



Fig. 5.128: Arterial supply of the caecum in the pig. Adapted from Chatelein (1973).

bands to the antimesenteric border of the caecum (Nickel et al., 1976, 1981). According to Schiltsky (1966), arterial vessels run in the interhaustral folds towards the medial and lateral taeniae. These can be crossed and the arterial branches can be followed towards the antimesenterial side of the caecum. Zahner and Wille (1996) deal with the intramural distribution of blood vessels of the large intestine; some of their Figures depict the caecum. The subepithelial capillaries are furnished with a fenestrated endothelium, whereas the capillaries of the pericryptal mucosa mainly show a continuous endothelium. This indicates that there is trans-endothelian transport in the subepithelial region of the caecal wall. There are no sure indications as to the existence of either arterio-venous anastomoses or haemodynamic regulatory structures. The venous drainage of the caecum is presented in publications by Hapke (1957) and Niiyama et al. (1978). Although the latter authors in their illustration of the caecum do neither show the site where the ileum enters the large intestine, nor the exit from the most proximal part of the colon ascendens (Fig. 5.129), they demonstrate a venous branching mode that is similar to that of the arteries. The caecum has three taeniae that separate three rows of haustra from each other. In studies on minipigs, Engelmann (1971) found three taeniae on the caecum. The medial and lateral taeniae are free and converge at the apex caeci. On the ventral side the plica ileocaecalis or ileocaecal fold is fixed (Fig. 5.128). The caecocolic junction lies ventral to the left kidney. From here, the caecum extends caudoventrally along the left abdominal wall, so that the apex comes to lie in the inguinal region (Engelmann, 1971; Nickel et al., 1967, 1973). The caecum “lies against the dorsal and cranial part of the left flank and

28 Cetartiodactyla, Artiodactyls, Suidae 

 443

Fig. 5.129: Venous drainage of the caecum in the pig. Adapted from Niiyama et al. (1978).

extends ventrally, so that its ventral blind end usually lies on the floor of the abdomen, near the median plane” (page 1278, Sisson, 1975). A comparison of mucosal surfaces in the caeca of domestic mammals produced differences, which, according to Wille (1975), are either related to different functional states or are possibly species-bound. A clear proof of the latter statement is not given. The thickness of the layers of the wall of the digestive tract is highly variable. For example, Hedemann et al. (2002) writes that the tunica muscularis externa of the large intestine is significantly thicker in the region of caecal and colonic taeniae than in the region of haustra. Over the taeniae of the caecum the crypt volume is smaller than over the haustra.

28.3.6.2 Remarks on the functions of the porcine caecum Even in a mammalian species like the domesticated form of Sus scrofa, most information on the process of digestion refers to the “large intestine” and not specifically to the caecum. The impression arises that authors lump findings from caecum plus colon – especially the “proximal” colon ascendens – or that they do not even consider the relatively small (see above!) caecum. For example, Ruckebusch et al. (1981) attribute a “simple caecocolon segment” to the pig. Keys and DeBarthe (1974) give quantitative information on the digestion in this caecocolon segment: Approximately 100% of the cell wall and cellulose digestion and 80% of the hemicellulose digestion took place in the “large intestine” of swine, and Rérat (1978) emphasises the contribution of the large intestine to the digestion of carbohydrates and nitrogenous matter. The large intestine as a whole has a high autoenzymatic

444 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

digestive capacity (Just et al., 1981). Sakata and Saito (2007) also speak of the total large intestine when they mention the influence of solid food particles on retardation of alloenzymatic digestion in the large intestine. An interesting account on digestive physiology in the large intestine of swine is given by Stevens (1978): Colonic retention time of digesta was prolonged throughout the length of the haustrated colon. However, VFA absorption is slightly higher from the caecum than from the proximal and distal colon. According to Vervaeke et al. (1989), production of volatile fatty acids is very similar in the caecum and proximal colon, but in the distal colon VFA production is smaller than in the proximal colon. The decrease in the level of fermentable substrate is certainly responsible for these digestive differences between the proximal and distal colon sections, but not for any difference between the caecum and the proximal colon ascendens. A very clear impression of the functional significance of the porcine caecum, as compared with three consecutive sections of the colon is given by Schnabel et al. (1990): When the concentration of volatile fatty acids (mmol/100 g dry matter) in the caecum is considered as 100%, the three thirds of the total colon are characterised in oral-aboral direction by 55, 35 and 21%. This corroborates the above-mentioned statement of Vervaeke et al. (1989) that the distal acid production is smaller than in the proximal sections. Freire et al. (2000) mention the high digestibility values of the neutral detergent fibre (determining plant cell well without pectins, Van Soest, 1982) and acid detergent fibre (determining cellulose and lignin) fractions in the caecum of pigs. Imoto and Namioka (1978) published interesting data contradicting the results of Schnabel et al. (1990). They show that the caecum in the pig does not contribute significantly to microbial fermentation – alloenzymatic digestion – in the total large intestine: 1.5 hours after the morning meal the caecum contributes less than 20% to the production of volatile fatty acids in the large intestine, whereas the proximal colon contributes about 60% to the total VFA production! In the pig, the large intestine retains digesta for 20 to 38 hours (Mason, 1983), which represents 73.4% of the total retention time between the stomach and the anus. From these data on retention in the caecum and in the colon, it is obvious that the total large intestine retains digesta for a much longer time than the stomach or the small intestine. The primary role of the large intestine of the pig is that of water and mineral absorption, and the alloenzymatic digestion of fibrous plant cell wall materials. In addition, bacterial synthesis and absorption of the K and B vitamins (Durst et al., 1989) may have survival value (Mason, 1979).

28.4 Babyrousa sp. 28.4.1 Introductory remarks Although the babirusa is considered here as a species belonging to the Suidae, reference should be made to the paper of Wu et al. (2006) who believe that Babyrousa has to be clearly separated from other Suidae. According to these authors, “…the molecular evidence, together with the morphological evidences support the suggestion that Babyrousa becomes an independent family” (page 199). According to Mohr (1958), this genus consists of one species, Babyrousa babyrussa, with four subspecies (B. b. babyroussa, B. b. celebensis, B. b. frosti, B. b. togeanensis). According to Groves (1980), Babyrousa babyrussa is divisible into only three living subspecies: B.  b.  babyrussa (syn. frosti), B.  b.  togeanensis and B.  b.  celebensis. On Sulawesi (Celebes), the babirusa remains abundant, but is decreasing (National Research Council USA, 1983). Macdonald, Burton et al. (2008a, b, c) in their Red Data Lists differentiated the same subspecies. Meijaard and Groves (2002) upgraded subspecies of the genus Babyrousa to species level. Grubb (2005) mentions four species: Babyrousa babyrussa (Buru babirusa, already described and depicted by Wallace, 1869, in his book “The Malay Archipelago”), B.  bolabatuensis (Bola Batu babirusa), B. celebensis (North Sulawesi babirusa) and B. togeanensis (Malenge babirusa). These members of the family Suidae are interesting inhabitants of zoos outside Indonesia, but in descriptions of these zoo animals, e.g. by Mohr (1958), anatomical information on abdominal organs is not given. In a publication by Ruskin (1983), the babirusa stomach is mentioned in passing, but nothing else about other sections of the digestive tract can be found. Van Wees et al. (2000) assume that forestomach fermentation in Babyrousa is more important than caecocolic fermentation. This speculation is based on some “probing” anatomical investigations on museum material by Langer (1974a, 1988).

28.4.2 Food eaten by the babirusa Babyrousa babyrussa browses leaves, eats mushrooms, fruit, bark, thistles, grubs, and some roots, but compared with pigs they do little rooting (Conklin et al., 1994). However, Leus et al. (2001) remark that only very few food items are clearly identified in the older literature, but food composition according to these authors resembles that presented above. From this food, the following apparent digestibilities have been determined for Babyrousa babyrussa by Van Wees et al. (2000): 71% for crude protein,



28 Cetartiodactyla, Artiodactyls, Babyrousa sp. 

 445

76% for crude fat and 78% for non-structural carbohydrates. These authors also write that much more hemicellulose (75%) than cellulose (19%) is digested, which means that cell wall constituents can be partially digested according to the degree of lignification (Van Soest, 1982).

28.4.3 Anatomy and mucosal lining of the stomach In the following description of the anatomy of the babirusa stomach, extensive reference will be made to previous publications of the present author (Langer, 1973, 1988). Davis (1940) mentions that the stomach is “much more elongate” than that of Sus scrofa, a fact that is also shown in illustrations that were published much more recently by Leus (1994) and Leus et al. (1999). Both Davis (1940) and Vrolik (1843) consider the stomach of Babyrousa babyrussa to consist of two compartments. Langer (1973) mentions that the corpus and fornix region are separated from each other by an externally visible constriction, but that they have a wide internal connection (Fig. 5.130). The stomach lies in the epigastric region and is covered ventrally by the liver. The fornix is a region of considerable volume. On it lies a cap-like diverticulum ventriculi that is curved in situ from left to right, lying caudoventrally (Langer, 1973). Caudally lies the pars pylorica of the stomach (Vrolik, 1843). Onethird of the gastric volume in the babirusa is represented by the fornix ventriculi (Fig. 5.130). Langer (1988) gives the following absolute and relative data: Fornix gastricus: 329.01 cm3 (34.8%); diverticulum ventriculi: 15.43 cm3 (1.6%); corpus gastricum plus pars pylorica: 601.90 cm3 (63.6%). Langer (1973) depicted the stomach of Babyrousa babyrussa and showed a fold at the border between the corpus gastricum and the fundus gastricus spiralling into the fundus (Fig. 5.130). The pyloric antrum has a greater diameter than both the corpus gastricum and the canalis pyloricus, which is funnel-shaped. The fornix is the most voluminous part of the stomach. Despite the groove and fold between both gastric regions, there is a wide opening between the corpus and the fornix (Davis, 1940; Langer, 1973). The diverticulum ventriculi is separated from the fornix by an external groove which can be called, according to the terminology applied in swine by Pernkopf and Lehner (1937), a diverticulo-gastric sulcus. The internal lining of the stomach of Babyrousa babyrussa has already been described by Langer (1973, 1988): The cardia is surrounded by non-glandular squamous epithelium, runs about 6 cm along the lesser curvature towards the pylorus, almost to the middle of the lesser curvature (Fig. 5.131). The oral parts of the two lips of the ventricular groove are also covered with non-glandular squamous mucosa. The diverticulum, the fornix and parts

Fig. 5.130: View into the opened stomach of Babyrousa babyrussa. Adapted from Langer (1973).

of the corpus are lined with cardiac epithelium. The proper gastric glandular region lies in the zone of the antrum pyloricum, the most proximal section of the pars pylorica. The border between the proper gastric and pyloric mucosa could not be clearly delineated. Nevertheless, this border between epithelial types could be approximated because of the more intensive folding found in the proper gastric glandular mucosa. According to investigations published by Leus (1994) and Leus et al. (1999), not only data of the mean area of gastric sections, but also of the areas lined with the four different mucosal types that were determined from two healthy and one animal suffering from chronic diarrhoea could be made available. The diverticulum ventriculi

Fig. 5.131: Glandular lining of the total stomach of Babyrousa babyrussa. Adapted from Langer (1973).

446 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

contributed about 9.3% to the total surface of the babirusa stomach; the fornix gastricus 34.7%, the corpus gastricum almost half of the total surface (48.7%) and the pars pylorica 7.3%. According to Leus (1994), the largest part of the stomach in Babyrousa was lined by cardiac mucosa (75.7%); squamous, non-glandular mucosa contributed 4.7%, proper gastric glands 12.3% and pyloric glands 7.3%. For an adult Babyrousa babyrussa Langer (1988) was able to determine the mucosal lining of the babirusa stomach. Cardiac mucosa: 52 cm2 (60%), squamous non-glandular mucosa: 8 cm2 (9%), proper gastric gland mucosa: 17 cm2 (20%) and pyloric gland mucosa: 9 cm2 (11%). Davis (1940) describes a “reticulated, honeycomb appearance” of the pars pylorica. This feature could neither be detected in the material available to the present author, nor in equivalent mucosal regions of the domestic pig and the white-lipped peccary (Leus, Macdonald et al., 2004a). In an interesting paper, Macdonald et al. (2008d) deal with a “honeycomb structure” in the cardiac gland region of Babyrousa. They found these structures to be formed from cells of the epithelial surface. Within the honeycomb structure bacteria can be found, associated with the mucus in the cardiac pits. “The micro-environment of the cardiac gland area may be consistent with specialised microbial fermentation of ingesta or may support microbial modification of glandular secretions” (Macdonald et al., 2008d, page 39). In a descriptive, but not functional, immunohistochemical study on the distribution of endocrine cells in the gastrointestinal tract Agungpriyono et al. (2000) mention that the situation in Babyrousa babyrussa is similar to the pig. It is difficult to understand the conclusion drawn in this study: “The distribution of gut endocrine cells might be related to the regulatory characteristics of the babirusa digestive tract” (page 173). As a consequence of intensive studies by Kristin Y. G. Leus form and function of the digestive tract in Babyrousa babyrussa have been further elucidated (Leus, 1994; Leus and Macdonald, 1997; Leus and Morgan, 1995; Leus et al., 1999, 2001, 2004b) and Macdonald and Leus (1995). Dissecting a wide range of material that was in a better state of conservation than that available to the present author, she was able to improve our knowledge of the gastric anatomy of this species. According to her studies and resulting illustrations, Leus (1994) was able to establish that the diverticulum ventriculi is externally separated from the fornix gastricus by a groove-like constriction and the fornix gastricus by another sulcus from the corpus gastricum. In the fornix gastricus, the bundles of the fibrae obliquae run first more or less parallel to the fold near the cardia and then turn towards the diverticulum. Similar to conditions in Sus, they take part in the formation of the fold running in

counterclockwise direction from the fornix into the diverticulum (Langer, 1973). Studying the architecture of the tunica muscularis Leus (1994) found that the corpus is more extended than Langer (1973, 1988) was able to show. However, the general distribution of muscular layers, as identified by both authors is not principally different (Fig. 5.132). On the incisura cardiaca “rides” the cardiac loop, the branches of which run parallel to the lesser curvature. They form the muscular basis of the lips lining the sulcus gastricus. Circular muscles that run perpendicular to the lesser curvature lie between the branches of this loop (Fig. 5.133). The two lips can be followed along the lesser curvature approximately to the border between corpus gastricum and pars pylorica, which comes very close, as Leus (1994) could show, to the pylorus. Fibrae obliquae corporis depart from the muscular lips of the sulcus gastricus and run towards the greater curvature thus forming the lower segmental loop of Forssell (1912, 1913) and Torgersen (1942, 1968). Leus (1994) and Leus et al. (1999) were able to demonstrate that all sections of the stomach wall are formed by a two-layered tunica muscularis. On

Fig. 5.132: Muscular architecture of the stomach of the babirusa, seen from the facies parietalis (A) and visceralis (B). Adapted from Langer (1973) and Leus (1994).



28 Cetartiodactyla, Artiodactyls, Babyrousa sp. 

 447

Fig. 5.133: Cardiac region of stomach of Babyrousa babyrussa after removal of the mucosal lining. Adapted from Langer (1973).

the lesser curvature, a pyloric torus is differentiated and forms the closing apparatus of the stomach together with the sphincter that is interwoven with the torus. 28.4.3.1 Gastric mesenteries in the babirusa This section cites the respective section of Langer (1988). In Babyrousa babyrussa, the course of the line of fixation of the mesogastrium dorsale, which contributes to the formation of the greater omentum, is principally identical to that in Sus scrofa (Fig. 5.134). Starting from the oesophagus it runs along the facies visceralis and through the groove that separates the ventricular corpus from the fornix. It crosses over to the facies parietalis and after a short course follows the greater curvature

down to the pylorus where it forms the short mesentery of the proximal duodenum (Langer, 1973). The mesogastrium dorsale might, according to Davis (1940), be responsible for the fact that the diverticulum ventriculi “is completely encased in a snugly fitting fold” of the dorsal mesogastrium (Davis, 1940, page 387). This latter author speaks of the “lesser omentum”, but his Figure 32 shows the greater omentum, formed by the mesogastrium dorsale! 28.4.3.2 Blood vessels of the babirusa stomach In Babyrousa babyrussa, both Davis (1940) and Langer (1973, 1988) found three branches of the A.  coeliaca, the Aa.  lienalis, gastrica sinistra, and hepatica communis

Fig. 5.134: Lines of fixation of both mesogastria to the stomach of Babyrousa babyrussa. Adapted from Langer (1988).

448 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

Fig. 5.135: Arterial supply of the stomach of Babyrousa babyrussa seen from the facies visceralis. Adapted from Langer (1973).

(Fig. 5.135). The fornix and the diverticulum ventriculi are supplied by two branches, one forming a branch of the A.  gastrica sinistra and the other branching off from the A. gastroepiploica accessoria (term coined by Davis, 1940). The cardiac gland zone, which lines the diverticulum ventriculi, fornix gastricus and proximal corpus gastricum of the babirusa (Leus, 1994) (Fig. 5.131) accounted for about 25% of the total gastric blood flow in Sus scrofa (Zamora and Reddy, 1981). Relatively the cardiac gland surface is much more extended in Babyrusa babyrussa than in Sus scrofa (75.7% according to Leus, 1994, pig about 30%). It can be assumed that the percentage of gastric blood flow that is directed to the cardiac gland area is relatively high in the babirusa. An ample vascular supply would be valuable.

characteristic of the Suidae (Davis, 1940; Mitchell, 1916) (Fig. 5.136, upper panel). The spiral coil agrees with that of Sus in that the centripetal part of the coil (diameter about 20 mm) exceeds the centrifugal part in diameter (10 mm). The following section, the colon transversum,

28.4.4 Small intestine of babirusa Only little information is given on the duodenum of the babirusa (Babyrousa babyrussa) (Davis, 1940). It is 370 mm long and forms a simple loop. Fifteen millimetre behind the pylorus, the ductus choledochus enters the duodenum. The ampulla duodeni has a length of 18 mm. 28.4.5 Colon of babirusa The colon ascendens of the babirusa has a length of about 3 m and is arranged in the double spiral coil that is

Fig. 5.136: The large intestine of Babyrousa and Hippopotamus under application of the terminology published by Schaller (1992). Illustrations are adapted from Mitchell (1916).



Fig. 5.137: External view of the caecum of Babyrousa babyrussa. Adapted from Vrolik (1843) and Davis (1940).

arches craniad. This region is followed by the descending colon, which runs straight caudad along the midline to the rectum (Davis, 1940). 28.4.6 Remarks on the caecum of babirusa The anatomical information available for the digestive tract of Babyrousa babyrussa is very unbalanced. For example, Agungpriyono et al. (2000) present a description of the distribution of endocrine cells in the gastrointestinal tract of that species, but any functional interpretation is missing. Even some speculative assumptions would be more stimulating than just a casuistic presentation of endocrine cells! Under these circumstances, it is not surprising that information on the morphology of the babirusa caecum comes from authors of the 19th (Vrolik, 1843) and the first half of the 20th century (Mitchell, 1916; Davis, 1940) (Fig. 5.137). The ileum opens into the caecum at an oblique angle (Fig. 5.138, Mitchell, 1916). Where the ileum ends and

Fig. 5.138: The caecum of Babyrousa babyrussa. Adapted from Mitchell (1916).

28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 449

the colon begins, Vrolik (1843) identified two semilunar folds, with a very small opening between them. Reflux from the caecum or the colon into the ileum is improbable. The folds can be found in a rather complex proximal part (Davis, 1940), which extends into a simple conical apex caeci. The terminal section is a simple diverticulum. Although anatomical knowledge on the digestive tract, and especially the caecum, in the babirusa is still very limited, Clauss et al. (2008a) endeavoured to study Sus scrofa and Babyrousa babyrussa and other species of the Suidae, as well as Tayassuidae, comparatively. “Forestomach” structures in babirusa do not have the effect of improved fibre digestion; the caecum is not mentioned by Clauss et al. (2008a). Regardless of differences in digestive anatomy, these authors remark that Babyrousa babyrussa and Sus scrofa appear to share a similar degree of dietary protein digestion and that fibre digestion is not principally different between both species.

28.5 Tayassuidae 28.5.1 Introductory remarks Presently, there live four genera of the Tayassuidae exclusively in the New World; the white-lipped pecari, Tayassu pecari (March, 1991; Grubb, 2005), the collared pecari, Pecari tajacu (Schmidt, 1991), Pecari maximus, the giant peccary, and the Chacoan pecari, Catagonus wagneri (Taber, 1991). Detailed comments on their distribution and biology can be found in Mayer and Brandt (1982), Sowls (1984, 1997), Eisenberg (1989), Redford and Eisenberg (1992), Eisenberg and Redford (1999). In all these studies, anatomical information is either very limited or nonexistent. Investigations on mitochondrial Cytochrome B DNA sequences by Theimer and Keim (1998) grouped Catagonus wagneri and Tayassu pecari in a closely related clade that is separate from Pecari tajacu. Groves and Grubb (1993) and Grubb and Groves (1993) describe the taxonomy and external form of collared, white-lipped and Chacoan peccary. The latter, which had been discovered in 1974, was described by Wetzel et al. (1975) as a surviving “fossil” species, which had persisted in a scrub-thorn and grass refugium in Paraguay and northeastern Argentina. A thorough study of the osteology of the cranium of Catagonus wagneri was published 2 year later by Wetzel (1977). In January 2000, another species, Pecari maximus, the giant peccary, was discovered on the banks of Rio Aripuanã, Brazilian Amazonia (Van Roosmalen et al., 2007, Gongora et al., 2011), with a range extending to northwestern Bolivia, as Moravec and Böhme (2007) described.

450 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

This “new” peccary species shows close relationship with collared peccaries and is therefore attributed to the same genus (Gongora et al., 2007); according to Van Roosmalen et al. (2007), Pecari maximus is a sister species to Pecari tajacu. Both species are separated from Tayassu pecari and Catagonus wagneri. The giant peccary, the “peccary that goes in pairs” according to local Indians, lives in very small groups. More recently, it was stated that the morphological and ecological data that were used to claim the species status of Pecary maximus seem to be deficient and inconclusive (Gongora et al., 2011). This means that this “new species of peccaries” is still questionable and uncertain. The family Tayassuidae seems to have originated in South East Asia (Ducrocq, 1994; Gongora and Moran, 2005), and then migrated into the New World as early as late Eocene. According to Ducrocq et al. (1998), the most primitive Eurasian tayassuid has been found in southern Thailand. However, this upper Eocene species (Egatochoerus jaegeri) does not represent a stem-taxon of both tayassuids and suids (Ducrocq, 1998). In the Oligocene and the earliest Miocene (Aquitanian) of Europe and North America, genera were identified with undoubted relationships to recent swine, but also to peccaries (Pearson, 1927). The earliest genus, Perchoerus, immigrated from Asia into North America about 36 million years ago (Prothero, 2009). Webb (1985) had already stated that the earliest fossil peccaries in South America represented none of the three living taxa. Migration of early peccaries also brought them to areas where they became extinct. For example, “Old World peccaries” (Palaeochoeridae) lived in Southern Germany during the Miocene (Van der Made, 2010) and according to Pickford and Morales (1989, 1998) tayassuids spread into central Spain during the early Miocene (later than 26 Ma.) and even as far south as Langebaanweg in the Western Cape Province of South Africa about 5 Ma, as Hendey (1976, 1982) reports. According to him, it is likely that the peccaries of the Americas never had links with the Old World. Carroll (1988) writes: “peccaries are first known in the early Oligocene of North America” (page 512). The relationships between surviving tayassuid species was studied by a couple of authors. For example, Theimer and Keim (1998) came to the conclusion that Catagonus wagneri and Tayassu pecari form a clade separate from Pecari tajacu. The divergence took place 3.4 to 7.4 Ma, i.e. before North American peccaries colonised South America. Based on studies of peccary chromosomes, Adega et al. (2007) classified the autosomal karyotype of Pecari tajacu as “more primitive” than that of Tayassu pecari. Gongora and Moran (2005) agree with the separate position of the collared peccary, Pecari tajacu, but mention that this is in contrast with earlier morphological taxonomy, which clusters white-lipped and collared

peccaries in the genus Tayassu. During the time when the present author studied the gastric anatomy of the collared peccary (Langer, 1978, 1979a), another name for this genus of this species was applied, Dicotyles tajacu.

28.5.2 Remarks on the biology of peccaries Sowls (1984, 1997) compiled data on biology, management and use of three peccary species. Pecari tajacu occupies the largest space in that book because it has the largest latitudinal range and has come into closer contact with European man than have Tayassu pecari and Catagonus wagneri. Ranging from the xeric regions of southwestern North America over Central American and the South American tropics to the xeric regions of northern Argentina, climatic tolerance of Pecari tajacu has to be extensive (Bodmer and Sowls, 1993). The collared peccary has to survive lack of open water, as well as frost and snow and the damp climate of the Amazonian rainforest. Generally food constituents of collared peccaries can be roots, tubers, fruits, nuts, green plant parts, also invertebrates and cladophylls of prickly pear cactus (Opuntia). The remarkable adaptability of P.  tajacu, which also lives in the Republic of Trinidad and Tobago (Eisenberg, 1989), stimulated trials by Young et al. (2012) to farm the collared peccaries as semi-domestic animals. Similar investigations of collared peccaries under confinement were also undertaken in Columbia (Montes Pérez et al., 2012). Pecari tajacu and Tayassu pecari are similar in relation to habitat and diet (Donkin, 1985). The white-lipped peccary (T. p.) forms larger groups – Lévi-Strauss (1955) speaks of troops of 50 or more, but sometimes herds of 100 and more individuals can be seen (March, 1993, Peres, 1996; Fragoso, 1998); Moreira-Ramírez et al. (2015) mention groups of over 300 individuals. The collared peccary (P. t.) generally form groups of 10 to 20, rarely 50. Tayassu pecari is a consumer of fruits and seeds of tropical forest trees. Over most of their range they occur together with collared peccaries. T. pecari are generally bigger (mean weight males: 33.5 kg) than the collared peccary (mean weight males: 19.4 kg, females: 19.7 kg) (Eisenberg and Redford, 1999). Although there is much overlap in the diet of the two species, only the whitelipped peccary can break open the seed wall of very hard palm nuts (Kiltie, 1981, 1982; Bodmer, 1989). March (1993) classifies Tayassu pecari as omnivorous, they eat fruits, seeds, roots, invertebrates, small vertebrates, fungi and carrion, but they have a strong tendency towards frugivory. Catagonus wagneri, with a mean weight of 37 kg, is, as Wetzel (1977) writes, a diurnal, cursorial animal, better equipped for escape from predators than the nocturnal Tayassu. Preliminary stomach analyses and tooth



structure indicate that C.  wagneri is a browser. Three species of living peccaries meet in the Gran Chaco during what may be an interim period between a more arid situation that will favour C. wagneri and a moist time that will increasingly favour the other two species in a more mesic forests and (Wetzel, 1977).

28.5.3 Food of the peccaries Monographic studies on the food of the collared peccary, Pecari tajacu, are rather unbalanced because most of them originate from North America. Working in Arizona, Eddy (1961) classified the collared peccary as herbivore, which adapts to material that is available. Prickly pear pads and fruits are eaten together with tubers and other materials, obtained by rooting (Schmidt, 1991). Comparable food was also found in Texas, for example, by Corn and Warren (1985), who also mention seasonal changes of plant material eaten. They corroborate findings of Everitt et al. (1981) and Hominick (2014). On the other hand, forbs are dominant year round as feeds of collared peccaries according to Ilse and Hellgren (1995). Studying collared peccaries in the Atlantic Forest of eastern Brazil, Senra Motta et al. (2008) mention that these animals crunch most of the seeds that are eaten, but intact guava seeds that have passed the gastrointestinal tract of these peccaries, showed improved germination. The white-lipped peccary, Tayassu pecari, is mainly omnivorous, as Mayer and Wetzel (1987) and March (1991) mentions. In the faeces of animals from Costa Rica studied by Altrichter et al. (2000), 61.6% consisted of fruits, but diet composition varies between seasons and habitats. Leaves and stems of herbaceous plants constitute a food resource when fruits are scarce (Altrichter et al., 2001). In the central llanos region of Venezuela, Tayassu pecari and Pecari tajacu coexist (Barreto et al., 1997). Fruits and seeds and roots and plant stems comprise most of the diet and proportions of these items differed significantly between species. The studies of Pérez-Cortez and ReynaHurtado (2008) in the state of Campeche, Mexico, revealed that Pecari tajacu consume a high percentage of fruits in their food (57.9%). Also in Tayassu pecari the main components of the food were fruits (82.2%). Although both peccary species shared 32 species on their diet, their feeding niches did not significantly overlap. In the Caatinga, a xeric thorn forest in northeastern Brazil, Pecari tajacu cannot be called a frugivore, as Olmos (1993) emphasised: Only 1.4% of the food consisted of fruits, but 40.0% of tubers and 32.8% of roots; in Tayassu pecari no fruit was eaten, but 6.0% of the food consisted of tubers and 78.9% of roots. Beck (2006) remarked that evolution

28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 451

of a strong mastication apparatus, unique interlocking canines and foraging ecology in both species are adaptations to exploit hard seeds, particularly palm seeds. There is considerable niche overlap (Desbiez et al., 2009). Prado (2013) published interesting comments on the regional variation of feeds of collared peccaries. In Arizona, Opuntia engelmanni is especially important in the collared peccary’s food. In North and Central America, the food of this species was composed mainly of leaves and roots. Pecari tajacu living in South America are mainly frugivorous. The principal food of Catagonus wagneri, the Chacoan peccary, are parts of cacti (Taber, 1991, 1993), which, according to Mayer and Brandt (1982), Mayer and Wetzel (1986) represent the major winter food. Fruits are eaten in early spring. It was hunting stress (Sowls, 1997), not lack of food, which is responsible for a decline in number of Catagonus wagneri.

28.5.4 Anatomy of gastric compartments More than 300 years ago Tyson (1683) gave a precise description of the gastric parts in “Tayassu”, which can clearly be correlated with the terminology applied by Langer (1978, 1988) (Fig. 5.139). On page 364 Tyson (1683) writes: “…what first of all we took notice of, was the remarkable structures of the Stomacks, for it had three. Into the middlemost, was inferred the Oesophagus or Gullet; which we therefore shall call the first ventricle or Stomack (gastric pouch, according to modern terminology). From this, on one side was a large passage into the second (blindsac junction); which pouching out had its two ends winding like a horn (caudodorsal and cranioventral blindsacs); and on the other side of the first or middle stomack, was a free open passage into the third (glandular stomach), which emptyed itself into the duodenum.” Tyson (1683) writes that the two blindsacs had already been mentioned as “duas appendices” or “cornua” by Adrianus Falcoburgius (*1581, +1650) from Leiden, Netherlands. Short accounts of the tayassuid stomach have been given by Cuvier, 1835; Vrolik, 1843; Alessandrini, 1857; Cordier, 1893b; Oppel, 1896. Pernkopf and Lehner (1937) and Stewart (1964) also supply some information of the gastric anatomy of the collared peccary. The following descriptions of the gastric anatomy will be based on data from that literature and own dissections (Langer, 1973, 1978, 1979a, 1988). From these previous publications the present author will cite intensively with emphasis on the gastric anatomy of P. tajacu. Benirschke et al. (1986) found that the stomach of Catagonus wagneri is essentially of the same structure as that of Pecari tajacu.

452 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

Fig. 5.139: External aspects of the stomach of Pecari tajacu from different perspectives. Adapted from Langer (1988).

From data that were compiled by Langer (1988) and discussed by Schwarm et al. (2010), peccaries have the smallest stomach – relative to body mass – of all “forestomach fermenters”. According to Edelmann (1889), the stomach of Pecari tajacu consists of three sacs of different shapes (Fig. 5.139). The left sac, called “blindsac junction” by Langer, 1978, 1988) has two diverticles or “cranioventral and caudodorsal blindsacs”. The middle sac or “gastric pouch” (Pernkopf and Lehner, 1937; Langer, 1978), into which the oesophagus opens, is separated from the left sac by a circular constriction and from the right sac or “glandular stomach” by a strong constriction on the lesser

curvature. Two blindsacs communicate with each other on the greater gastric curvature via the blindsac junction (Fig. 5.140). Edelmann (1889) differentiates a shorter, but wider, blindsac from a narrow one. The right gastric sac, the glandular stomach, first bends ventrad at the ostium connecting it with the gastric pouch (Langer, 1973), its greater curvature is the most ventral part of the stomach. From this region the pyloric part of the glandular stomach turns dorsally and opens into the duodenum. A thickening of the muscular wall, especially near the greater curvature, can be seen, although a true pyloric torus is not present (Langer, 1973, 1978).



28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 453

Fig. 5.140: Caudodorsal (A) and right dorsal (B) view of a dried and opened stomach of Pecari tajacu. Abbreviations: A: gastric pouch, A’: blindsac junction; B: caudodorsal blindsac; C: cranioventral blindsac; D: glandular stomach; E: duodenum; O: oesophagus; a: cardia; b: fornical fold; b’: fornical sulcus; c: caudodorsal blindsac fold; c’: caudodorsal blindsac sulcus; d: cranioventral blindsac fold; d’: cranioventral blindsac sulcus; e: ventricular sulcus; f: ventral transverse fold; f’: ventral transverse sulcus; g: dorsal transverse fold; g’: dorsal transverse sulcus; h: pylorus. Asterisks mark the border between squamous and glandular mucosa. From Langer (1988), drawing by C. Thiele.

The two blindsacs lie in a parasagittal plane in the left hypochondrial region. It has already been mentioned above that the two blindsacs join the stomach on the socalled blindsac junction which itself is a wide connection (60 × 60 mm, Langer, 1978) with the gastric pouch (Fig. 5.140). The perimeter of the opening between the gastric pouch and the blindsac junction is formed by the fornical sulcus and the fornical fold (Langer, 1978). In addition to the ventral transverse sulcus that runs perpendicular to the greater curvature, a dorsal one can also be found that corresponds externally to the internal dorsal transverse fold (Fig. 5.141). The aperture between gastric pouch and glandular stomach has the following diameter: 40 × 50 mm (Langer, 1978). It should be noted that in relation to the gastric longitudinal axis, the ventral transverse fold lies more orad than the dorsal transverse fold. A clear and effective sphincter between the above-mentioned forestomach regions and the glandular stomach, where HCl is produced and acid digestion takes place, is not differentiated; only transverse folds inhibit the aborad transport. In the forestomach, extensive motility has been shown (Wyburn, 1979, personal communication). Mixing of gastric content could be the result of this motility. The ventral part of the gastric pouch and the upper blindsac touch the left abdominal wall (Langer, 1978). The anterior blindsac lies in the left hypochondriac region. The dorsally situated blindsac junction is covered by the

left part of the cupula of the diaphragm. Viewed from the left side of the animal the most dorsal part of the blindsac junction covers the cardia. The region of the blindsac junction near the hiatus of the oesophagus is affixed to the diaphragm by the dorsal mesogastrium or omentum minus (Fig. 5.142). Wyburn (1979, personal communication) of Murdoch University, Western Australia, supplied the present author with outlines, taken from a radiographic cine-film of a peccary stomach during a contraction cycle (Fig. 5.143). These illustrations show that the gastric pouch and the blindsac junction change their volumes and positions considerably during the contraction cycle. The caudodorsal blindsac can be almost completely contracted, but also the cranioventral one can change from a voluminous sac to a narrow tube. From the above, it should be clear that all data that will be given can only represent an approximation because of the high variability of the capacities of different gastric sections. The following information will be cited from Langer (1988): All deep sulci of the peccary stomach correspond to internal folds and make it possible to determine the volume of different gastric compartments. In one adult female collared peccary with a body weight (BW) of 26.9 kg, Langer (1979a) determined – 19 hours after the animal had its last meal – an absolute gastric volume of 1.16 L. This means that the peccary stomach amounts to 4.3 L/100 kg BW. Dyson (1969) investigated eight adult

454 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

Fig. 5.141: Internal view of the left parietal wall of the stomach of Pecari tajacu showing internal structures (A) and mucosal lining (B). Adapted from Langer (1988).

Fig. 5.142: Laterodorsal aspect of the gastric region of a newborn Pecari tajacu. Adapted from Langer (1988).

collared peccaries and found a mean body weight of 26.3 kg. In an animal that had survived its last meal for 24 hours, the wet weight of the gastric content amounted to 1.11 kg. This means 4.2 L/100 kg BW. This is approximately twice the daily food intake: According to Shively (1979), the

dry matter (DM) intake as percent of body weight, determined in three feeding trials done in two collared peccaries amounted to a mean of 2.14 + 0.26 kg DM intake/day/100 kg BW. The food material used in the trials of Shively (1979) had a mean dry matter content of 92.4% which means a



28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 455

Fig. 5.143: Three outline drawings of different stages of contraction in peccary stomach from a radiographic cine-film. CVB: cranioventral bllindsac; CDB: caudodorsal blindsac; BJ: Blindsac junction; GP: Gastric pouch; GS: Glandular stomach. Adapted from R.S. Wyburn, Murdoch University, W.A., personal communication.

wet weight (WW) intake amounting to 2.32 kg WW/day/ 100 kg BW. The relative volume of the gastric compartment in the female adult peccary investigated by Langer (1979a) expressed as percent of the total gastric volume amounted to: caudodorsal blindsac: 11.3%, cranioventral blindsac: 29.4%, gastric pouch: 44.6%, glandular stomach: 14.7%. 28.5.4.1 Internal mucosal lining The internal folds or plicae are formed by all layers of the whole gastric wall. All the other internal, macroscopically discernible differentiations are connected with mucosal lining. The general distribution of the different epithelial types has already been depicted and described by Tyson (1683), Cuvier (1835), Edelmann (1889), Cordier (1893b), Oppel (1896), Pernkopf (1937), Pernkopf and Lehner (1937), Stewart (1964), Langer (1973, 1978, 1988). Forty-six percent of the interior of the total stomach in the adult collared peccary is lined with non-glandular squamous epithelium (Fig. 5.141 B), the surface lined with cardiac mucosa amounts to 40%, but only 14% of the total stomach is lined with mucosa of the proper gastric and pyloric type (Langer, 1978, 1979a). Proper gastric and pyloric mucosa show macroscopically similar aspects as found in other mammals. Squamous epithelium covers the gastric pouch of the collared peccary. This is, however, not completely so.

The tip of this compartment on the greater curvature is lined with cardiac mucosa, but surrounded on all sides by squamous mucosa (Fig. 5.141 B and Fig. 5.144). From the cardia runs a ventricular sulcus or gastric groove that had already been mentioned by Cordier (1893b) along the lesser curvature (Fig. 5.141 A). This structure can be followed to the dorsal transverse fold (Pernkopf and Lehner, 1937). The two lips of this groove are not very prominent, but this might well be so during early life or in specific phases of gastric movements; changes of external form have been made visible by the X-ray studies of Wyburn (personal communication) that were already mentioned above (Fig. 5.143). On the bottom of the gastric pouch nonglandular and cardiac glandular mucosa are separated from each other by a small “wall” of mucosa that could be called a “margo plicatus”, which is not as prominent, but similar to the margo plicatus of the horse (Krölling and Grau, 1960). In the blindsac junction and the two blindsacs, two types of mucosal lining can be observed, the cardiac and the non-glandular. Anatomists from Venezuela (García and Leal, 2003) studied the gastric anatomy of the collared peccary. They give a systematic account of the mucosal lining of the gastric compartments, but not in all cases a clear correlation with earlier data was possible. However, the greatest part of the study is worth to be cited (pages 124/125) and to

Fig. 5.144: Internal mucosal lining of the gastric wall in Pecari tajacu, seem from the left (A) and right (B). Adapted from Langer (1978).

456 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

be related with anatomical terminology applied in a study by Langer (1988). Reference to Fig. 5.141 B and Fig. 5.144 should be made. Nonglandular squamous mucosa in the gastric pouch (“Region 1” of García and Leal, 2003): It is characterised by a whitish non-glandular mucosa with short longitudinal ridges. It extends from the oesophageal opening to the left of the medial plane up to the fold which separates the gastric pouch from the blindsac junction and can be followed towards the pyloric region which García and Leal (2003) call right lateral chamber. Gastric pouch (“Region 2”): It has a greyish mucosa with numerous low folds and occupies the greater part of the area of the oesophageal chamber. Folds separate it from the left lateral chamber (“fornical” and “blindsac folds”, Figs. 5.140 and 5.141 A). An approximately circular area of cardiac mucosa lies ventrally and represents “Region 6” of García and Leal (2003) (Figs. 5.141 B and 5.144). Blindsac junction (“Region 3” or left lateral chamber, the blindsac junction) is lined with cardiac mucosa. Glandular stomach, fundic region (“Region 4”): Smooth aspect, colour greyish-pale reddish, with very thin walls, occupying the greater curvature and the right region of the true stomach (“estómago verdadero”). It is connected with the aglandular mucosa, but delineated clearly towards the gastric pouch. It becomes pale towards the pyloric region, into which it insensibly continues. Glandular stomach, pyloric region (“Region 5”): An area of whitish to pale reddish colour with thick (fat) longitudinal folds with smooth glandular mucosa and well-developed walls, especially towards the duodenum, where the pyloric sphincter is formed. Edelmann (1889) gave a description of the stomach and its mucosal lining in Pecari tajacu. The bottoms of the gastric pouch and of the two blindsacs have a mucosa with a welldefined muscularis mucosa. The largest part of the pouch and the blindsac junction, as well as a small strip of the glandular stomach are lined with a non-glandular cutaneous mucosa with a squamous multi-layered epithelium (Figs. 5.141 B and 5.144). The bottom part of the glandular stomach has a mucosa with parietal cells, i.e. proper gastric glands are differentiated. The upper section of the right gastric sac (very probably Edelmann, 1889, means the canalis pyloricus) has glands of pyloric character. In a histological study, Tamate and Yamada (1983) document a narrow zone of cardiac gland mucosa at the entrance

into the glandular stomach. This zone is not shown in the highly schematic illustration of the mucosal lining of a white-lipped peccary, published by Leus et al. (2004a). 28.5.4.2 Muscle architecture of the tunica muscularis Reference will be made in the following to the publications of Langer (1973, 1988). In his paper on the comparison of gastric forms in the Craniota, Pernkopf (1937) presented an illustration on the subdivision of the stomach of the peccary on the basis of the investigation of the tunica muscularis. He was able to differentiate the following gastric compartments: the gastric groove = sulcus ventriculi, the fornix with the two blindsacs, the corpus in the region of the gastric pouch and the pars pylorica. The pouch clearly has to be taken as part of the corpus region and not as a part of the fornix, as Pernkopf (1937) assumed. Both deep and superficial layers together form a net-like architecture of the tunica muscularis, which is just appropriate for a hollow organ. The superficial layer is formed by longitudinal musculature coming out of the oesophagus (Fig. 5.145) and running over the blindsac junction and along the glandular stomach. Along the greater curvature of the glandular stomach the longitudinal superficial layer can clearly be distinguished, but along the lesser curvature of gastric pouch and glandular stomach this layer is thinned out (Langer, 1973). In the pyloric region, longitudinal fibres are clearly discernible along the greater curvature. Branching off from the musculature of the superficial layer of the blindsac junction are muscular cell bundles that run obliquely towards the tips of both blindsacs. On both blindsacs a vortex is formed near the tip. Below the incisura cardiaca there is the fornical fold (Fig. 5.146). When it is cross-sectioned, musculature belonging to the superficial layer can be found to take part in the formation of this cardiac fold.

Fig. 5.145: Schematic illustration of the muscular layers in Pecari tajacu. Adapted from Langer (1988).



Fig. 5.146: Muscle architecture of the gastric groove in the stomach of Pecari tajacu. Adapted from Langer (1988).

Stewart (1964) gave an illustration of the internal muscular layer of the parietal side of the stomach of Pecari tajacu. Along the lesser curvature of the gastric pouch the internal layer consists only of circular musculature that forms the bottom of the gastric sulcus or groove (Fig. 5.146) (Langer, 1973). On both sides of this groove there are muscular ridges that are formed by oblique fibres. Both ridges form a loop (ansa cardiaca) or upper segmental loop around the opening of the oesophagus. They can be followed aborally towards the dorsal transverse fold where they thin out. From both muscular lips oblique fibres branch off and can be followed to the greater curvature of the gastric pouch. Those fibres that branch off from the cardiac loop further caudally contribute to the deep layer of the glandular stomach. The deep layer of the most aboral part of the glandular stomach is formed by circular musculature. A pyloric torus cannot be discerned, but the circular fibres of the pyloric sphincter form a thickened cushion of the greater curvature. 28.5.4.3 Arteries of the gastric region Information on blood vessel supply was obtained during topographical dissections (Langer, 1973, 1978) and the following information is partly cited from Langer (1988). The coeliac artery divides into three major vessels: lienal, left gastric, and common hepatic arteries (Fig. 5.147 A). This is fully corroborated by Cavalcante Filho et al. (1998) who investigated gastric arteries in Tayssu pecari and Pecari tajacu (Fig. 5.147 C): In 71.41% of the studies (28 whitelipped peccaries and 8 collared peccaries), the A. coeliaca subdivides into the above-mentioned three branches. A branch from the lienal artery supplies the upper blindsac,

28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 457

but the main vessel can be followed to the hilus of the spleen and a vessel with thicker calibre, the left gastroepiploic artery, runs to the greater curvature of the glandular stomach. It later forms an anastomosis with the right gastroepiploic artery. The A. gastrica sinistra is the second branch of the coeliac artery. This vessel can arise from the A. coeliaca together with the A. splenica or lienalis. The left gastric artery supplies the visceral side of the gastric pouch and glandular stomach. The diverticular artery, a branch of the A.  gastrica sinistra, runs in the fornical sulcus to supply the blindsac junction and anterior blindsac. The third and very prominent branch of the coeliac artery is the common hepatic artery. Its first branch is the A.  hepatica propria. On the lesser curvature of the right gastric part or glandular stomach the A. hepatica subdivides into the A. gastrica dextra and A. gastroduodenalis, which itself represents the origin of the A. gastroepiploica dextra (Cavalcante Filho et al., 1998). The A. gastrica sinistra ramifies into a cranial and a caudal branch, which runs to the gastric sac and the cranioventral and caudodorsal blindsacs. The described subdivision of the A.  coeliaca into A. lienalis, A. gastrica sin. and A. hepatica communis can be called a typical “tripus halleri”. This situation is not only typical for conditions in man, but has also been described in Sus scrofa as a variation in the gastric arterial supply by Schwarze and Schröder (1964). Cavalcante Filho et al. (1998) mention alternatives from the “tripod-type” of branching into A. lienal, A. gastrica sinistra and A.  hepatica communis: A truncus gastrosplenicus, formed by the left gastric and lienal (splenic) artery forms a two-branched “fork” together with the A. hepatica in 17.85% of the investigated cases. Langer (1988) describes the same situation with two branches of the A.  coeliaca, consisting of the A. hepatica and A. gastrica sinistra plus A. lienalis or splenica (A. gastrosplenica or gastrolienalis). In only 3.75% of all cases, Cavalcante Filho et al. (1998) found a truncus gastrohepaticus and an isolated A. splenica (Fig. 5.147 B, C).

28.5.4.4 Mesenteries of the peccary stomach The mesogastria of the peccary have only been described in papers of Alessandrini (1857) in foetal material and by Langer (1973, 1988). Reference to the chapter on gastric mesenteries in Langer (1988) will be made. As in other mammals, the spleen lies within the dorsal mesogastrium below the greater curvature (Langer, 1978). The area of the cardia is fixed to the diaphragm (Fig. 5.142). From here the ventral mesogastrium can be followed along the lesser curvature of both gastric pouch and glandular stomach down to the pylorus where it ends (Fig. 5.148) (Langer, 1973). The starting point of the line of fixation of the dorsal mesogastrium is the area of fixation of the cardiac region

458 

 V Laurasiatheria – 28 Cetartiodactyla, Artiodactyls

Fig. 5.147: Arterial supply of the stomach in Pecari tajacu. Modified from: Langer (1973) (A, B); Cavalcante Filho et al. (1998) (C).



28 Cetartiodactyla, Artiodactyls, Tayassuidae 

 459

Fig. 5.148: Semischematic illustration with the two mesogastria in Pecari tajacu. Adapted from Langer (1988).

to the diaphragm (ligamentum gastrophrenicum). From here the line of fixation of the mesogastrium dorsale can be followed via the parietal side of the stomach between the blindsac junction and the upper blindsac to the upper blindsac sulcus, from where it runs over the greater curvature of the gastric pouch. It then crosses the ventral transverse sulcus and runs over the glandular stomach to the pylorus and the right aspect of the loops of the small intestine are covered by the greater omentum.

28.5.4.5 Functional remarks on the peccary stomach According to Carl and Brown (1983), who studied Pecari tajacu, volatile fatty acids (VFA’s) have been found in the forestomach. Their research has shown forestomach pH as 5.0–6.2, which is a range that would be suitable for a microbial population (Dyson, 1969; Shivley, 1979). VFA’s are by-products of microbial fermentation and are the primary source of energy in ruminants (Church, 1969). Wild Pecari tajacu, eating prickly pear cactus, had high amounts of VFAs in their stomach contents, but no bacteria were observed (Hayer, 1961). However, Oliveira et al. (2009) found bacteria in the forestomach of collared peccaries. Most of the protozoa found in peccaries belonged to the genus Entodinium. The presence of a high population of VFA-producing protozoa lends more evidence to the possibility that the collared peccary may utilize VFAs for energy. Not only protozoa, but also archaea are resident in forestomach and stomach compartments of the peccaries, and its community composition is relatively constant between different compartments and different

specimens (Oliveira et al., 2009). It was speculated that the collared peccary might have a digestive physiology similar to ruminant animals. On the other hand, Shively (1979) writes that the collared peccary apparently lack the ability to digest fibre to any great extent. Pecari tajacu is unable to reduce particle size of its food to any great extent because of its oral anatomy and inability to chew the cud and prepare it for microbial digestion. In another study, the same author and her colleagues (Shively et al., 1985) present the information that dry matter disappearance from the commercial swine feed was nearly identical for the peccary and steer (73% vs. 75%), but cellulose digestion was negligible for the peccary (