Quaternary Fossil Vertebrates from Continental Portugal: Paleobiodiversity, Revision of Specimens and New Localities

Home , Abrigo do Lagar Velho, Algoz, Almonda, Continental Portugal, Furninha, Lapa do Santo

Darío Estraviz López Graduado em Biología

Quaternary fossil vertebrates from continental Portugal: Paleobiodiversity, revision of specimens and new localities

Dissertação para obtenção do Grau de Mestre em Paleontologia

Orientador: Prof. Associado Octávio Mateus, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa

Júri: Presidente: Prof. Auxiliar Paulo Alexandre Rodrigues Roque Legoinha (FCT- UNL) Arguente: Prof. Catedrático João Luís Cunha Cardoso (Universidade Aberta) Vogal: Prof. Associado Octávio João Madeira Mateus (FCT-UNL)

Júri:

Junho 2019

Darío Estraviz López Graduado em Biología

Quaternary fossil vertebrates from continental Portugal: Paleobiodiversity, revision of specimens and new localities

Dissertação para obtenção do Grau de Mestre em Paleontologia

Orientador: Prof. Associado Octávio Mateus, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa

A Faculdade de Ciências e Tecnologia e a Universidade Nova de Lisboa tem o direito, perpétuo e sem limites geográficos, de arquivar e publicar esta dissertação através de exemplares impressos reproduzidos em papel ou de forma digital, ou por qualquer outro meio conhecido ou que venha a ser inventado, e de a divulgar através de repositórios científicos e de admitir a sua cópia e distribuição com objetivos educacionais ou de investigação, não comerciais, desde que seja dado crédito ao autor e editor.

“In vertebrate paleontology, increasing knowledge leads to triumphant loss of clarity.”

Alfred Romer

ACKNOWLEDGEMENTS

It is a strange feeling to finally write this. I have been working on this master thesis for one and a half years and at the same time that I am happy about finishing it I feel like this is the end of an era: the end of the straight road that took me from the preschool to this master, always with paleontology in mind. I could have never reached this stage without the people that have accompanied me and for them are these words:

First I have to say thanks to my family. This dream of mine is only possible because of their tireless support. Thanks to my grandparents (Mariví, Josefa and Nicolás). Infinite thanks to my father Antonio for understanding my passion for paleontology and caring for me while I am in other country, if the computer where I am writing this is still working is thanks to you. Infinite thanks to my mother Silvia and my stepfather Suso, I do not have enough words to tell how much I appreciate that you come every few months to help me at my home in Portugal. All of you are the best family of the world and I am so lucky of having you by my side. You all make me happy in every possible way.

I want to thank my advisor professor Octávio Mateus. He has given me an immense freedom to follow the research path that I felt that was the most appropriate for each topic. The freedom to do my own failures (and his help to get out of them) has made me a better scientist.

I would like to say thanks to Miguel Moreno Azanza, that have really acted a lot of time as a de facto co-advisor, even when he had to take care of his real students. I won´t forget it. Also I would like to thank Eduardo Puértolas Pascual for his comments on the thesis and his help to overcome concrete problems during it.

Thanks to my fellow master students at the Lourinhã PaleoTeam for their companionship and support, Hugo Campos (And his family that opened me their house in Algarve), João Pratas, André Saleiro, Cátia Ribeiro, Francisco Costa, Vincent Cheng and Alexandra Fernandes; especially to my classmates of the fifth master edition Filippo Maria Rotatoria and Alexandre Guillaume. I won´t forget the dinners on Thursday nights in Almada!

Big thanks to the personnel of the Museum da Lourinhã Alexandre Audigane, Bruno Pereira, Carla Alexandra Tomás, Carla Abreu, Jorge Beja and Raquel Furtado. Thanks for letting me use the museum facilities and also for the paraffin heater to prevented me to freeze during winter while working.

Thanks to all the other people that stood by my side in Lourinhã: Isabel Mateus, Marta Mateus, Simão Mateus, Silvia Batista, Ana Luz, Micael Martinho, João Marinheiro, João Russo, Marco Marzola, João Muchagata, Marta, Isabel and all the English class, my training partners at “Miga fight team” and everyone else; you all complete my “second family” in Portugal and you make me feel at home in a foreign country.

But this thesis couldn´t have been done without many people that work outside Lourinhã.

I would like to thank the entire GPS (Grupo Proteção Sicó) especially to Sérgio Medeiros and all the other speleologists that have helped me several times to descend to the Algar do Vale da Pena, for their disinterested help. I owe you my life and that is not an exaggeration!

Thanks to my former advisor at University of A Coruña, Aurora Grandal d´Anglade, that has helped me immensely during the process of doing the chemical analysis of my samples. I promise I will get something interesting for ancient DNA soon.

iii

Many thanks to the paleontologist that helped me with the identification of the fauna of Santa Margarida: Pedro Callapez, Raquel Moyá Costa, Pavlos Piskoulis and Juan Manuel López García. I really hope to be able to keep working with you in further studies about this locality in the future.

Thanks to Miguel Ramalho, Jorge Sequeira and all the personnel of the Geological Museum in Lisbon, for letting me to visit their collections on short notice twice. Thanks to Jorge Graça, the Algarvian photographer that contacted Octávio Mateus and discovered the Santa Margarida locality.

Also thanks to Alexandra Elbakyan, without her work I could have never done this thesis.

Finally thanks to the students of other editions, personnel of the Department of Earth Sciences of the FCT-UNL and everyone that has helped me one way or the other and could not see his/her name on this thesis, I thank you a lot.

ABSTRACT

The Quaternary fossil record of Portugal is important for our understanding of the paleobiodiversity in Iberia. In the present master thesis a series of studies augment our knowledge about this topic. A census of Quaternary paleobiodiversity is carried out in order to test how reliable the fossil record is for detecting living species, resulting in that ~38% of living terrestrial tetrapods are recognized in the fossil record for Portugal, although the number of species recognized varies between groups. The body mass of a Portuguese proboscidean (Palaeoloxodon antiquus) is calculated via numerical methods for the first time (11metric tons) and morphometric comparisons of this species with Mammuthus primigenius are presented using an extensive Proboscidean sample. A new fossil brown bear (Ursus arctos) locality, Algar do Vale da Pena, with numerous claw mark in the walls of the cave (the first of this type of marks described in Portugal) is presented and the fossil bear remains identified and compared to a sample from NW Spain. The bears from Algar do Vale da Pena contrast with other previously known Portuguese brown bear specimens by relative small size. A new microvertebrate locality from Algarve, Santa Margarida, is presented. It is an extraordinary rich site with one fossil for every two grams of sediment selected and processed. The locality provided the first record of two arvicoline taxa in Portugal (Iberomys huescarensis and Victoriamys chalinei), which allows giving a minimal age of around 800.000 YBP for at least part of it. This makes Santa Margarida one of the oldest three localities in the Pleistocene of Portugal.

KEYWORDS

Holocene, Pleistocene, Mammalia, mammals, cave deposits, paleontology

RESUMO

O registro fóssil Quaternário de Portugal é importante para nossa compreensão da paleobiodiversidade na Ibéria. Na presente tese de mestrado, uma série de estudos aumenta o nosso conhecimento sobre a matéria. Um censo de paleobiodiversidade do Quaternário é realizado para testar a fiabilidade do registro fóssil para a detecção de espécies vivas, resultando em que 38% dos tetrápodes terrestres vivos são reconhecidos no registro fóssil de Portugal, ainda que este número varie entre grupos. A massa corporal de um proboscídeo português (Palaeoloxodon antiquus) é calculada através de métodos numéricos pela primeira vez (11 toneladas) e comparações morfométricas desta espécie com Mammuthus primigenius são apresentadas, utilizando uma extensive amostra de proboscídeos. É apresentada uma nova localidade de urso pardo (Ursus arctos); o Algar do Vale da Pena, com numerosas marcas de garras nas paredes da gruta (as primeiras deste tipo de marcas descritas em Portugal) e os restos fósseis de ursos são identificados e comparados com uma amostra procedente do Noroeste de Espanha. Os ursos do Algar do Vale da Pena contrastam com outros exemplares portugueses fósseis de urso pardo pelo seu tamanho relativamente pequeno. Uma nova localidade de microvertebrados do Algarve, Santa Margarida, é apresentada. É um local extraordinariamente rico, com um fóssil por cada dois gramas de sedimentos selecionados e processados. A localidade forneceu o primeiro registro de dois taxa de Arvicolinae em Portugal (Iberomys huescarensis e Victoriamys chalinei), o que permite determinar uma idade mínima de cerca de 800.000 YBP para, pelo menos, parte da jazida. Isto faz de Santa Margarida uma das três localidades mais antigas do Plistocénico de Portugal.

PALAVRAS-CHAVE

Holocénico, Plistocénico, Mammalia, mamíferos, depósitos de grutas, paleontologia.

vii

viii

RESUMO GRÁFICO | GRAPHICAL ABSTRACT

INDEX OF MATERIALS

1. INTRODUCTION ...... 1 1.1 History of the study of Quaternary fossil vertebrates from continental Portugal ...... 1 1.2 Vertebrates in the Quaternary of Portugal ...... 7 1.3 Aim, context and objectives of the present thesis...... 22 2. PALEOBIODIVERSITY OF THE QUATERNARY FOSSIL TETRAPODS IN CONTINENTAL PORTUGAL ...... 23 2.1 Introduction and methods ...... 23 2.2 Lists of localities and fossil tetrapods of the Quaternary of Portugal ...... 23 2.3 Results ...... 36 2.4 Discussion ...... 41 2.5 Conclusions ...... 42 3.THE PALAEOLOXODON ANTIQUUS OF SANTO ANTÃO DO TOJAL ...... 43 3.1 Introduction and methods ...... 43 3.2 Geographical and geological settings ...... 43 3.3 The material ...... 44 3.4 Measurements of femur and tibia ...... 48 3.5 Body mass estimation ...... 52 3.6 Analysis of metrical data ...... 54 3.7 Conclusions ...... 66 4. ALGAR DO VALE DA PENA: A NEW FOSSIL BEAR SITE FROM PORTUGAL ...... 67 4.1 Overview ...... 67 4.2 Introduction ...... 67 4.3 Methods, expeditions to the cave and preparation of the material ...... 73 4.4 The bone assemblages ...... 77 4.5 Results: The recovered bone remains ...... 80 4.6 The claw marks...... 118 4.7 Discussion ...... 119 4.8 Conclusions ...... 121 4.9 Further work ...... 122 5. SANTA MARGARIDA: A NEW MICROVERTEBRATE LOCALITY FROM ALGARVE ...... 123 5.1 Overview ...... 123 5.2 Introduction: Location, discovery, geological settings and nature of the site ...... 123 5.3 Preparation methods ...... 126 5. 4 Results ...... 129

5.5 Discussion ...... 143 5.6 Conclusions ...... 146 6. CONCLUSIONS ...... 147 Chapter 2: Paleobiodiversity of the Quaternary fossil tetrapods in continental Portugal ...... 147 Chapter 3: The Palaeoloxodon antiquus of Santo Antão do Tojal ...... 157 Chapter 4: Algar do Vale Da Pena: A new fossil bear site from Portugal ...... 148 Chapter 5: Santa Margarida: A new microvertebrate locality from Algarve...... 148 7. REFERENCES ...... 149

xii

INDEX OF FIGURES

Figure 1.1: Bust of Carlos Ribeiro at the Geological Museum, Lisbon (Photo by Octávio Mateus)...... 1 Figure 1.2: Picture of Nery Delgado at exposition in the Geological Museum, Lisbon...... 2 Figure 1.3: Hyena coprolite from Furninha cave recovered by Nery Delgado (MG). Note the glued identification.Geological Museum (Lisbon) ...... 2 Figure 1.4: Picture of Paul Choffat at exposition in the Geological Museum, Lisbon...... 3 Figure 1.5: Georges Zbyszewski (Picture by unknown author)...... 4 Figure 1.6: Miguel Telles Antunes in 2009 (Photo by Natacha Cardoso)...... 5 Figure 1.7: João Luis Cardoso in 2017 (Photo by Octávio Mateus)...... 6 Figure 1.8: “Contribuição para o conhecimento dos grandes mamíferos do Plistocénico Superior de Portugal” ...... 6 Figure 1.9: Vertebra of Salamandra salamandra from Figueira Brava cave in dorsal view; scale bar 1mm. Catalog number of the specimen not provided in the original (Crespo et al., 2000)...... 8 Figure 1.10: Mandible from an indeterminate lacertid recovered at the site of Santa Margarida (Portugal). FCT-UNL...... 8 Figure 1.11: Right dentary of Blanus cinereus from Figueira Brava in lingual view, scale bar 1 mm. Catalog number of the specimen not provided in the original (Crespo et al., 2000)...... 9 Figure 1.12: Hyoplastron of Testudo hermanni from Figueira Brava in ventral view; scale bar 1 cm. Catalog number of the specimen not provided in the original (Lapparent de Broin & Antunes, 2000). 9 Figure 1.13: Right femur of Bubo bubo (Linnaeus, 1758) from Furninha cave in anterior view. Geological Museum, Lisbon...... 10 Figure 1.14: Right humerus of Cygnus olor from Furninha cave in anterior view. Geological Museum, Lisbon...... 10 Figure 1.15: Pinguinus impennis proximal humerus from Furninha in lateral view. Geological Museum, Lisbon...... 12 Figure 1.16: Right hemimandible of Erinaceus europaeus from Furninha in lingual view. Geological Museum, Lisbon...... 13 Figure 1.17: Rodenia incisor recovered at the site of Santa Margarida (Portugal). FCT-UNL...... 13 Figure 1.18: Molar of Palaeoloxodon antiquus (MG 3561) recovered by Paul Choffat in Condeixa during the XIX century in lateral view...... 15 Figure 1.19: Partial molar of Stephanorhinus hemitoechus from Gruta da Serra de Molianos. Geological Museum, Lisbon...... 16 Figure 1.20: Type specimen of Equus ferus antunesi in lateral view (MG). Geological Museum, Lisbon...... 16 Figure 1.21: Anterior mandibular fragment of Hippopotamus incognitus from Condeixa in dorsal view. Geological Museum, Lisbon...... 17 Figure 1.22: Skull of female Cervus elaphus from Gruta das Fontainhas in lateral view. Geological Museum, Lisbon...... 18

xiii

Figure 1.23: Distal part of a humerus of Capra pyrenaica in anterior view from Caldeirão cave. Catalog number and scale of the specimen not provided in the original (Davis, 2002)...... 19 Figure 1.24: Fragment of left hemimandible of Cuon alpinus with P4 and fragment of M1 from Escoural cave in oclusal (left) and labial (right) views. Catalog number of the specimen or not provided in the original, scale do not appear in the image (Cardoso, 1993a)...... 20 Figure 1.25: Left Ursus arctos mandible from Furninha cave in lingual view. Geological Museum, Lisbon...... 20 Figure 1.26: Skull of Hyaena hyaena prisca from Furninha cave in dorsal view. Geological Museum, Lisbon...... 21 Figure 1.27: The leopard (Panthera pardus) skull from Algar da Manga Larga in anterolateral view. Geological Museum, Lisbon...... 22 Figure 2.2: Percentages of different parameters about the paleodiversity of the Quaternary in Portugal...... 37 Figure 2.3: Status of the different big mammals that appear in the fossil record in Portugal...... 38 Figure 2.4: Percentage of species from each group in extant, fossil and fossil species (without cave deposits)...... 41 Figure 3.1: The location of the Santo Antão do Tojal Palaeoloxodon antiquus is marked by the pushpin, in the Lisboa Peninsula (left) and at local level (right). Credit Google Earth...... 43 Figure 3.2: General view of the Palaeoloxodon antiquus femur of Santo Antão do Tojal (MG 5788) at display in the geological museum in posterior view. Scale bar 7 cm. The chalk mark signals a cut in the bone where a stone tool was found, being it regarded by Zbyszewski as a proof that the animal was butchered by Neanderthals...... 45 Figure 3.3: Right tibia of the Palaeoloxodon antiquus of Santo Antão do Tojal (MG 5793) at display in the Geological Museum in anterior view. Scale bar 7 cm...... 46 Figure 3.4: Proboscidean tibia of Santo Antão do Tojal (MG 5793) in left, anterior and right proximal views. The principal anatomical features are marked, in white the tibial depression and in red the medial articular surface. Scale bar 7 cm...... 46 Figure 3.5: The spinal process of the proboscidean of Santo Antão do Tojal (MG 8081) in A) anterior and B) posterior views...... 47 Figure 3.6: The phalanx from the proboscidean of Santo Antão do Tojal in A) dorsal; B) lateral; C) ventral and D) lateral views...... 48 Figure 3.7: Measurements of the femur of mammals by Von den Driesch, 1976...... 49 Figure 3.8: Measurements of the tibia of mammals by Von den Driesch, 1976...... 51 Figure 3.9: Plot of means for the previously selected measurements of the femur of Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with an interval of confidence of 95%. All measurements are in millimeters...... 57 Figure 3.10: Plot of means for the previously selected measurements of the tibia of Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with an interval of confidence of 95%. All measurements are in millimeters...... 58

xiv

Figure 3.11: Measurements of Santo Antão do Tojal specimen (in red) compared to the means for them in Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with a 95% confidence interval...... 59 Figure 3.12: Estimation of density for maximum length and minimum diaphysis width of the femur of Mammuthus primigenius and Palaeoloxodon antiquus...... 60 Figure 3.13: Estimation of density for maximum length and minimum diaphysis width of the tibia of Mammuthus primigenius and Palaeoloxodon antiquus ...... 61 Figure 3.14: From upper to bottom: Percentage of variation accounted by the first three Principal Component of the femur; PC1, PC2 and PC3. In PC1 and PC2 the highest contribution comes from the maximum length and the minimum diaphysis circumference respectively...... 63 Figure 3.15: PC1 vs PC2 of the femur of selected proboscideans. The Santo Antão do Tojal specimen is marked with the red square, Mammuthus primigenius in blue and Palaeoloxodon antiquus in black...... 63 Figure 3.16: From upper to bottom: Percentage of variation accounted by the first three Principal Component of the tibia; PC1, PC2 and PC3. In PC1 and PC2 the highest contribution comes from the maximum length and the minimum diaphysis circumference respectively...... 65 Figure 3.17: PC1 vs PC2 of the tibia of selected proboscideans. The Santo Antão do Tojal specimen is marked with the red square, Mammuthus primigenius in blue and Palaeoloxodon antiquus in black. 65 Figure 4.1: Location of the Algar do Vale da Pena. Upper left, at peninsular level, it is situated in the Portuguese Estremadura. Bottom, view at regional level, the closest city is Alcobaça. Upper right, view at local level. Credit Google Earth...... 67 Figure 4.2: Litostratigraphic log of the “Maciço Calcário Estremenho”, with the Candeeiros Formation marked (Carvalho, 1996)...... 68 Figure 4.3: Location of the cave (red and orange dot) in the corresponding sheet (26 D, Caldas da Rainha) of the geologic map (1:50000) of Portugal, including legend (Zbyszewski & de Matos, 1959)...... 68 Figure 4.4: Topography of the Algar do Vale da Pena. In black, some landmarks of the cave are marked and in red, the bone findings (Modified from Amendoeira et al., 2015)...... 70 Figure 4.5: A) Entrance of the cave showing the first ramp and the top of two meters step. B) Picture taken from the base of the two meters step, showing the second ramp. C) Final vertical well of the entrance from the bottom. D) Picture from the bottom of the cave showing the final well (white line), the second ramp (green line), the two meter step (red line) and part of the firs ramp (blue line). Pictures by Carlos Ferreira...... 71 Figure 4.6: Claw marks (left) and dog skeleton (right) in the small conduct at the right of the cave. .. 72 Figure 4.7: Thick deposit of sands below the calcareous crust (left) and claw marks (right)...... 72 Figure 4.8: Equipment needed to descend (in black) and climb out of the cave again (in red). Left; A) Shunt, B) Descensor, C) Climbing fist. Right; D) Ventral croll and security knot to stop descending, note how the fingers secure the shunt...... 73 Figure 4.9: First excavation around the skull of “bear number one” in October 2016...... 74 Figure 4.10: The main block as in October 2017, during the second expedition, no new diggings where carried out this time. The skull is near the scale of seven centimeters, ...... 74

Figure 4.11: A) Digging of the main block of “bear one” (note the camera of Rui Guerra). B) First layer of plaster after the block was cut. C) Putting the block in the net. D) Moving the block with the tree trunk. E) Final plastering of the block. F) Current state of the block. Pictures by Carlos Ferreira...... 75 Figure 4.12: Before ( left) and after( right) for some ursid remains recovered in Algar do Vale da Pena: a) distal radius, b) mandibular ramus and c) basicranium...... 76 Figure 4.13: The main block of “bear” one during the second expedition. Scapula and protruding long bones are marked with black arrows. In red, line where the block was cut, main skull with the scale of 7 cm and muddy area with numerous smaller bone remains...... 77 Figure 4.14: Remains of the “bear number two” before (left) and after (right) removal of the block that was hiding them...... 78 Figure 4.15: Muddy sediment below the calcareous crust, during the preparation of the hemimandible (ADVP 2.1). Scale 7 cm...... 78 Figure 4.16: Material of bear number four encased in the calcareous crust. Left, humerus; right, femur and pelvis...... 79 Figure 4.17: Distal humerus of the bear number five...... 79 Figure 4.18: Damaged scapula (ADVP 1.1) from “bear number one” (Ursus arctos) of Algar do Vale da Pena...... 81 Figure 4.19: Right scapholunate bone (ADVP 1.3) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) proximal; B) distal; C) anterior and D) exterior views...... 81 Figure 4.20: Measurements for the scapholunate bone in ursids according to the system of Tsoukala & Grandal-d’Anglade (2002). All of them could be recorded for the specimen...... 82 Figure 4.21: Transverse diameter vs anteroposterior diameter of the scapholunate of brown bears (García Vázquez, 2015) as brown dots and cave bears (unpublished own data from Cova da Ceza population) as green squares; plus the scapholunate ADVP 1.3, marked as a yellow star...... 83 Figure 4.22: Transverse diameter vs anteroposterior diameter of the scapholunate of brown bears (data from García Vázquez, (2015)) as brown dots; plus the scapholunate ADVP 1.3, marked as a yellow star...... 83 Figure 4.23: Right hamate bone (ADVP 1.4) from “bear number one” (Ursus arctos) of Algar do Vale da Pena...... 84 Figure 4.24: Measurements for the scapholunate ursid bone according to the system of Tsoukala & Grandal-d’Anglade (2002).The recorded measurement is marked...... 84 Figure 4.25: Length of the right hamate bone (ADVP 1.4) from “bear number one” of Algar do Vale da Pena compared to the length of the same bone of other brown bears (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002). Two groups (males and females) are classified according to this measurement...... 85 Figure 4.26: Second phalanx (ADVP 1.6) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) ventral; C) proximal; D) distal views...... 86 Figure 4.27: Measurements for the second phalanx of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002). All of them could be recorded...... 86 Figure 4.28: Length vs distal transverse diameter of the second phalanx (ADVP 1.6, marked as blue square) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the same

xvi ration of a recent brown bear, data from García Vázquez, (2015) . The 95% ellipses are drawn in the graphs. Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 87 Figure 4.29: Proximal epiphysis of the third right metatarsal (ADVP 1.7) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) exterior; B) dorsal; C) interior and D) ventral views. 88 Figure 4.30: Measurements for the third metatarsal of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002).The recorded measurements are marked...... 88 Figure 4.31: Proximal transverse diameter vs Proximal anteroposterior diameter of the third metatarsal (ADVP 1.6, marked as blue square) from “bear number one” (Ursus arctos) of Algar do Vale da Pena (As green triangle) compared to the same ration of fossil brown bears from NW Spain, data from García Vázquez, (2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 89 Figure 4.32: Distal epiphysis of the right radius (ADVP 1.8) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) anterior; B) exterior; C) Posterior; D) Interior and E) Distal views...... 90 Figure 4.33: Measurements for the radius of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). Recorded measurements are marked...... 90 Figure 4.34: Transverse diameter of distal radius from ADVP 1.8 and fossil and recent brown bears from NW Spain and Portugal (Cardoso, 1993; García Vázquez, 2015). Specimens with more than 60 mm are classified as male and less than 58 mm as females...... 91 Figure 4.35: Transverse diameter vs anteroposterior diameter of distal radius from ADVP 1.8 (Green in the graph) and fossil and recent brown bears from NW Spain (García Vázquez, 2015). Two clusters of individuals are classified as females or males...... 92 Figure 4.36: Distal distal part of the ulnar diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) anterior; B) exterior; C) Posterior and D) Interior views...... 92 Figure 4.37: Measurements for the ulna of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurement is marked with a circle...... 93 Figure 4.38: Minimum anteroposterior diameter of the ulnar distal diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared graphically to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Suggested males and females are marked...... 94 Figure 4.39: Posterior part of right hemimandible (ADVP 2.1) from “bear number two” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) posterior; C) exterior; D) anterior; E) interior and F) ventral views...... 94 Figure 4.40: Measurements for the mandible of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurements are marked with a circle...... 95 Figure 4.41: Total height of the mandible vs length from the mandibular condyle to the third molar from ADVP 2.1 (blue square in the graph), fossil and recent brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as green triangles (unpublished data from Cova da Ceza). Two clusters of individuals are classified as females or males...... 97 Figure 4.42: Total height of the mandible vs height of the mandibular condyle from ADVP 2.1 (blue square in the graph) and fossil and recent brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as red inverted triangles (unpublished data from Cova da Ceza)...... 97

xvii

Figure 4.43: Diameter of the mandible between M2 and M3 vs height of the mandibular corpus labially at M3 from ADVP 2.1 (blue square in the graph) and fossil and recent brown bears from NW Spain as black dots (García Vázquez, 2015) and a cave bear as a green diamond (unpublished own data from Cova da Ceza)...... 98 Figure 4.44: Partial occipital and parietal bones (ADVP 2.2) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) posterior and C) anterior views...... 99 Figure 4.45: Temporal and partial parietal bones (ADVP 2.3) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) exterior and B) interior views...... 99 Figure 4.46: Partial canine teeth (ADVP 2.5) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in lateral view...... 100 Figure 4.47: Fourth upper premolar (ADVP 2.6) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views...... 100 Figure 4.48: Measurements for the fourth upper premolar of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 102 Figure 4.49: Breadth vs length of the fourth upper premolar ADVP 2.6 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015)...... 102 Figure 4.50: First upper molar (ADVP 2.7) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views...... 103 Figure 4.51: Measurements for the first upper molar according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurements are marked with a circle...... 103 Figure 4.52: Breadth vs length of the first upper premolar ADVP 2.7 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as orange triangles from Liñares cave, Galicia, NW Spain (González & Martelli, 2003)...... 105 Figure 4.53: Second lower molar (ADVP 2.8) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views...... 105 Figure 4.54: Measurements for the second lower molar according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 107 Figure 4.55: Breadth vs length in mm of the second lower molar ADVP 2.8 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as green diamonds(García Vázquez, 2015) plus cave bears from A Ceza cave as black dots, Galicia, NW Spain (Unpublished own data)...... 107 Figure 4.56: Third lower molar (ADVP 2.9) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views...... 108 Figure 4.57: Measurements for the third lower molar according to the system of Tsoukala & Grandal- d’Anglade (2002)...... 108 Figure 4.58: Breadth vs length in mm of the third lower molar ADVP 2.9 (black dot in the graph) and the same measurements of recent and fossil brown bears from NW Spain as blue diamonds (García Vázquez, 2015) plus cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal- d’Anglade, 1993; González & Martelli, 2003; unpublished own data) as purple triangles...... 110

xviii

Figure 4.59: Third phalanx (ADVP 2.10) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) Distal; B) and D) Lateral; C) Proximal view...... 110 Figure 4.60: Measurements for the third phalanx of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002). Taken measurements are marked with a circle...... 111 Figure 4.61: Height and diameter of the third phalanx (ADVP 2.10) (black dot in the graph) and the same measurements of recent and fossil brown bears from NW Spain as blue squares (García Vázquez, 2015) plus cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal- d’Anglade, 1993; González & Martelli, 2003; unpublished own data) as crimson diamonds. 95% ellipses are drawn...... 112 Figure 4.62: Distal epiphysis of femur (ADVP 3.1) from the “bear three” of Algar do Vale da Pena in A) distal B) lateral views...... 112 Figure 4.63: Anteroposterior diameter of the distal epiphysis of the femur (ADVP 3.1) and the same measurement of recent and fossil brown bears from NW Spain (García Vázquez, 2015)...... 113 Figure 4.64: Vertebral centrum (ADVP 3.2) from the “bear three” of Algar do Vale da Pena in A) dorsal and B) anterior views...... 114 Figure 4.65: Proximal epiphysis of second right metatarsal (ADVP 4.1) from the “bear four” of Algar do Vale da Pena in A) Interior; B) Ventral; C) Exterior and D) Dorsal...... 114 Figure 4.66: Measurements for the second metatarsal according to the system of Tsoukala & Grandal- d’Anglade (2002).Taken measurements are marked with a circle...... 115 Figure 4.67: Anteroposterior vs transversal diameter of the proximal epiphysis (upper) and diaphysis (bottom) of the second metatarsal, ADVP 4.1 (blue square) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015)...... 116 Figure 4.68: Proximal epiphysis of fifth right metatarsal (ADVP 4.2) from the “bear four” of Algar do Vale da Pena in A) Interior; B) Ventral; C) Exterior and D) Dorsal views...... 116 Figure 4.69: Measurements for the fifth metatarsal according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurements are marked with a circle...... 117 Figure 4.70: Anteroposterior proximal diameter vs transversal diameter of diaphysis of the fifth metatarsal, ADVP 4.2 (blue square) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015)...... 117 Figure 4.71: Claw marks in the walls of the Algar do Vale da Pena. Some of them are today partially covered by calcareous crusts. Scale bar 7 cm...... 118 Figure 5.1: Location of the site of Santa Margarida. Figure by Cátia Ribeiro. Credit of the maps, google maps...... 123 Figure 5.2: Maximum concentration point of fossil bearing blocks in the wall of the site as it was discovered in 2017. Note the difference of color between the bone bearing breccified blocks of cemented terra rossa and the grey tone of the calcareous rocks. Photo by Octávio Mateus...... 124 Figure 5.3: Location of Santa Margarida site. In blue the calcareous Picavessa Formation. Modified by Cátia Ribeiro from the geologic map of Algarve by LNEG...... 124 Figure 5.4: Left; comparison of freshly cut cemented terra rossa block and the underlying grayish calcareous rock of the Picavessa formation. Right; calcareous crust covering the “terra rossa”..... 125 Figure 5.5: A) Abundant microfossils in rich block; B) Rabbit humerus in situ; C) and D) larger long bones protruding from block...... 125

xix

Figure 5.6: Upper left, possible leporid phalanx in dorsolateral view; upper right, arvicoline mandible in lingual view; bottom left, shrew mandible in labial view; bottom right, lizard mandible in labial view...... 126 Figure 5.7: Left, consolidated arvicoline mandible with PARALOID B72™; center, extraction of the mandible using a air-scribe and right the extracted mandible...... 126 Figure 5.8: Right, rodent teeth mounted in the plastic box and other fossils stored in similar settings (Scale 7cm); left, the binocular lenses used to take the pictures (Photo by Alexadre Guillaume)...... 127 Figure 5.9: Workflow followed in the preparation of the Santa Margarida material (Portugal) in the Museum of Lourinhã...... 128 Figure 5.10: Amount of processed sediment from Santa Margarida (Portugal) at Museum of Lourinhã...... 129 Figure 5.11: Paralaoma servilis from Santa Margarida in A) lateral, B) ventral and C) dorsal views...... 130 Figure 5.12: Bisected shell of Pomatia elegans from Santa Margarida...... 130 Figure 5.13: Lacertid mandible from Santa Margarida (Portugal) in lingual view...... 131 Figure 5.14: Lacertid mandible from Santa Margarida (Portugal) in A) lingual and B) oclusal views...... 131 Figure 5.15: Unicusp teeth ascribed to a bat from Santa Margarida (Portugal) in meso-lingual view (Image by Cátia Ribeiro)...... 132 Figure 5.16: Right upper second bat molar from Santa Margarida (Portugal) in oclusal view. The rectangle marks the w shaped labial ridge and the circle the lingually recurved hypocone shelf and the strong postprotocrista. Metastyle/paracone (1,3 mm) and metastyle/lingual (2 mm) lengths are marked. Drawing by Pavlos Piskoulis...... 132 Figure 5.17: Soricidae upper incisors recovered from Santa Margarida (Portugal) in labial view. A) Sorex, B) and C) Crocidura...... 133 Figure 5.18: Broken second upper molar of Crocidura recovered from Santa Margarida (Portugal) in oclusal view. Labial part to the right...... 133 Figure 5.19: Fragment of Crocidura mandible (bearing P4, M1 and M2) recovered from Santa Margarida (Portugal) in A) oclusal, B) lingual and C) labial views...... 134 Figure 5.20: Tip of a rodent incisor from Santa Margarida (Portugal)...... 135 Figure 5.21: A) third lower molar and B) first upper molar of Apodemus cf. sylvaticus from Santa Margarida (Portugal) in oclusal view...... 135 Figure 5.22: Left upper first or second molar of Eliomys quercinus from Santa Margarida (Portugal) in oclusal view...... 136 Figure 5.23: Left upper first molar of Allocricetus bursae from Santa Margarida (Portugal) in oclusal view...... 136 Figure 5.24: Nomenclature and measurements for an arvicoline first lower tooth according to López- García et al., (2015). ABBREVIATIONS: a: length of the anteroconid complex; ACC, anteroconid complex; AC, anterior cap; BRA, buccal reentrant angle; BSA, buccal salient angle; c: width of the opening of triangles T4-T5; L, length; La, mean width of T4; Li, mean width of T5; LRA, lingual re-

xx entrant angle; LSA, lingual salient angle; PL, posterior lobe; TTC, trigonidtalonid complex, T1-T7, triangles 1-7; W, width...... 137 Figure 5.25: Right lower first molar of Victoriamys chalinei from Santa Margarida (Portugal) in oclusal view...... 137 Figure 5.26: Left lower first molar of Iberomys huescarensis from Santa Margarida (Portugal) in oclusal view...... 138 Figure 5.27: Right and left (respectively) lower first molars of Iberomys brecciensis from Santa Margarida (Portugal) in oclusal view...... 138 Figure 5.28: Left lower first molar (without the posterior lobe) of Iberomys cabrerae from Santa Margarida (Portugal) in oclusal view...... 139 Figure 5.29: A/L x100 vs C/W x100 ratios for a collection of arvicoline teeth. Victoriamys chalinei from El Chaparral locality appear as pink circles. Allophaiomys lavocati from El Chaparral appear as blue pluses symbols. Iberomys huescarensis, from a collection of Iberian and Italian sites appear as black dots. Data from López-García et al., (2012, 2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle...... 140 Figure 5.30: A/L x100 vs La/Li x100 ratios for a collection of arvicoline teeth. Victoriamys chalinei from El Chaparral locality appear as red inverted triangles. Allophaiomys lavocati from El Chaparral appear as blue squares. Iberomys huescarensis, from a collection of Iberian and Italian sites appear as black dots. Data from López-García et al., (2012, 2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle...... 140 Figure 5.31: A/L x100 vs La/Li x100 ratios for Iberomys brecciensis (green squares) and I. chalinei (gold inverted triangles) from a collection of Iberian and Italian sites. Data from López-García et al., (2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle...... 141 Figure 5.32: Left distal Oryctolagus cuniculus humerus from Santa Margarida (Portugal) in A) Anterior; B) Exterior; C) Posterior and D) Medial views...... 142 Figure 5.33: Diverse postcranial elements of microvertebrates recovered from the sample of Santa Margarida. A) Distal femur in posterolateral view, B) Partial distal humerus? in anterior view and C) Distal radius in lateral view...... 142 Figure 5.34: Evolution of the first lower molar of the genus Iberomys with examples from the site of Santa Margarida (Portugal)...... 144 Figure 5.35: The two possible stratigraphy scenarios that explain the faunal assemblage recovered from the ex situ blocks from Santa Margarida (Portugal). Upper: Fossils from different but similar layers that became mixed during preparation. Lower: Reworked fossils situated in the same layer as the younger ones...... 145

xxi

INDEX OF TABLES

Table 1.1: Presence of different species of Microtus/Iberomys in different Quaternary sites in Portugal, data from Antunes et al., 1986b, Antunes et al., 1989; Póvoas et al., 1992; Jeannet, 2000; Moreno-García & Pimenta, 2002; Bicho et al., 2003; Valente & Haws, 2006; Pimenta, 2014, López-García et al., 2018)...... 14 Table 2.1: Localities of the Quaternary of Portugal...... 23 Figure 2.1: Satellital picture showing the location of 30 sites with Quaternary fossil vertebrates in Portugal, mentioned in the table 2.1. In black, the localities situated in the Lusitanian basin and in red the ones situated in the Algarve basin. Credit Google Earth. .... 25 Table 2.2: Amphibians of the Quaternary of Portugal...... 26 Table 2.3: Reptiles of the Quaternary of Portugal...... 26 Table 2.4: Birds of the Quaternary of Portugal...... 27 Table 2.5: Mammals of the Quaternary of Portugal...... 31 Table 2.6: Resume table showing the number, percentage and status of extant and fossil species of the Quaternary of Portugal...... 36 Table 2.7: Table showing the tetrapods and localities known outside the cave deposits in the Quaternary of Portugal...... 38 Table 2.8: Resume table showing the number, percentage and status of extant and fossil species in the Quaternary of Portugal, excluding the cave deposits...... 40 Table 3.1: Measurements of the phalanx of the proboscidean of Santo Antão do Tojal. “Thickness” refers to anteroposterior diameter of the bone and “width” to the mediolateral diameter of it. All the measurements are in millimeters...... 48 Table 3.2: Measurements of 48 individual femurs of Palaeoloxodon antiquus and Mammuthus primigenius taken from the literature. The specimen of Santo Antão do Tojal is marked. All measurements are in millimeters...... 49 Table 3.3: Measurements of 30 individual tibias of Palaeoloxodon antiquus and Mammuthus primigenius taken from the literature. The specimen of Santo Antão do Tojal is marked. In order to fit the table, abbreviations have been used: ML, maximum length; AL, articular length; PW, proximal width; PT, proximal thickness; PAW, proximal articular width; PAT, proximal articular thickness; MDW, minimum diaphysis width; MDT, minimum diaphysis thickness; MDC, minimum diaphysis circumference; DW, distal width; DT, distal thickness; DAW, distal articular width; DAT, Distal articular thickness. All measurements are in millimeters...... 51 Table 3.4: Maximum length vs minimum diaphysis circumference ratios of femur and tibia, for selected Palaeoloxodon antiquus specimens. All measurements in mm.”E” stands for “estimated”...... 53 Table 3.5: Femur vs tibia lengths for selected Palaeoloxodon antiquus specimens. All measurements in mm.”E” stands for “estimated”...... 53 Table 3.6: Shoulder height and body masses of different Palaeoloxodon antiquus specimens. The specimens below to the Santo Antão do Tojal one had not been calculated using the same methods as presented. Possible males are marked in blue...... 54 Table 3.7: Number of individual measurements taken for proboscidean femurs and tibias of the considered sample. Percentage of specimens where the measurement has been taken compared to the total is also presented. The measurements that have been taken in 50% or more of the specimens are marked...... 55

xxii

Table 3.8: Mean for selected measurements in the hind limb of Palaeoloxodon antiquus and Mammuthus primigenius, with percentage of difference between the species for each measurement and bone. All measurements are given in millimeters...... 56 Table 4.1: Measurements from the right scapholunate bone (ADVP 1.3) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to a set of measurements of other brown bears (García Vázquez, 2015) and cave bears (unpublished own data from Cova da Ceza population). Measurements are in mm according to the system of Tsoukala & Grandal- d’Anglade (2002)...... 82 Table 4.2: Length of the right hamate bone (ADVP 1.4) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the length of the same bone of other brown bears (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 85 Table 4.3: Length vs distal transverse diameter of the second phalanx (ADVP 1.6) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the length of the same bone of a recent brown bear (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 87 Table 4.4: Proximal transverse diameter and proximal anteroposterior diameter of the third metatarsal (ADVP 1.7) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to same measurements of fossil brown bears from NW Spain (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 89 Table 4.5: Measurements from the radio (ADVP 1.8) from “bear number one” of Algar do Vale da Pena compared to same ratio of fossil brown bears from NW Spain (García Vázquez, 2015) as well as some of Portuguese caves. Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 91 Table 4.6: Minimum anteroposterior diameter of the ulnar distal diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 93 Table 4.7: Measurements from the partial hemimandible (ADVP 2.1) from “bear number two” (Ursus arctos) of Algar do Vale da Pena compared to same measurements of fossil brown bears from NW Spain (García Vázquez, 2015) as well as some cave bears from the same region (Unpublished own data from Cova de A Ceza). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 96 Table 4.8: Measurements from the fourth upper premolar (ADVP 2.6) from “bear number two” of Algar do Vale da Pena compared to broad an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 101 Table 4.9: Length and breadth from the first upper molar (ADVP 2.7) from “bear number two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from Liñares cave, Galicia, NW Spain (González & Martelli, 2003). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 104 Table 4.10: Breadth and length of the second lower molar (ADVP 2.8) from the “bear two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza cave,

xxiii

Galicia, NW Spain (Unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 106 Table 4.11: Breadth and length of the third lower molar (ADVP 2.9) from the “bear two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal-d’Anglade, 1993; González & Martelli, 2003; unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 109 Table 4.12: Height and diameter of the third phalanx (ADVP 2.10) from the “bear two” of Algar do Vale da Pena compared to the height and diameter of the same phalanx of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal-d’Anglade, 1993; González & Martelli, 2003; unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 111 Table 4.13: Anteroposterior diameter of the distal epiphysis of the femur (ADVP 3.1) from the “bear three” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 113 Table 4.14: Measurements of the second metatarsal (ADVP 4.1) from the “bear four” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 115 Table 4.15: Measurements of the fifth metatarsal (ADVP 4.2) from the “bear four” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002)...... 117 Table 4.16: Nitrogen and Carbon percentages in the samples of ADVP 1.10 and ADVP 3.4 from Algar do Vale da Pena (Portugal)...... 122 Table 5.1: Results of the tests for determining the product to be use in the chemical preparation of Santa Margarida samples...... 127 Table 5.2: Number of specimens from broad groups recovered from the processed sediment of Santa Margarida...... 129 Table 5.3: Measurements and ratios of the most complete lacertid mandible fragment picked from the disaggregated sediments of Santa Margarida, Portugal, following the methods of Blain et al., 2007: h, is height of the teeth; d, its diameter and a, the height of the teeth that protrudes over the dental crest. All measurements are in mm...... 131 Table 5.4: Classification of the different Soricidae remains recovered from Santa Margarida (Portugal)...... 133 Table 5.5: Classification of the different rodent dental remains recovered from Santa Margarida (Portugal)...... 134 Table 5.6: Measurements and ratios of the arvicoline first lower molars from Santa Margarida ...... 139

xxiv

LIST OF ABBREVIATIONS

Institutions

EAVP: European Association of Vertebrate Paleontology. MG: Geological Museum of Lisbon. FCT-UNL: Facultade de Ciência e tecnología da Universidade Nova de Lisboa. LNEG: Laboratório Nacional de Energía e Geología. ML: Museu da Lourinhã. GPS: Grupo de Proteção Sicó. GEM: Grupo de espeleologia e Montanhismo.

Other

PC: Principal Components. YBP: Years Before Present.

xxv

xxvi

1.INTRODUCTION

1.1 History of the study of Quaternary fossil vertebrates from continental Portugal

This section has been partially published in an abstract at meeting of EAVP 2018 (Estraviz-López & Mateus, 2018b). The study of the remains of Quaternary fossil faunal assemblages cannot be understood (mostly in their early years) without its close relation with archeological works and the more general geologic survey. The discovery of tools first and then human remains (nowadays we now that they were in fact younger) in the terraces of the Somme River in France, related with the remains of extinct fauna like rhinoceroses and elephants in the beginning of the second half of the XIX century, sparked the interest of the scientific community about the antiquity of humanity and the search of new bone material and artifacts in Pleistocene deposits. Also, other two factors in this period made the interest in the study of the past rise in the scientific community as never before: The publication of “The origin of species” by Charles Darwin and the outbreak of the nationalist sentiment that made the cult peoples of Europe to search for clues and material proof in the past, even competing with other nations and scientists (Cardoso, 2015a).

In Portugal, the scientific study of the Quaternary started thanks to the work of the members of the 2nd Geological Commission of Portugal in 1857; mostly Carlos Ribeiro and Nery Delgado, whose work proved that it was possible to know about the distant past (well before the written texts, oral traditions and collective memory) just thanks to the study of material remains (being them bones or tools) coming from alluvial and cave deposits (Cardoso, 2015a).

Carlos Ribeiro (1813-1884) (Figure1.1) was the father of the prehistoric archeology (and therefore the Quaternary paleontology) in Portugal. From humble origins he developed a career in the army as military engineer that he left in 1847, being the leader of the Geological Commission of Portugal with Pereira da Costa (more focused to collection management than fieldwork) since 1857 until his death (Daveau, 2000). He published a surprisingly accurate description of the Quaternary deposits, that in that moment also included some of Neogene origin, in the basins of Tagus and Sado rivers that he never finished (a second part of that work was never published) (Cardoso, 2015b) and several other works including descriptions of silex and quartzite tools from the terrains of those rivers (Ribeiro, 1871). He raised the question about how old the humanity could be and pointed towards the possibility that some tools that he collected could be from Tertiary origin (Cardoso, 2000) . He also carried fieldwork, excavating the prehistoric settlement of Leceias in 1878 and the cave of Ponte da Laje in 1879 among other diggings (Cardoso, 2013), probably searching for material to show to other fellow scientist in the important IX session of the International Congress of Anthropology and Prehistoric Archaeology that was carried in Lisbon in September of 1880; the culmination of 25 years of his work.

Figure 1.1: Bust of Carlos Ribeiro at the Geological Museum, Lisbon (Photo by Octávio Mateus).

Joaquim Felipe Nery Delgado (1835-1908) (Figure 1.2) was also a major figure in the study of the Quaternary in Portugal. He graduated as military engineer in 1855 just in time to incorporate as adjunct to the Director of the Geological Comission of Portugal in 1857 that he later lead since the death of Carlos Ribeiro in 1884 to his own death in 1908, with more than 51 years of study and research in the areas of geologic cartography and prehistoric archeology on his back. As geologist his major interest was the study of the Paleozoic rocks in Portugal, mostly in Alentejo as well as the nature of the enigmatic marks of “bilobites” (Cruziana) (Cardoso, 2008a).

Figure 1.2: Picture of Nery Delgado at exposition in the Geological Museum, Lisbon.

In 1865 and 1866 he undergone the first scientific excavation of a cave occupied by prehistoric humans, Casa da Moura, in the municipality of Óbidos and Cova da Moura, Lourinhã and Bombarral municipalities. He later published his findings, along with other excavations in nearby caves in his work “Da Existência do Homem no Nosso Solo em Tempos Mui Remotos Provada pelo Estudo das Cavernas. Notícia ácerca das Grutas da Cesareda” published in 1867 that included a human skull, which being published before the Cro Magnon man would be the first Paleolithic anatomically modern human remains known (Delgado, 1867; Zilhão, 1993). The same year of 1865 he started some preliminary diggings in the Furninha cave in Peniche, but it was not until 1879 and 1880 when he undertook this pivotal excavation in the history of the Portuguese prehistoric archeology and Quaternary paleontology (Delgado, 1884). The methodology of this excavation is even nowadays exemplar, with each piece of bone or tool identified not only with the site but with the layer from where it came and with its coordinates in a grid system (Figure 1.3). So Nery Delgado was a pioneer of the 3D location of the archeological artifacts (Cardoso, 2015a).

Figure 1.3: Hyena coprolite from Furninha cave recovered by Nery Delgado (MG). Note the glued identification.Geological Museum (Lisbon)

Both Nery Delgado and Carlos Ribeiro were respected scientists in his time, being his work known across Europe, proof of that is the correspondence that they kept with other fellow foreign scientists (Cardoso & Avila de Melo, 2001) and the visits that they did to other countries, like for example Nery Delgado to Spain in 1878 (Carneiro, 2007). The quality of their work put Portugal in the frontline of this type of studies in Europe at that time, and all of this crystallized in the celebration of the IX session of the International Congress of Anthropology and Prehistoric Archaeology in Lisbon, 1880. Some new diggings were made to provide with new information and pieces to show in the congress, for example the interesting site of Mealhada where some proboscidean remains were uncovered (Cardoso, 1993).

In the last years of the XIX century and the first of the XX century Nery Delgado focused more in the study of the geology of the Paleozoic of Portugal until his death (Delgado, 1886) but the study of the Quaternary animal remains continued with less intensity. Paul Choffat was another important geologist and paleontologist of those years, that inherited the lead of Nery Delgado over the geological services and overall the study of the geology in Portugal; he uncovered some hippopotamus and proboscidean remains in the calcareous tuffs of Condeixa in 1895 and 1898. Years later, in 1909, he also promoted the diggings in the Pleistocene sites of Algar de João Ramos and Molianos (Cardoso, 1993b).

Paul Choffat (figure 1.4) realized that the materials recovered by Nery Delgado and Carlos Ribeiro deserved an attention that he (being focused on the investigation of the Paleozoic and Mesozoic of Portugal) could not properly give, so thanks to his initiative the classic Portuguese material of sites like Furninha and Fontainhas was studied by Edouard Harlé, from Toulouse (One of the top experts in the Pleistocene of that time). He published a work that would be a major reference for the study of Quaternary paleontology in Portugal during the next 80 years, cataloging the fauna found until that moment and comparing it with other countries of Europe (Harlé, 1910; Cardoso, 1993b).

Figure 1.4: Picture of Paul Choffat at exposition in the Geological Museum, Lisbon.

After this outburst of the early XX century the study of the Quaternary animals in Portugal entered a phase of stasis. Paul Choffat said: “By reasons that we should not comment, although that our faculties are filled with professors of great erudition, there are few of them that are dedicated to the progress of observation sciences”, that was his way of saying that there was not much interest in carrying fieldwork in Portugal during those years (Daveau, 2000). We should wait until the middle of the century and the works of Octávio da Veiga Ferreira and mostly Georges Zbyszewski to restart the findings in the Quaternary of the Portuguese caves and alluvial deposits.

Octávio da Veiga Ferreira (1917-1997) was a mine engineer that incorporated to the Geological Services in 1950. He already had carried some archeological excavations by that time and during his many years of activity he was a major figure in the Iberian archeology. In the area of paleontology, although it was not his main research field, he published 32 papers, four of them about Pleistocene vertebrates and one about Holocene mollusks (Cardoso, 2008 b). One of his major contributions is his study about the rhinoceroses of Portugal (Veiga Ferreira, 1975).

Georges Zbyszewski (1909-1999) (Figure 1.5) was the direct superior of Octávio da Veiga Ferreira; born in Russia, but later nationalized French citizen. He visited Portugal for first time in 1935 searching for clues about the tectonic movements in the Portuguese littoral during the Quaternary for his doctoral thesis. In 1940 he incorporated to the Geological Services and during the next 40 years he covered an immense array of geological topics, writing around 300 scientific publications from the geological cartography of Portugal, to the volcanism in Açores and Madeira, to the Mesozoic reptiles and over all of that; more than 100 publications about Quaternary lithic industries and fauna (Soares de Carvalho & Cardoso, 1999; Teixeira, 1979). During the years “Zby” (as he was called by friends and family) reunited a lot of information about the recent fluvial terraces of the Tagus River and its tributaries in the work “Le Quaternaire du Portugal” published in 1957. A lot of publications were made from mammal material coming from those terraces, either Miocene or Pleistocene, for example some remains like phalanges and tusks of Paleoloxodon antiquus (Falconer & Cautley, 1847) from that area and the classic site of Mealhada. Is also remarkable the different excavations and publications that he did with the material of a lot of different Pleistocene caves like: Almonda (1941 and 1947), lapa da Bogalheira (1941), the caves of Maceira and Vimeiro (1949), the cave of Ponte da Laje (excavated before by Carlos Ribeiro) (1957), Gruta das Salemas (1962), Gruta da Columbeira (1963), Alcobertas (1971), and lapa do Bugio the same year. Beside all of this he also studied the Pleistocene faunas coming from Algoz, in Algarve (Soares de Carvalho & Cardoso, 1999; Teixeira, 1979). All this work would have been probably impossible without the hard work of dozens of unrecognized collectors that were the eyes and ears of not only him but also great scientist before and after him.

Figure 1.5: Georges Zbyszewski (Picture by unknown author).

The work of Veiga Ferreira and Zbyszewski in the Geological services of Portugal served well to kick start the next phase of the study of the Quaternary faunas in Portugal, with the lead of Miguel Telles Antunes first and João Luis Cardoso later. A good visualization of this is the work that both of them published in 1990 to synthesize their many years of study in this matter with the new studies of the next generation in the decade of 1980 (Zbyszewski & Veiga Ferreira, 1990).

Before continuing it should be noted the work of Torres Pérez-Hidalgo, a Spanish researcher that in 1979 published an extensive study about the bears of Portugal, including measurements of old material recovered even in the XIX century and agreeing with the work of Harlé in that there is no other recognized fossil bear in the country besides the brown bear, Ursus arctos Linnaeus, 1758 (Torres Pérez-Hidalgo, 1979).

Miguel Telles Antunes (1937) (Figure 1.6) is probably until this day the most prolific figure in the History of the Portuguese paleontology. He started his career in the University of Lisbon in 1960 after his doctorate in geology and in 1974 he moved to the Universidade Nova de Lisboa. During his many years in paleontology he covered practically all the topics in the area writing more than 450 scientific articles and directing 9 doctoral theses, from the Miocene vertebrates in the area of Lisbon, to the Eocene of Silveirinha (Baixo Mondego), or the dinosaur faunas in the Lusitanian Basin (https://www.fct.unl.pt/noticias/2016/12/sessao-de-homenagem-ao-professor-doutor-miguel-telles- antunes). His collaboration with Zbyszewski at the end of the decade of 1970 produced some new insights into the Neogene of the Tagus Basin and during the decade of 1980 he carried several studies in Pleistocene material: The faunas of Morgadinho and Goldra in Algarve, the review of the Algoz fauna (also from Algarve) and the Mealhada material (that lead to the identification of Homotherium latidens Owen, 1846 in Portugal) and also the studies carried out in the cave contexts of Figueira Brava (where he impulse excavations), Caldeirão cave (where he recognized Castor fiber Linnaeus, 1758 for first time in Portugal) (Antunes, 1989a) or Pedreira das Salemas (Cardoso, 1993a). Over all of this work, the impulse that Antunes gave to the study of Pleistocene faunas through the former CEPUNL (Centro de Estratigrafía e Paleobiología da Universidade Nova de Lisboa) is everlasting even today, with numerous radiometric datings and training of new personal.

Figure 1.6: Miguel Telles Antunes in 2009 (Photo by Natacha Cardoso).

More or less at the same time during the decade of 1980 and the beginning of the 1990, two personalities started producing scientific investigation at high level. João Zilhão (1957) now professor of the University of Barcelona is an archaeologist, therefore his core area is situated more in the paleoanthropology, but his work in the Almonda cave system (Zilhão et al., 1991) and other caves in the Estremadura like Caldeirão cave produced a wide array of Pleistocene faunal material (Zilhão, 1987). This material was subject of studies the following years.

The other personality is João Luis Cardoso (1956) (Figure 1.7) the most well known researcher in the study of Quaternary faunas from Portugal. Professor of the Universidade Aberta de Lisboa has written more than 550 publications, including books, chapters of books and scientific articles; a lot of them about the paleontology and archeozoology, but also in archeology and geology (https://www.cm- obidos.pt/jornadas-arqueologia). In the second half of the decade of 1980 and the first of the decade of 1990 he worked in the CEPUNL publishing together with Antunes or in solitary plenty of scientific articles such the first reference of several species in the Pleistocene of Portugal, like Hippopotamus incognitus Faure, 1984 in Portugal from the deposits of Mealhada (Antunes et al., 1988), Cuon alpinus europaeus Bourguignat, 1868 (Cardoso, 1992), Panthera spelaea Goldfuss, 1810 (Antunes & Cardoso, 1986), Crocuta crocuta intermedia (M. de Serres, 1828) (Cardoso, 1994), Equus hydruntinus Regalia, 1905 (Cardoso, 1995), Dicerorhinus hemitoechus (Falconer, 1868) (Cardoso, 1990), Mammuthus primigenius Blumebach, 1799 (Antunes & Cardoso, 1992) and Rupicarpa rupicapra (Linnaeus, 1758) (Cardoso & Antunes, 1989). Also he has described a new subspecies of horse from the Pleistocene of Portugal, Equus ferus antunesi (Cardoso & Eisenmann, 1989).

Figure 1.7: João Luis Cardoso in 2017 (Photo by Octávio Mateus).

But his main contribution on this period is his doctoral thesis defended in October, 1992 and published in 1993 by the municipality of Oeiras titled “Contribuição para o conhecimento dos grandes mamíferos do Plistocénico Superior de Portugal” (Figure 1.8); a book of more than 500 pages that is the most comprehensive and at the same time detailed study ever made in Portugal about the Quaternary big mammal fauna of the country, covering remains of 26 taxa from 28 localities across all Portugal (Cardoso, 1993a).

Figure 1.8: “Contribuição para o conhecimento dos grandes mamíferos do Plistocénico Superior de Portugal”

During the next 20 years João Luis Cardoso has continued to develop some more work in that investigation line, doing a catalog of Iberian equids (Morales-Muñiz et al., 1998), describing a mammoth lamella near Lisbon (Cardoso & Regala, 2002) or a particularly strange leopard in the Algar da Manga Larga (Cardoso & Regala, 2006) plus some important syntheses about the study of Quaternary mammals and archeozoology (Cardoso, 1993 b; Cardoso, 2002) or the life and work of important scientist that we have covered before (Cardoso & Avila de Melo, 2001; Cardoso, 2008a; Cardoso, 2015b). Beside his labor, since the year 2000 other works have started to fill the gaps in our knowledge of the fauna of that time period that Cardoso already noted: the study of Quaternary bird fossils and microfauna like rodents, amphibians and reptiles (Cardoso, 2002).

About the Quaternary birds, two works have vastly augmented our knowledge during the last years. In 2010 a comprehensive thesis about the birds of this age in Portugal was presented, in which bird remains of more than 13 sites were studied for first time or reviewed (Figueiredo, 2010) and also recently another work was published about avian remains found in archeological contexts across the country (Pimenta et al., 2015) making our understanding of the Quaternary avifauna of Portugal improve considerably. Meanwhile, on amphibians and reptiles, only few publications about Pleistocene material from Portugal have been made (Jiménez Fuentes et al., 1998; Crespo, 2002). The micromammals (and even more the fishes) are still poorly studied compared to Spain but there are some works published (Antunes et al., 1986a; Antunes et al., 1989), mostly related to archeological contexts (Monteiro, 2016; Póvoas et al., 1992) although new works aim to a change this (López- García et al., 2018).

1.2 Vertebrates in the Quaternary of Portugal

Quaternary fishes, amphibians, reptiles and micromammals are almost always treated in an archaezoological way, not to study the small vertebrate’s ecology or paleodistribution per se (Dean & Carvalho, 2011; Nabais & Rodrigues, 2017; Pereira et al., 2017).

Studied fish remains from the Pleistocene and Holocene of Portugal overwhelmingly come from archeological contexts, mostly Holocene (Pereira et al., 2017) although some are older, Pleistocene, like in Caldeirão cave (Zilhão, 1997). There are marine fishes recognized (although rare) in the after mentioned cave and in Lapa do Suão or Lapedo. They appear in bigger numbers in Lapa do Picareiro and Lapa dos Coelhos (Bicho & Haws, 2008). Fish remains recovered in geological context are the pharyngeal teeth of cyprinids from the Pleistocene site of Morgadinho, Algarve (Antunes et al., 1986a).

The Quaternary amphibians have been recognized in numerous recent archeological sites of Holocene age; for example they have been found in the Mesolithic shell midden of Cabeço de Morros, although not specific identification is given (Detry, 2008). According to Hockett & Haws (2009) amphibian remains have been found in three out of the eight archeological sites treated in that work: Figueira Brava, Picareiro and Lagar Velho. In the Lagar Velho site four amphibian species were recognized, being two of them unidentified anurans and the other two cf. Triturus marmoratus (Latreille, 1800) and cf. Salamandra salamandra (Linnaeus, 1758) (Moreno-García & Pimenta, 2002). In Picareiro cave two species were recognized: Bufo bufo (Linnaeus, 1758) and Pelobates cultripes Cuvier, 1829 (Hockett & Haws, 2009). The cave of Figueira Brava has yielded remains of Salamandra salamandra (Figure 1.9), Pelobates cultripes and one indeterminate anuran (Crespo et al., 2000). The largest amphibian assemblage of this time period from Portugal comes from Guia, Algarve, a site in geologic seasonal context with no archaeological remains known and five species of amphibians being identified: Pelobates cultripes (Almost ¾ of all fossil remains), Pleurodeles waltl (Michahelles, 1830), Bufo bufo, Epidalea calamita Laurenti, 1758 and Pelophylax perezi (Seoane, 1885) (Antunes et al., 1989). From the Gruta da Aroeira (Almonda System) there are undetermined anurans and salamandrines as well as one species of Pelodytes Bonaparte, 1838 and Bufo spinosus (Daudin, 1803) (López-García et al., 2018). Also there are some undetermined amphibian remains come from the Morgadinho site in Algarve (Antunes et al., 1986a).

Figure 1.9: Vertebra of Salamandra salamandra from Figueira Brava cave in dorsal view; scale bar 1mm. Catalog number of the specimen not provided in the original (Crespo et al., 2000).

Reptiles have been recognized in at least 12 sites of the Pleistocene of Portugal: Gruta da Aroeira (Almonda System), Mealhada, Furninha, Caldeirão, Almonda, Gruta Nova da Columbeira, Lagar Velho, Portocovo, Figueira Brava, Escoural, Foz do Enxarrique and Gruta de Ibn Amar. All of the reptile species recognized still exist in Portugal nowadays except one. Most of the remains found are Testudines Linnaeus, 1758 probably because this group of animals has a lot of sturdy bony plates that favor conservation, compared with the relative fragile bones of other reptiles, but there are also some references of Squamata Oppel, 1811 like Lacertilia Gunther, 1867 (Figure 1.10), Amphisbaenia Gray, 1844 and Serpentes Linnaeus, 1758 (Crespo, 2002; Morales Pérez & Serra, 2009). Timon lepidus Daudin, 1802 is the most common lacertid, appearing in Figueira Brava and Guia. Psammodromus algirus (Linnaeus, 1758) is referred in Figueira Brava, although the remains probably should be ascribed to Psammodromus manuelae Busack et al., 2006 (Busack et al., 2006). From the same locality there is a reference to an undetermined Podarcis Wagler, 1830. In Guia (Algarve) appears Acanthodactylus erythrurus (Schinz, 1833) and from Lagar Velho come other two undetermined lacertids (Antunes et al., 1989; Crespo et al., 2000; Moreno-García & Pimenta, 2002). Undetermined lacertids appear as well at Gruta da Aroeira (Almonda System) (López-García et al., 2018).

Figure 1.10: Mandible from an indeterminate lacertid recovered at the site of Santa Margarida (Portugal). FCT-UNL.

The Serpentes of the Quaternary of Portugal are mostly known from the cave of Figueira Brava, where at least two species of Colubridae Oppel, 1811 are recognized, being them probably Rinechis scalaris (Schinz, 1822) and/or Hemorrhois hippocrepis (Linnaeus, 1758), but probably more are present (like Coronella girondica Daudin, 1803 or Vipera lastei Bosca, 1878). In the Gruta da Aroeira (Almonda System), there are references to cf. Coronella girondica and Natrix cf. maura (López-García et al., 2018). In Guia there is a reference to the presence of Natrix Laurenti, 1758. Also two undetermined Serpentes are present at Lagar Velho (Antunes et al., 1989, Crespo et al., 2000; Crespo, 2002; Moreno-García & Pimenta, 2002). The Amphisbaenia group is represented by Blanus cinereus Vandelli, 1797 of Figueira Brava, where it is known by a right dentary (Figure 1.11) and vertebrae (Crespo et al., 2000).

Figure 1.11: Right dentary of Blanus cinereus from Figueira Brava in lingual view, scale bar 1 mm. Catalog number of the specimen not provided in the original (Crespo et al., 2000).

Testudines are represented in Portugal by three species during this time period: Testudo hermanni Gmelin, 1789, Mauremys leprosa (Schweiger, 1812) and Emys orbicularis (Linnaeus, 1758). Testudo hermanni, nowadays extinct in Portugal, is the most common one appearing in the literature under different nomenclatures like Agrionemys in the sites of Mealhada, Furninha, Figueira Brava (Figure 1.12), Gruta Nova da Columbeira, Caldeirão, Escoural and the Gruta of Ibn Amar; being the later one probably of Holocene age. Mauremys leprosa appears in Mealhada (being this the oldest appearance of the species in Portugal), Caldeirão, Porto Covo (Cascais) and probably Gruta Nova da Columbeira. Finally, Emys orbicularis appears in Figueira Brava and probably Gruta Nova da Columbeira. It should be noted that Gruta Nova da Columbeira has yielded more than 80 specimens of turtles, being this the most important locality for the group made (Jiménez Fuentes et al., 1998; Lapparent de Broin & Antunes, 2000; Morales Pérez & Serra, 2009).

Figure 1.12: Hyoplastron of Testudo hermanni from Figueira Brava in ventral view; scale bar 1 cm. Catalog number of the specimen not provided in the original (Lapparent de Broin & Antunes, 2000).

The avifauna of the Pleistocene of Portugal is known in around 15 different sites: Gruta Nova da Columbeira, Gruta das Salemas, Gruta das Fontainhas, Gruta da Furninha, Gruta do Pego do Diabo, Gruta da Casa da Moura, Gruta do Caldeirão, Lapa da Rainha, Lagar Velho, Lapa do Picareiro, Galería Pesada/Almonda, Lapa do Suão, Escoural, Gruta da Figueira Brava and Vale Boi (Figueiredo, 2010; Manne et al., 2012; Pimenta et al., 2015). Practically all of them (if not all) are situated in archaeological contexts. More than 78 species have been recognized in diverse works, the immense majority of them are already present in the country (Figure 1.13) but five species are extremely rare in Portugal, four of them no longer inhabit the territory and three are extinct (Figueiredo, 2010; Pimenta et al., 2015). It is also worth mentioning that in the Portuguese islands of Açores and Madeira there are more examples of extinct birds, but given their insular nature those will not be discussed here (Alcover et al., 2015).

Figure 1.13: Right femur of Bubo bubo (Linnaeus, 1758) from Furninha cave in anterior view. Geological Museum, Lisbon.

Between the species that have been discovered in the Quaternary fossil record of Portugal, there are some (most of them being of cold climate) today considered “rarities” and whose sights are homologated by the “Portuguese Committee of Rarities” belonging to the SPEA, “Sociedade Portuguesa para o Estudo das Aves”. All of the following homologated cites come from them: Falco rusticolus Linnaeus, 1758; is known by one left tarsometatarsus from Gruta da Moura that have the characteristics of the genus Falco but is bigger than any other species. The gyrfalcon today lives in the high Arctic and has only one homologated sight in Portugal in 1991 (Poole & Boag, 1988; Figueiredo, 2010). Anser albifrons (Scopoli, 1767) is known by one carpometacarpus from Pego do Diabo. This species today nest in the Arctic and winters in Northern Europe. It has ten homologated sights in Portugal (Ely et al., 2005; Figueiredo, 2010). Somateria mollisima (Linnaeus, 1758) is recognized in Furninha cave, being particularly abundant in this site (the second most common duck). Although this fact, today it only lives in Northern Europe and has five homologated sights in Portugal (Swennen, et al., 1989; Brugal et al., 2012). Mergus merganser Linnaeus, 1758 has been recognized in Escoural cave, but nowadays is rare in Iberian Peninsula and has only been seen eight in Portugal (Pimenta et al., 2015). Cygnus olor (Gmelin, 1789) is known by one right humerus from Gruta da Furninha (Figure 1.14) and in Galería Pesada, Almonda cave system. It lives today mostly in Northern Europe and has been seen at least 20 times in Portugal (Figueiredo, 2010; Pimenta et al., 2015).

Figure 1.14: Right humerus of Cygnus olor from Furninha cave in anterior view. Geological Museum, Lisbon.

Between the species that no longer live in Portugal (most of them being also linked to cold, like the previous group) we have:

Pyrrhocorax graculus (Linnaeus, 1766) is probably the most common species of this category, being it represented by hundreds of bones in Lapa da Rainha, Furninha, Gruta de Casa da Moura, Lapa do Picareiro, Fontainhas, Caldeirão and Gruta Nova da Columbeira (Figure 1.12). Although its widespread distribution an abundance, it no longer appears in the mountains of Portugal according to the SPEA, probably due to climatic reason (Figueiredo, 2010; Smith et al., 2013). Tetrao urogallus Linnaeus, 1758; known by one tarsometatarsus from Lapa da Rainha, that lived in the area of Peneda-Gêres during the second half of the XVIII century and went extinct around 100 years ago in Portugal (Dantas da Gama, 1998; Figueiredo, 2010). Perdix perdix Linnaeus, 1758; is known by several bone remains from Lapa da Rainha and Gruta Nova da Columbeira. Recently it has appeared back in the Northeast of Portugal (Martínez Nistal et al., 1986; Figueiredo, 2010). Lagopus muta (Montin, 1781) has been recognized by one tarsometatarsus from Gruta Nova da Columbeira. Today the closer populations to Portugal are in the Pyrenees (Figueiredo, 2010; Smith et al., 2013). Lagopus lagopus Linnaeus, 1758 was discovered in Portugal in Gruta Nova da Columbeira (two tarsometatarsus) and probably Lapa da Rainha (one femur, one tarsometatarsus). There is also an interesting reference of possible Holocene survival of this species in Portugal until Roman times in Quinta das Longas, but nowadays it is extinct from the Iberian Peninsula as a whole (Figueiredo, 2010; Smith et al., 2013; Pimenta et al., 2015).

Then we have three fossil species of birds found in the Quaternary of Portugal now extinct: Grus primigenia (Milne-Edwards, 1869), Puffinus holeae Walker et al., 1990 and Pinguinus impennis (Linnaeus, 1758). Grus primigenia, the cave crane, was a larger crane than the Grus grus Linnaeus, 1758; almost as big as the Asiatic sarus crane, Grus antigone (Linnaeus, 1758). It inhabited Western Europe during the Pleistocene and part of the Holocene, being known from France, Germany, Mallorca, Ibiza and England, surviving until the Iron Age in that case. It is known in Portugal by one left scapula from Figueira Brava cave. This species was sturdier than the sarus crane and probably had a different way of living. The cause and precise date of its demise remains in mystery (Harrison & Cowles, 1977; Northcote & Mourer-Chauviré, 1985; Northcote & Mourer-Chauviré, 1988; Mourer- Chauviré & Antunes, 2000). Puffinus holeae Walker et al., 1990; the Canary dune shearwater, whose remains have been recovered from different levels of Figueira Brava (Coracoids, humerus, ulnas, etc). This was the first time that the species was recorded outside its nesting grounds in the Canary Islands. The fact that it appears until recent levels in the cave pointed towards that its demise was triggered by the human colonization of the islands that led to the destruction of its colonies and not climate change at the end of Pleistocene, as was later proved (Mourer-Chauviré & Antunes, 2000; Rando & Alcover, 2010). Pinguinus impennis (Linnaeus, 1758), the great auk, has been recognized in two Portuguese sites (besides the island of Porto Santo): In Figueira Brava by two humerus and in Furninha by one humerus (Figure 1.15) that was recovered more than 100 years ago by Nery Delgado. Its demise was caused by human persecution and it went extinct in the XIX century (Mourer-Chauviré & Antunes, 1991; Pimenta et al., 2008).

Figure 1.15: Pinguinus impennis proximal humerus from Furninha in lateral view. Geological Museum, Lisbon.

Finally there is a reference of Corvus antecorax Mourer-Chauviré, 1975 (the ancestor of the crow) in the Almonda Cave and other of Alectoris barbara (Bonnaterre, 1791) in Gruta Nova da Columbeira that became restricted to North Africa after the Pleistocene and have been reintroduced in Portugal, in contrast with the other regionally extinct species that are linked towards a colder environment (Figueiredo, 2010).

The study of Quaternary micromammals in Portugal is less developed than in Spain. Most of these studies are centered on relatively recent archeological contexts (Morales & Rodríguez, 2009), and although other go into the Pleistocene almost always they revolve around the archaezoological question, not usually going deep into the paleoclimatic reconstruction or the past distribution of the species of micromammals (Manne et al., 2006).

The “insectivores” of Portugal during the Quaternary are represented by at least seven species. The hedgehog, Erinaceus europaeus Linnaeus, 1758 is represented mainly by an extraordinary array of 321 bones, belonging to at least 40 individuals from Furninha (Figure 1.16), also it is known by two teeth from Figueira Brava cave, and some more remains from Lagar Velho and Lapa do Picareiro (Mein & Antunes, 2000; Moreno-García & Pimenta, 2002; Bicho et al., 2003; Brugal et al., 2012). Crocidura russula Hermann, 1780 is the most common member of this group, that appears in Goldra, Guia, Figueira Brava, Lapa do Picareiro, Lagar Velho and Bom Santo Cave (Antunes et al., 1986b; Antunes et al. , 1989; Mein & Antunes, 2000; Moreno-García & Pimenta, 2002; Bicho et al., 2003; Pimenta, 2014). Crocidura suaveolens Pallas, 1811 is present in Figueira Brava cave and Lapa do Picareiro (Mein & Antunes, 2000; Bicho et al., 2003). Undetermined species of the genus Crocidura Wagler and Sorex Linnaeus, 1758 are presented in the sites of Gruta da Aroeira (Almonda System) and Lagar Velho (Moreno-García & Pimenta, 2002; López-García et al., 2018). Sorex araneus Linnaeus, 1758 is identified in Goldra (Antunes et al., 1986b). Talpa occidentalis Cabrera, 1907 is known from Figueira Brava cave, Lapa do Picareiro and Bom Santo Cave (Mein & Antunes, 2000; Bicho et al., 2003; Pimenta, 2014), Talpa europea Linnaeus, 1758 is cited in Goldra (Antunes et al., 1986b) and an unidentified Talpa Linnaeus, 1758 species is known from Lagar Velho, although this last two probably belong to the first species. Finally Galemys kormosi (Shereuder, 1940) is the only extinct insectivore that has been found during this time period in Portugal, concretely by one upper third premolar from the site of Morgadinho (Antunes et al., 1986a), its relative Galemys pyrenaicus Geoffroy, 1811 that still exists today in the North of Portugal has been discovered in Lapa do Picareiro (Bicho et al., 2003).

Figure 1.16: Right hemimandible of Erinaceus europaeus from Furninha in lingual view. Geological Museum, Lisbon.

The fossil Chiroptera Blumenbach, 1779 have mostly been studied in detail in Portugal in the Figueira Brava cave, where seven species are recognized: Rinolophus ferrumequinun (Shereber, 1774), Rinolophus hipposideros (Bechstein, 1800), Rinolophus euryale Blasius, 1853; Myotis myotis (Borkhausen, 1797), Myotis blythi (Tomes, 1857), Myotis nattereri (Khul, 1818) and Miniopterus schreibersi (Khul, 1819). From Gruta da Aroeira (Almonda System) five species are cited: Myotis myotis, another undetermined Myotis Borkhausen, 1797; Rinolophus ferrumequinun and another undetermined Rhinolophus Lacepede, 1799 plus possibly Miniopterus schreibersi. Then we have other two species recognized in Lapa do Picareiro: Myotis myotis and Myotis mystacinus Khul, 1817. There is also one indeterminate bat from Bom Santo Cave (Mein & Antunes, 2000; Bicho et al., 2003; Pimenta, 2014, López-García et al., 2018).

The rodents (figure 1.17) are the most common micromammals from the Quaternary of Portugal, with at least 20 species known in eleven localities across Portugal (Morgadinho, Gruta da Aroeira, Guia, Vale Boi, Bom Santo Cave, Figueira Brava, Galería Pesada, Lagar Velho, Lapa do Suão, Lapa do Picareiro and Caldeirão).

Figure 1.17: Rodenia incisor recovered at the site of Santa Margarida (Portugal). FCT-UNL.

The Muridae Illiger, 1811 are represented by at least four species. Apodemus sylvaticus (Linnaeus, 1758) is the most common one, being recognized in Goldra, Guia, Caldeirão, Bom Santo (Figure 1.15), Lagar Velho and Lapa do Picareiro; meanwhile the bigger Apodemus flavicollis (Melchior, 1834) is present in Figueira Brava Cave and Gruta da Aroeira. There are also two other species more closely related to human activity; the house mouse, Mus musculus Linnaeus, 1758 that appears in Guia and Figueira Brava; and the black rat, Rattus rattus Linnaeus, 1758 that is found in Bom Santo Cave and Figueira Brava. Their presence in the latter site is due to the inclusion of more recent sediments in the Pleistocene deposits. Finally there is an indeterminate Muridae in the site of Morgadinho (Antunes et al., 1986a; Antunes et al., 1986b; Antunes et al., 1989; Póvoas et al., 1992; Jeannet, 2000; Moreno- García & Pimenta, 2002; Bicho et al., 2003; Pimenta, 2014; López-García et al., 2018).

Sciuridae Fischer de Waldheim, 1817 is only represented by one species: Sciurus vulgaris Linnaeus, 1758 that appear in Figueira Brava and Lagar Velho (Jeannet, 2000; Moreno-García & Pimenta, 2002). Gliridae Muirhead in Brewster, 1819 is represented by two species: Eliomys quercinus Linnaeus, 1766 that appear in Gruta da Aroeira (Almonda System), Goldra, Guia, Caldeirão, Bom Santo, Figueira Brava and Lapa do Suão and Glis glis (Linnaeus, 1766) that only appear in Lapa do Picareiro. There is also one indeterminate glirid from Lagar Velho (Antunes et al., 1986b; Antunes et al., 1989; Póvoas et al., 1992; Jeannet, 2000; Moreno-García & Pimenta, 2002; Bicho et al., 2003; Valente & Haws, 2006; Pimenta, 2014, López-García et al., 2018).

The Cricetidae Fischer, 1817 is the most common family of rodents in Portugal, being represented by at least fifteen species: Allocricetus bursae Schaub., 1930 is an extinct species of rodent closely related to actual hamsters that has been found in Gruta da Aroeira (Almonda System) and Caldeirão (Póvoas et al., 1992; López-García et al., 2018). Complete teeth of another extinct rodent, a species of Mimomys Forsyth-Major, 1902 has been located in the site of Morgadinho (Antunes et al., 1986 a). Pliomys episcopalis Méhely, 1914 is yet another extinct member of this group cited in the middle Pleistocene sites of Galería Pesada and Gruta da Aroeira, in the Almonda system (Marks et al., 2002a; López-García et al., 2018) . There is also a strange reference of a Myodes Pallas, 1811 from an unknown locality in Atouguia da Baleia (Nehring, 1899). Chionomys nivalis Martins, 1842 appears in Caldeirão and Lapa do Picareiro, but is no longer present in Portugal since it is related to colder conditions (Póvoas et al., 1992; Bicho et al., 2003). At least 3 species of water voles have been recorded in Portugal: Arvicola sapidus Miller, 1908 is the most common one appearing in Caldeirão, Figueira Brava and Lagar Velho. Arvicola amphibious (Linnaeus, 1758) (Formerly known as Arvicola terrestris) and Arvicola cantiana Hinton, 1910, the probable ancestor of A. sapidus, appear in Lapa do Suão and Figueira Brava respectively (Póvoas et al., 1992; Jeannet, 2000; Moreno-García & Pimenta, 2002; Valente & Haws, 2006). The genus Microtus Schrank, 1798 and /or Iberomys Chaline, 1972 is represented by at least 6 species from 9 localities in Portugal that will be treated in the Table 1.1.

Bom Lapa do Lapa do Lagar Gruta da Figueira Brava Caldeirão Guia Goldra TOTAL Santo Suão Picareiro Velho Aroeira Microtus lusitanicus (Gerbe, X X X X X 5 1879) Microtus duodecimcostatus (de Selys-Longchamps, X X X X X X 6

1839) Microtus arvalis (Pallas, X X X 3 1778) Microtus agrestis(Linnaeus, X X 2 1761) Iberomys brecciensis X X X X X 5 (Forsyth Major, 1905) Iberomys cabrerae X X 2 (Thomas, 1906) Microtus sp. X X 2

TOTAL 4 7 3 1 1 2 4 2 1

Several pairs of species of Microtus are difficult to tell apart from each other, being the pair M. lusitanicus/M. duodecimcostatus the most common one. It should be also addressed that the most common rodent taxa today (Mus, Apodemus, Rattus, etc) appear to be under-represented in the fossil record, probably because of their “lack of scientific interest” that have caused them to be neglected sometimes in the literature. To end with rodents, the beaver, Castor fiber also appear in Portugal, namely in Caldeirão cave (Antunes, 1989a).

The lagomorphs are overwhelmingly represented in Portugal by the Oryctolagus cuniculus (Linnaeus, 1758), the European rabbit. It appears in all the major Quaternary sites including hundreds or thousands of bones that are a testimony of its hunt by ancient human populations (Hockett & Bicho, 2000). The hares, Lepus Linnaeus 1758, seem to be way less abundant but are mentioned in Caldeirão and Pego do Diabo (Davis, 2002; Valente, 2004) and some tracks in the eolianites of the South of Portugal are attributed to hares, with a new ichnospecies being described: Leporidichnites malhaoi (Neto de Carvalho, 2010). Also from Algarve, from the old site of Morgadinho (Almost 1 million YBP) comes a Mediterranean pika of the genus Prolagus Pomel, 1853 probably P. calpensis Major, 1905 (Antunes et al., 1986a). In the nearby and probably contemporaneous site of Algoz is cited also Oryctolagus lacosti (Pomel, 1853) (Antunes et al., 1986c).

The only non human primate known in the Pleistocene of Portugal is Macaca sylvanus (Linnaeus, 1758) that has been recognized in the Galería Pesada of the Almonda system during the middle Pleistocene (Marks et al., 2002b).

Now we will start to discuss the big mammals of the Quaternary from Portugal, an area that has been better covered than the microvertebrates or the bird fauna since the XIX century, as we have seen in the previous section. Besides an African elephant (Loxodonta africana Blumenbach, 1797) lamella and other ivory elements brought to Portugal by humans from Africa since the middle Holocene (Schuhmacher et al., 2009) two species of proboscideans have been cited in the Quaternary of Portugal: Palaeoloxodon antiquus and Mammuthus primigenius. The straight tusked elephant was cited already in the XIX century by Paul Choffat, who found a molar (Figure 1.18) in the calcareous tuffs of Condeixa (Choffat, 1895). Most of the remains known until today are dental remains, like the tusks known from Meirinha, Algés and Condeixa or the molars or molar laminae found in Furninha, Almonda, Carregado, Mealhada, the before mentioned Condeixa molar and Foz do Enxarrique; this last one was regarded as the most recent known remain for the species with around 30.000 YBP; and although new radiometric datings (Cunha et al., 2019) have pushed back this date to around 44.000 YBP it is still the most recent known remain for this species. Other bone remains include several fragments of ribs, tibia and humerus from Mealhada, an unciform from Santa Cruz and a set of bones coming from Santo Antão do Tojal (Femur, tibia, phalanx…etc) that will be treated later in this work (Antunes & Cardoso, 1992; Figueiredo, 2012). The Mammuthus primigenius presence is well more cryptic. There is a fragment of femur from Algar de João Ramos that is badly damaged, but its C14 dating of 14000 YBP make its attribution only possible to this species, being it also the most recent proboscidean remain from Portugal (Antunes & Cardoso, 1992). Also two molar lamellas are known, one from Figueira Brava cave and another one that was recovered by a fisherman from the ocean bottom in front of the coast of Oeiras (Cardoso & Regala, 2002). Also there are some tracks of proboscideans conserved in the Pleistocene eolianites from South Portugal, in Pessegueiro Island (Neto de Carvalho, 2010).

Figure 1.18: Molar of Palaeoloxodon antiquus (MG 3561) recovered by Paul Choffat in Condeixa during the XIX century in lateral view.

Rhinoceroses were identified for first time in the Quaternary of Portugal by Nery Delgado, from the material of Furninha cave (Delgado, 1884). Until now it seems that only one species is known in this time period in Portugal: Stephanorhinus hemitoechus (Falconer, 1859), mostly thanks to several dental specimens (Figure 1.19) but also some postcranial (two metatarsals, four astragalus and one calcaneus). The species is known only from cave material and appears in the localities of Serra dos Molianos, Furninha, Gruta Nova da Columbeira, Lapa da Rainha, Lorga de Dine, Correio-Mor, Figueira Brava, Pedreira das Salemas and Escoural, only between 30000 and 20000 YBP (Veiga Ferreira, 1975; Cardoso, 1993a).

Figure 1.19: Partial molar of Stephanorhinus hemitoechus from Gruta da Serra de Molianos. Geological Museum, Lisbon.

Other Perissodactyla, Owen 1848, include three species of equids that have been recognized in the Quaternary of Portugal: Equus hydruntinus, Equus ferus (Linnaeus, 1758) and Equus africanus (Linnaeus, 1758). The later one arrived to Portugal thanks to human introduction around 4000 YBP, 1500 years earlier than previously thought (Cardoso et al., 2013) but was absent of the country for all the Pleistocene. Equus ferus has a native subspecies described in Portugal: Equus ferus antunesi; whose type specimen (Figure 1.20) comes from Fointainhas cave. It was represented by a superbly preserved skull and some other limb bones that allow reconstructing this subspecies like a slender and small horse of 140 cm to the withers (Cardoso & Eisenmann, 1989). Other material that cannot be compared with the type of E. f. antunesi appear in many Portuguese sites and has been simply attributed to Equus ferus appearing quite commonly in the Portuguese records (mostly dental material), from the prewürmian sites of Mealhada to the archaeological contexts of Furninha, Vale Boi and Foz do Enxarrique to the relatively modern sites like Casais Robustos (Cardoso, 1993a; Raposo, 1995; Manne et al., 2012). Equus hydruntinus appear at least in Pedreira das Salemas, Caldeirão cave and the site of Vale Boi (Cardoso, 1995; Davis, 2002; Manne et al., 2012). If the “zebro” of medieval chronicles is this species, it only disappeared in Portugal during the XV century (Antunes, 2006).

Figure 1.20: Type specimen of Equus ferus antunesi in lateral view (MG). Geological Museum, Lisbon.

There is no big associated or articulated cetacean skeleton from geological contexts during this period in Portugal, but several bones collected in archaeological contexts have been recovered, some of them only centuries old, that are testimony of the hunt of whales during medieval and modern times (Teixeira et al., 2014). In similar contexts but way older we have a cetacean bone from the site of Vale Boi, Algarve, with around 28000 years (Manne et al., 2012) a possible cetacean bone from Lagar Velho (Moreno-García & Pimenta, 2002) and six vertebrae ascribed to Delphinus delphis Linnaeus, 1758 from Figueira Brava with around 30.000 years (Antunes, 2000). The hippopotamuses are known in Portugal in three localities: Condeixa, Mealhada and Algoz. The remains of Condeixa (Figure 1.21) are dental and mandibular remains (a fragment of inferior mandible, a molar and an isolated canine) already collected by Paul Choffat in the XIX century (Choffat, 1895). In Mealhada there is also a left cuboid (Antunes et al., 1988). Both have been attributed to Hippopotamus incognitus Faure, 1884 by Cardoso, 1993 although some authors considered it a synonym of Hippopotamus amphibious Linnaeus, 1758 (Petronio, 1995). The remains from Algoz are the better preserved ones, including one femur head, a distal part of a tibia, a calcaneum, an incomplete astragalus; a fragmentary cuboid and navicular, a fourth left metatarsal and two phalanges that are regarded as belonging to the same animal. They had been attributed to Hippopotamus major Cuvier, 1825 (Antunes et al., 1986c) although by priority rule they should be named Hippopotamus antiquus Desmarets, 1822 (Magri et al., 2010).

Figure 1.21: Anterior mandibular fragment of Hippopotamus incognitus from Condeixa in dorsal view. Geological Museum, Lisbon.

Most authors agree that nearly all of the suid remains on Europe from the Pleistocene belong to the extant species Sus scrofa Linnaeus, 1758; although the presence of Sus strozzi Forsyth Major, 1881 in the Early Pleistocene of Europe has been proved (Alcalá et al., 1985). Sus scrofa is divided nowadays in five subspecies across Europe, but the distinction at sub specific level in the fossil record (mostly based on dental remains and size, like for example the subspecies S. s. priscus) is specially problematic (Cardoso, 1993a; Lister et al., 2010). In Portugal, although this species is nowadays in expansion and quite common (Sáez-Royuela & Tellería, 1986) its remains from the Upper Pleistocene and Holocene are rare. It is known in sites like Fontainhas, Pedreira e Gruta das Salemas, Escoural, Lapa do Picareiro, Figueira Brava and Caldeirão (Cardoso, 1993a; Hockett & Haws, 2009). The later ones have been suggested to belong to domestic animals (Davis, 2002).

The red deer (Cervus elaphus Linnaeus, 1758) (Figure 1.22) is probably the most common big mammal from the Pleistocene and Holocene of Portugal. It is known since the very first excavation of Furninha (Delgado, 1884), spanning from the oldest sites of Mealhada (Antunes et al., 1988) to Holocene sites in Archaeological contexts (Dean & Carvalho, 2011). It seems that there is a smaller morphotype in the Würm of Portugal (Cardoso, 1993a) and genetic studies of recent deer have pointed towards the genetic separation of European and Iberian red deer during the LGM (Queirós et al., 2019); therefore given the abundance of materials, ancient DNA analysis of the red deer fossil occurrences would be interesting. The roe deer (Capreolus capreolus Linnaeus, 1758) is rare, being it known from the caves of Caldeirão, Columbeira, Lapa da Rainha, Picareiro and Porto Covo (Cascais) the later ones probably from Holocene (Cardoso, 1993a, Hockett & Haws, 2009). The fallow deer (Dama dama Linnaeus, 1758) has been recognized in the Algar de João Ramos by an inferior tooth, Gruta Nova da Columbeira with dental series and Pedreira das Salemas by a first phalanx. The chronology of all the Pleistocene remains is younger than 30000 and older than 14000 YBP (Cardoso, 1989). There are not known remains of this species of Pre-Roman Holocene time, coming the oldest of this period from the Roman sites of São Pedro da Fronteira and Torre de Palma, so it is possible that Pleistocene remains belonged to a different species of the genus Dama (Davis & MacKinnon, 2009). An extinct species of the genus; Dama vallonnetensis De Lumley et al., 1988, is cited from the older middle Pleistocene deposits in the Almonda system (Marks, Monigal, et al., 2002). Eucladoceros sp. Falconer, 1868 has also been cited from the lowermost Middle Pleistocene of Algoz, Algarve (Antunes et al., 1986c).

Figure 1.22: Skull of female Cervus elaphus from Gruta das Fontainhas in lateral view. Geological Museum, Lisbon.

The bovids in Portugal during the Quaternary are represented by the auroch, Bos primigenius Bojanus, 1827; the Iberian ibex, Capra pyrenaica Schinz, 1838; the Pyrenean chamois, Rupicapra pyrenaica Bonaparte, 1845 and the sheep/mouflon Ovis orientalis Linnaeus, 1758; arriving the last one by human hand to Portugal in the Holocene only 3000 thousands year after its domestication, 5450 YBP (Davis & Simões, 2016). The auroch, Bos primigenius, is common in Portugal, appearing in all major Quaternary sites. It is necessary to remark that the difficulties to differentiate this species from members of Bison Hamilton Smith, 1827 make the former´s presence in Portugal probable, although not proved, because not

18 enough well preserved material have been recovered (the easiest way to differentiate between them is from cranial material, quite scarce), (Cardoso, 1993a). The Iberian ibex, Capra pyrenaica, is an endemism of the Iberian Peninsula. The subspecies typical of Portugal, Capra pyrenaica lusitanica went extinct at the end of the XIX century but nowadays the species is again living in Portugal (Moço et al., 2006). It is known from sites like Fontainhas, Caldeirão (Figure 1.23), Casais Robustos (that we will discuss later), Pedreira das Salemas, Figueira Brava, Escoural, Pego de Diablo, Algar of João Ramos or Vale Boi (Cardoso, 1993a, Manne et al., 2012). The Pyrenean chamois, Rupicapra pyrenaica is known in Portugal from the localities of Gruta das Salemas, Caldeirão, Pego do Diabo and Lapa do Picareiro (Cardoso, 1993a; Bicho et al., 2000). The fossil population of Portugal was regarded as a subspecies of the chamois Rupicapra rupicapra pyrenaica (Linnaeus ,1758) but further genetic investigation revealed that the animals living in the Cantabric mountains, the Pyrenees and the Apennines have been separated for around 30000 years from the main population in the Alps, that came from the East during the Würm, so the specimens from Portugal probably belonged to Rupicapra pyrenaica (Pérez et al., 2002).

Figure 1.23: Distal part of a humerus of Capra pyrenaica in anterior view from Caldeirão cave. Catalog number and scale of the specimen not provided in the original (Davis, 2002).

Concerning Carnivora Bowdich, 1821 the pinnipeds are represented by one right ulna from Figueira Brava attributed to Pusa hispida (Schereber, 1775), (Antunes, 2000) that today lives in the high Arctic, in locations like Svalbard (Krafft et al., 2007). There is also a part of a mandible at exposition in the Geological Museum in Lisbon that is labeled as “Foca (Phocus monachus)”, but it does not appear to be in the bibliography.

Mustelidae Fischer de Waldheim, 1817 is present in the Quaternary of Portugal. In the site of Vale Boi there is a reference to the appearance of Mustela Linnaeus, 1758. Martes Pinel, 1792 appear in Vale Boi and Furninha and the badger, Meles meles Linnaeus, 1758 appear in Furninha, Caldeirão and Lapa do Suão (Davis, 2002; Valente & Haws, 2006; Brugal et al., 2012; Manne et al., 2012).

Three canid species have been recognized from this time period in Portugal: Canis lupus Linnaeus, 1758, Vulpes vulpes (Linnaeus, 1758) and Cuon alpinus Pallas, 1811. The first two are relatively common, being known since the XIX century from diverse Portuguese caves, like Furninha, Fontainhas or Caldeirão, but Cuon alpinus is only known in Portugal thanks to a fragment of hemimandible from the Escoural cave (Figure 1.24) ( Cardoso, 1993a), although studies suggest that many Cuon remains could be misidentified in late Pleistocene and Holocene contexts so it would be risky to assess the relative abundance of that species compared to the other two (Ripoll et al., 2010).

Figure 1.24: Fragment of left hemimandible of Cuon alpinus with P4 and fragment of M1 from Escoural cave in oclusal (left) and labial (right) views. Catalog number of the specimen or not provided in the original, scale do not appear in the image (Cardoso, 1993a).

At least three works until this date agree in the conclusion that there is only one species of ursid recognized from the Quaternary of Portugal: the brown bear, Ursus arctos Linnaeus, 1758 (Harlé, 1910; Torres, 1979; Cardoso, 1993a). There are bear remains known in most of major Portuguese Quaternary sites: Furninha (Figure 1.25), Fontainhas, Molianos, Pedreira and Gruta das Salemas, Algar de Cascais, Figueira Brava, Gruta de Almonda (Galería Pesada), Caldeirão and the hesperic massif sites of Lorga de Dine and Escoural, although it is only abundant in the two first sites. This is not a surprise given the affinity of this species for caves during winter periods (Cardoso, 1993a; Marks et al., 2002a; Davis, 2002). Besides this locations it has been detected in recent Holocene archaeological contexts, for example in Leceias (Pires, 2002). It is also worth to mention that the cave bear Ursus spelaeus Rosenmüller, 1794 in known to have lived near the Portuguese territory in the Galician cave of Rebolal, only 50 km away from Portugal (Grandal-d’Anglade et al., 2006) .

Figure 1.25: Left Ursus arctos mandible from Furninha cave in lingual view. Geological Museum, Lisbon.

Two species of hyenas have been cited in the Pleistocene of Portugal: Hyaena hyaena (Linnaeus, 1758) and Crocuta crocuta (Erxleben, 1777). Hyaena hyaena prisca de Serres, 1828 is cited from Furninha cave with some abundance (Figure 1.26), which is curious given that the species is not cited in the Pleistocene of Spain, meanwhile it appears in France. It seems that it survived longer in Portugal that in other locations of its distribution (Cardoso, 1996; Brugal et al., 2012).

Crocuta crocuta spelaea (Goldfuss, 1832) is more common, appearing in at least ten sites in the Pleistocene of Portugal including Caldeirão or Pego do Diabo (Cardoso, 1996; Valente, 2004). There is also a mention to the presence of Crocuta crocuta intermedia, an older and smaller subspecies of spotted hyena in Lorga de Dine (Cardoso, 1994).

Figure 1.26: Skull of Hyaena hyaena prisca from Furninha cave in dorsal view. Geological Museum, Lisbon.

Felidae Fischer, 1817 is represented in the Quaternary of Portugal by at least five species: Felis silvestris Schreber, 1777, Panthera spelaea, Panthera pardus Linnaeus, 1758, at least one species of Lynx Kerr, 1792, and Homotherium latidens (Cardoso, 1993a; Antunes, 1986). Felis silvestris in known from at least ten sites of this chronology in Portugal (Villaluenga, 2016), in more recent archaeological contexts (younger than 10000 YBP) the specimens could belong to domestic cat (Baca et al., 2018). The lynx is the most common felid in the Pleistocene of Portugal, appearing at least in more than eleven sites (Villaluenga, 2016) including a new one that will be presented later in this work. If all that remains are from one or various species of lynx is a question that is not clear. Until now all of them have been attributed either to the extant endangered Lynx pardinus Temminck, 1827 or to Lynx spelaea (Boulé, 1906), being the latter regarded as a junior synonym of the former by some (Boscaini et al., 2016). The possible presence of European lynx, Lynx lynx (Linnaeus, 1758) cannot be ruled out under the light of new works that have proved its presence in the Iberian Peninsula in the Holocene (Rodríguez-Varela et al., 2016). There are also tracks attributed to lynxes in the eolianites of the South of Portugal (Neto de Carvalho, 2014). The leopard, Panthera pardus, is present in twelve sites of the Quaternary of Portugal (Cardoso & Regala, 2006; Villaluenga, 2016) including one mostly complete and well preserved skull from the Algar da Manga Larga (Figure 1.27). The cave lion, Panthera spelaea is known by scarce material from seven localities in Portugal: Vale Boi (non specified remains), Pedreira das Salemas (three canine teeth and one phalanx), Gruta de Penacova (one canine teeth), Lorga de Dine (one fourth right metacarpal), Escoural (one third premolar), Caldeirão (one phalanx) and Figueira Brava (one phalanx) (Antunes & Cardoso, 1986; Cardoso, 1993a; Manne et al., 2012; Villaluenga, 2016). Homotherium latidens has been identified in the Pleistocene of Portugal by one left astragalus in the Mealhada site, probably older than 100000 YBP (Antunes, 1986).

Figure 1.27: The leopard (Panthera pardus) skull from Algar da Manga Larga in anterolateral view. Geological Museum, Lisbon.

1.3 Aim, context and objectives of the present thesis.

This thesis will focus in Pleistocene tetrapods (with special emphasis on macro-mammals), although some Holocene material will be included on it:

1. Paleodiversity census. Compile a state of the art census of the fossil Quaternary tetrapods known in Portugal across the literature, with a list of species, localities and publications. Quantify bias in the fossil record, the spatial distribution of the sites, extinction rates and differences between groups (What groups suffered more severe extinctions? Which are the better represented in the fossil record?) Establish the quality of the fossil record, i.e., what percentage of biodiversity is represented in the fossil record. 2. The Palaeoloxodon antiquus of Santo Antão do Tojal: Estimate for first time with the most up to date numerical methods the body mass of a Portuguese proboscidean and study its implications. Perform a metric study with a dataset of more than 40 Eurasian Palaeoloxodon antiquus and Mammuthus primigenius (including the Santo Antão do Tojal specimen) to determine metric differences between the two species. 3. The bears of Algar do Vale da Pena: Excavation and description of the new locality and recovered bear materials. Compare the skeletal measurements of the recovered remains to with other Iberian populations. Provide a specific determination. Asses the implications of this new information in the context of Portugal. 4. The microvertebrates of Santa Margarida: Describe the new locality. Elaborate a list of fauna. Provide a chronology for the site.

Given this objectives the thesis will be separated in 4 sections according to each one of them.

2. PALEOBIODIVERSITY OF THE QUATERNARY FOSSIL TETRAPODS IN CONTINENTAL PORTUGAL

2.1 Introduction and methods

On this chapter an inventory of localities and tetrapod species recognized in the Quaternary of Portugal will be provided in order to give us an approach to the paleobiodiversity of the territory. For each locality some general information will be given. For each group a table will be provided including the species name, the localities where it has been recognized and the references to that occurrence, also specifying if the species is found today in Portugal or not. Then a resume table with some numerical considerations for each group will be given, like the number and percentage of extinct species in the sample, the number and percentage of species locally extinct in Portugal and the number and percentage of species of the group that are only known as fossils (Species extinct + Species locally extinct). Also there will be the number and percentage of species detected and undetected in the fossil record for each group, in comparison with the species that live today in Portugal. Then there will be a general overview, providing some comparison between groups, some numerical examples about the importance of the cave deposits in the Quaternary record and some inferences about the paleobiodiversity of other periods. This section has been partially published in an abtract and presented in the 1st Paleontological Virtual Congress (Estraviz-López & Mateus, 2018a).

2.2 Lists of localities and fossil tetrapods of the Quaternary of Portugal

In the following table 2.1 a list of the principal Portuguese Quaternary localities will be provided, including name, coordinates, municipality, approximate age(s) and some references. It should be noted that not all the sites are on it; some minor localities where isolated remains of elephants or other fauna have been found are not listed, like Meirinha, Carregado or Porto Covo (Cascais). Also the coordinates of some localities are not totally precise, since they were not explicitly pointed in the literature.

Table 2.1: Localities of the Quaternary of Portugal.

MUNICIPALIT NAME Y COORDINATES AGE REFERENCES Late Pleistocene Pre Würm (80.000 YBP?) 41° 54' 35" N and mainly Würm Cardoso, 1994; Lorga de Dine Vinhais 6° 55' 49" W (20.000 YBP) Cardoso, 1996 Middle Pleistocene Zbyszewski, 40° 22' 35" N (interglacial Mindel- 1977a; Antunes Mealhada Mealhada 8° 27' 30" W Riss?) et al., 1988 Condeixa-a- 40° 05' 49" N Middle Pleistocene Choffat, 1895; Condeixa Nova 8° 29' 59" W (Mindel?) Cardoso, 1993a Cardoso, 39° 38' 59" N 1993a; Davis, Caldeirão Tomar 8° 24' 54" W 30.000 - 5.700 YBP 2002 Antunes & Cardoso, 1989; Raposo, 1995; Foz do Vila Velha de 39° 38' 58" N Cunha et al., Enxarrique Ródão 7° 40' 11" W 140.000-35.000 YBP 2019 Lapa do 39° 32' 33" N Bicho et al., Picareiro Fátima 8° 38' 02" W 12.000 - 6.500 YBP 2006 Marks et al., Almonda 39° 30' 19" N Middle Pleistocene 2002 a; López- System Torres Novas 8° 36' 58" W (~250.000 YBP) García et al.,

2018 Cardoso, 1993a; Estraviz-López 39° 30' 12" N & Mateus, Casais Robustos Alcanena 8° 40' 20" W Recent Würm? 2018c. Gruta da Serra de Molianos Alcobaça Unknown Recent Würm? Cardoso, 1993a Algar de João 39° 23' 57" N Ramos Alcobaça 8° 56' 10" W ~15.000 YBP Cardoso, 1993a 39° 21' 23" N Aprox 80.000 YBP(But Brugal et al., Furninha Peniche 9° 24' 14" W with great uncertainty) 2012 39° 19' 36" N Casa da Moura Óbidos 9° 15' 14" W ~25.000 YBP Cardoso, 1993a Gruta Nova da 39° 18' 06" N Columbeira Bombarral 9° 11' 58" W ~25.000 YBP Cardoso, 1993a 39° 17' 59" N Valente & Lapa do Suão Bombarral 9° 12' 01" W 15.000 - 10.000 YBP Haws, 2006 39° 11' 37" N Fontainhas Cadaval 9° 02' 39" W ~22.000 YBP Cardoso, 1993a 39° 10' 51" N Lapa da Rainha Torres Vedras 9° 19' 38" W 25.000 - 20.000 YBP Cardoso, 1997 39° 10' 29" N Bom Santo Cave Alenquer 9° 02' 02" W 5.800 - 5.400 YBP Pimenta, 2014 Pedreira e Gruta 38° 52' 38" N das Salemas Loures 9° 11' 58" W 30.000 -20.000 YBP Cardoso, 1993a Cardoso, 38° 51' 50" N 1993a; Valente, Pego do Diabo Loures 9° 13' 00" W 28.000 - 23.000 YBP 2004 Santo Antão do 38° 50' 26" N 9° Tojal Loures 8' 37" W 80.000 YBP Raposo, 1995 38° 49' 44" N Correio-Mor Loures 9° 10' 50" W Recent Würm Cardoso, 1993a 38° 45' 06" N Porto Covo Cascais 9° 25' 23" W Recent Würm Cardoso, 1993a 38° 42' 10" N Algar de Cascais Cascais 9° 25' 08" W 18.000 YBP Cardoso, 1993a Motemor-o- 38° 32' 59" N Escoural Novo 8° 08' 15" W 20.000-15.000 YBP Cardoso, 1993a 38° 28' 23" N Figueira Brava Setúbal 8° 59' 04" W 30.000 YBP Antunes, 2000 37° 09' 07" N Antunes et al., Algoz Silves 8° 17' 51" W Nearly 1 million YBP 1986c 37° 07' 38" N Antunes et al., Goldra Loulé 7° 59' 19" W Early Würm 1986b 37° 07' 13" N Late Pleistocene-Early Antunes et al., Guia Albufeira 8° 17' 43" W Holocene 1989 37° 05' 48" N Antunes et al., Morgadinho Tavira 7° 42' 43" W Nearly 1 million YBP 1986a 37° 05' 25" N Mostly 27.000-17.000 Manne et al., Vale Boi Vila do Bispo 8° 48' 26" W YBP (8.000 YBP) 2012

Figure 2.1: Satellital picture showing the location of 30 sites with Quaternary fossil vertebrates in Portugal, mentioned in the table 2.1. In black, the localities situated in the Lusitanian basin and in red the ones situated in the Algarve basin. Credit Google Earth.

As we can see in figure 2.1.; practically all the localities are situated in two areas over de Mesozoic basins of Portugal (Terrinha et al., 2006; Kullberg et al., 2014). The stretch of the Lusitanian basin between Lisbon and Leiria (plus the Setúbal Peninsula) covers an area of 5000 square kilometers, about 5,5% the surface of the country, yet it comprises 18 of the 30 localities considered (More than 50% of them). Most of the other localities are situated in the Algarve basin, although it is interesting to note that most of them are situated in open air sites and not in caves, opposed to what happens in the North. Only three localities are outside these main areas: Lorga de Dine (Northernmost point), Escoural (Southernmost point outside the main areas) and Foz do Enxarrique (Between them). Those two caves are the only ones whose fauna have been studied in detail in the Hesperian Massif of Portugal and Foz do Enxarrique not only is outside the main areas of findings, but it is also an open air site near the Tagus River. Only ten localities are situated in open air sites, most of them in Algarve.

Tables 2.2 (amphibians), 2.3 (reptiles), 2.4 (birds) and 2.5 (mammals) summarize the known vertebrates from the Quaternary of Portugal. Some nomenclature is corrected, see previous chapter for further information. Table 2.2: Amphibians of the Quaternary of Portugal.

EXTANT IN PORTUGAL TAXON LOCALITIES REFERENCES ?

Moreno-García & Salamandra salamandra Lagar Velho, Figueira Pimenta, 2002; (Linnaeus, 1758) Brava Crespo et al., 2000 YES Triturus marmoratus Moreno-García & (Latreille, 1800) Lagar Velho Pimenta, 2002 YES Hockett & Haws, 2009; Antunes et al., YES Bufo bufo (Linnaeus, 1758) Lapa do Picareiro, Guia 1989 Hockett & Haws, Pelobates cultripes Cuvier, Lapa do Picareiro, Figueira 2009; Crespo et al., YES 1829 Brava, Guia 2000 Pleurodeles waltl YES (Michahelles, 1830) Guia Antunes et al., 1989 Epidalea calamita Laurenti, YES 1758 Guia Antunes et al., 1989 Pelophylax perezi (Seoane, YES 1885) Guia Antunes et al., 1989

Table 2.3: Reptiles of the Quaternary of Portugal.

EXTANT IN TAXON LOCALITIES REFERENCES PORTUGAL? Timon lepidus Daudin, Crespo et al., 2000, 1802 Figueira Brava, Guia Antunes et al., 1989 YES Psammodromus manuelae Busack et al., 2006 Figueira Brava Crespo et al., 2000 YES

Podarcis sp. Wagler, 1830 Figueira Brava Crespo et al., 2000 YES Acanthodactylus erythrurus (Schinz, 1833) Guia Antunes et al., 1989 YES Rinechis scalaris (Schinz, 1822) Figueira Brava Crespo et al., 2000 YES Hemorrhois hippocrepis (Linnaeus, 1758) Figueira Brava Crespo et al., 2000 YES Gruta da Aroeira Crespo et al., 2000; Coronella girondica (Almonda System), López-García et al., YES Daudin, 1803 Figueira Brava 2018 Vipera lastei Bosca, 1878 Figueira Brava Crespo et al., 2000 YES Gruta da Aroeira Antunes et al., 1989; YES (Almonda System), López-García et al., Natrix sp. Laurenti, 1758 Guia 2018 Blanus cinereus Vandelli, 1797 Figueira Brava Crespo et al., 2000 YES Testudo hermanni Gmelin, Mealhada, Furninha, Lapparent de Broin &

1789 Figueira Brava, Gruta Antunes, 2000; Nova da Columbeira, Morales Pérez & Caldeirão, Escoural, Serra, 2009. Gruta de Ibn Amar NO Mealhada, Caldeirão, Lapparent de Broin & Porto Covo (Cascais), Antunes, 2000; Mauremys leprosa Gruta Nova da Morales Pérez & (Schweiger, 1812) Columbeira Serra, 2009. YES Lapparent de Broin & Antunes, 2000; Emys orbicularis Figueira Brava, Gruta Morales Pérez & (Linnaeus, 1758) Nova da Columbeira Serra, 2009. YES

Table 2.4: Birds of the Quaternary of Portugal.

EXTANT IN TAXON LOCALITIES REFERENCES PORTUGAL? Tetrao urogallus Linnaeus, 1758 Lapa da Rainha Figueiredo, 2010 NO Lapa do Picareiro, Furninha, Lapa do Suão, Gruta Nova da Alectoris rufa Columbeira, Escoural, Figueira Linnaeus, 1758 Brava, Bom Santo Cave Pimenta et al., 2015 YES Alectoris barbara (Bonnaterre, 1791) Gruta Nova da Columbeira Figueiredo, 2010 YES

Perdix perdix Lapa do Picareiro, Furninha, Gruta Figueiredo, 2010; Linnaeus, 1758 Nova da Columbeira, Caldeirão Pimenta et al., 2015 NO Coturnix coturnix Furninha, Gruta Nova da Linnaeus, 1758 Columbeira Figueiredo, 2010 YES Phasianus YES colchicus Linnaeus, 1758 Gruta Nova da Columbeira Figueiredo, 2010 Lagopus lagopus Linnaeus, 1758 Lapa da Rainha Figueiredo, 2010 NO Lagopus muta Montin, 1781 Gruta Nova da Columbeira Figueiredo, 2010 NO Anas platyrhynchos Lapa do Suão, Escoural, Figueira Linnaeus, 1758 Brava Pimenta et al., 2015 YES Anas acuta Linnaeus, 1758 Gruta Nova da Columbeira Pimenta et al., 2015 YES Anas crecca Linnaeus, 1758 Furninha Pimenta et al., 2015 YES Melanitta nigra (Linnaeus, 1758) Furninha, Figueira Brava Pimenta et al., 2015 YES Melanitta fusca (Linnaeus, 1758) Figueira Brava Pimenta et al., 2015 YES Clangula hyemalis Linnaeus, 1758 Figueira Brava Pimenta et al., 2015 YES Tadorna tadorna (Linnaeus, 1758) Furninha Figueiredo, 2010 YES Tadorna ferruginea Pallas, 1761 Furninha Figueiredo, 2010 YES Somateria Furninha Figueiredo, 2010

27 mollisima NO (Linnaeus, 1758) Mergus albellus (Linnaeus, 1758) Galería Pesada (Almonda) Figueiredo, 2010 YES Anser albifrons (Scopoli, 1769) Pego do Diabo Figueiredo, 2010 NO Cygnus olor (Gmelin, 1789) Galería Pesada (Almonda), Furninha Pimenta et al., 2015 NO Scolopax rusticola Galería Pesada (Almonda), Figueira (Linnaeus, 1758) Brava Pimenta et al., 2015 YES Numeius phaeopus (Linnaeus, 1758) Furninha, Figueira Brava Pimenta et al., 2015 YES Calidris canutus Linnaeus, 1758 Lapa do Picareiro, Figueira Brava Pimenta et al., 2015 YES Larus fuscus Linnaeus, 1758 Figueira Brava Pimenta et al., 2015 YES Pinguinus impennis (Linnaeus, 1758) Furninha, Figueira Brava Pimenta et al., 2015 NO Alca torda Linnaeus, 1758 Figueira Brava Pimenta et al., 2015 YES Vanellus vanellus Gruta Nova da Columbeira, (Linnaeus, 1758) Fontainhas Figueiredo, 2010 YES Pluvialis squatarola YES (Linnaeus, 1758) Gruta Nova da Columbeira Figueiredo, 2010 Pluvialis apricaria (Linnaeus, 1758) Lapa do Picareiro Pimenta et al., 2015 YES Podiceps nigricollis Brehm, 1831 Figueira Brava Figueiredo, 2010 YES Otis tarda Linnaeus, 1758 Gruta de Casa da Moura Figueiredo, 2010 YES Grus grus Linnaeus, 1758 Gruta de Casa da Moura Figueiredo, 2010 YES Grus primigenia (Milne-Edwards, NO 1869) Figueira Brava Figueiredo, 2010 Puffinus puffinus Furninha, Gruta Nova da (Brünnich, 1764) Columbeira Pimenta et al., 2015 YES Puffinus holeae Walker et al., 1990 Figueira Brava Pimenta et al., 2015 NO Phalacrocorax aristotelis YES (Linnaeus, 1761) Furninha, Figueira Brava Pimenta et al., 2015 Phalacrocorax carbo (Linnaeus, YES 1758) Figueira Brava Pimenta et al., 2015 Morus bassanus (Linnaeus, 1758) Figueira Brava Pimenta et al., 2015 YES Phoenicopterus ruber Pallas, 1811 Furninha Figueiredo, 2010 YES Gavia stellata (Pontoppidan, YES 1763) Figueira Brava Pimenta et al., 2015

Accipiter nisus (Linnaeus, 1758) Figueira Brava Pimenta et al., 2015 YES Aquila chrysaetos Linnaeus, 1758 Figueira Brava, Vale Boi Pimenta et al., 2015 YES Aquila adalberti Brehm, 1861 Lapa da Rainha Figueiredo, 2010 YES Milvus migrans (Boddaert, 1783) Figueira Brava Pimenta et al., 2015 YES Falco tinnuncus Linnaeus, 1758 Lagar Velho, Figueira Brava, Pimenta et al., 2015 YES Falco rusticolus Linnaeus, 1758 Gruta de Casa da Moura Figueiredo, 2010 NO Aquila fasciata (Viellot, 1822) Figueira Brava Pimenta et al., 2015 YES Aegypius monachus (Linnaeus, 1766) Caldeirão Figueiredo, 2010 YES Gyps fulvus Hablizl, 1783 Furninha, Caldeirão Pimenta et al., 2015 YES Buteo buteo (Linnaeus, 1758) Galería Pesada (Almonda) Figueiredo, 2010 YES Haliaeetus albicilla Linnaeus, 1758 Galería Pesada (Almonda) Pimenta et al., 2015 YES Bubo bubo Galería Pesada (Almonda), (Linnaeus, 1758) Caldeirão, Furninha, Figueira Brava Pimenta et al., 2015 YES Galería Pesada (Almonda), Caldeirão, Lapa do Picareiro, Gruta Athene noctua Nova da Columbeira, Lapa da Scopoli, 1769 Rainha, Escoural, Figueira Brava Pimenta et al., 2015 YES Strix aluco Linnaeus, 1758 Gruta Nova da Columbeira Figueiredo, 2010 YES Asio flammeus (Potoppidan, 1763) Lapa do Picareiro, Furninha Pimenta et al., 2015 YES Columba livia Galería Pesada (Almonda), Lapa do Gmelin, 1789 Picareiro, Furninha, Figueira Brava Pimenta et al., 2015 YES Columba palumbus Galería Pesada (Almonda), Linnaeus, 1758 Caldeirão, Escoural, Figueira Brava Pimenta et al., 2015 YES Coracias garrulus Lapa do Picareiro, Lapa do Suão, Linnaeus, 1758 Gruta Nova da Columbeira Pimenta et al., 2015 YES

Egretta sp. Forster, Figueiredo, 2010, 1817 Galería Pesada (Almonda) Pimenta et al., 2015 YES Galería Pesada (Almonda), Caldeirão, Lapa do Picareiro, Pyrrhocorax Furninha, Lapa do Suão, Gruta Nova pyrrhocorax da Columbeira, Fontainhas, Lapa da Linnaeus, 1758 Rainha, Figueira Brava Pimenta et al., 2015 YES Galería Pesada (Almonda), Lapa da Rainha, Lapa do Picareiro, Furninha, Pyrrhocorax Gruta de Casa da Moura, graculus (Linnaeus, Fontainhas, Caldeirão, Gruta Nova Figueiredo, 2010, NO 1766) da Columbeira Pimenta et al., 2015

Galería Pesada (Almonda), Lagar Velho, Lapa do Picareiro, Furninha, Lapa do Suão, Gruta Nova da Corvus monedula Columbeira, Fontainhas, Lapa da (Linnaeus, 1758) Rainha, Figueira Brava Pimenta et al., 2015 YES Caldeirão, Gruta Nova da Corvus corax Columbeira, Lapa da Rainha, Gruta Figueiredo, 2010, (Linnaeus, 1758) de Casa da Moura Pimenta et al., 2015 YES Corvus corone Galería Pesada (Almonda), Lapa do Linnaeus, 1758 Picareiro, Furninha Pimenta et al., 2015 YES Corvus frugilegus Linnaeus, 1758 Furninha, Escoural Pimenta et al., 2015 YES Corvus antecorax Mourer-Chauviré, NO 1975 Galería Pesada (Almonda) Figueiredo, 2010 Pica pica Linnaeus, Galería Pesada (Almonda), 1758 Caldeirão, Furninha, Escoural Pimenta et al., 2015 YES Garrulus glandarius Lapa do Picareiro, Lapa do Suão, YES Linnaeus, 1758 Gruta Nova da Columbeira Pimenta et al., 2015 Turdus iliacus Linnaeus, 1766 Furninha Figueiredo, 2010 YES Turdus pilaris Furninha, Gruta Nova da Linnaeus, 1758 Columbeira Figueiredo, 2010 YES Turdus philomelos Brehm, 1831 Gruta Nova da Columbeira Figueiredo, 2010 YES Turdus merula Linnaeus, 1758 Furninha Figueiredo, 2010 YES

Picus viridis Galería Pesada (Almonda), Gruta Figueiredo, 2010, Linnaeus, 1758 Nova da Columbeira Pimenta et al., 2015 YES Cyanopica cyanus (Pallas, 1776) Gruta Nova da Columbeira Figueiredo, 2010 YES Sturnus vulgaris Linnaeus, 1758 Lagar Velho, Lapa do Picareiro Figueiredo, 2010 YES Carduelis carduelis Linnaeus, 1758 Gruta Nova da Columbeira Figueiredo, 2010 YES Ptyonoprogne rupespris (Scopoli, YES 1769) Gruta Nova da Columbeira Figueiredo, 2010 Motacilla alba Linnaeus, 1758 Lapa da Rainha Figueiredo, 2010 YES Caprimulgus europaeus Galería Pesada (Almonda), Gruta YES Linnaeus, 1758 Nova da Columbeira Pimenta et al., 2015 Cuculus canorus (Linnaeus, 1758) Furninha, Gruta de Casa da Moura Figueiredo, 2010 YES

Table 2.5: Mammals of the Quaternary of Portugal.

EXTANT IN TAXON LOCALITIES REFERENCES PORTUGAL?

Erinaceus Mein & Antunes, 2000; Moreno- europaeus Furninha, Figueira Brava, Lagar García & Pimenta, 2002; Bicho et Linnaeus, 1758 Velho, Lapa do Picareiro al., 2003; Brugal et al., 2012 YES

Antunes et al., 1989; Mein & Figueira Brava, Lapa do Antunes, 2000; Moreno-García & Crocidura russula Picareiro, Lagar Velho, Guia, Pimenta, 2002; Bicho et al., 2003; Hermann, 1780 Bom Santo Cave Pimenta, 2014 YES Crocidura suaveolens Pallas, Figueira Brava cave, Lapa do Mein & Antunes, 2000; Bicho et al., YES 1811 Picareiro 2003 Sorex araneus Linnaeus, 1758 Goldra Antunes et al., 1986b NO Sorex sp. Linnaeus, 1758 Lagar Velho Moreno-García & Pimenta, 2002 YES Talpa occidentalis Figueira Brava, Lapa do Mein & Antunes, 2000; Bicho et al., Cabrera, 1907 Picareiro, Bom Santo Cave 2003; Pimenta, 2014 YES Talpa europaea Linnaeus, 1758 Goldra Antunes et al., 1986b NO Galemys kormosi (Shereuder, 1940) Morgadinho Antunes et al., 1986a NO Galemys pyrenaicus YES Geoffroy, 1811 Lapa do Picareiro Bicho et al., 2003 Rinolophus ferrumequinun Gruta da Aroeira (Almonda Mein & Antunes, 2000; López- (Shereber, 1774) System), Figueira Brava García et al., 2018 YES Rinolophus hipposideros YES (Bechstein, 1800) Figueira Brava Mein & Antunes, 2000 Rinolophus euryale Blasius, 1853 Figueira Brava Mein & Antunes, 2000 YES Myotis myotis Gruta da Aroeira (Almonda), (Borkhausen, Figueira Brava, Lapa do Mein & Antunes, 2000, Bicho et al., YES 1797) Picareiro 2003; López-García et al., 2018 Myotis blythi (Tomes, 1857) Figueira Brava Mein & Antunes, 2000 YES Myotis nattereri (Khul, 1818) Figueira Brava Mein & Antunes, 2000 YES Myotis mystacinus Khul, 1817 Lapa do Picareiro Bicho et al., 2003 YES Miniopterus schreibersi (Khul, YES 1819) Figueira Brava Mein & Antunes, 2000

Antunes et al., 1986b; Antunes et Apodemus Goldra, Guia, Caldeirão, Lagar al., 1989; Póvoas et al., 1992; sylvaticus Velho, Lapa do Picareiro, Bom Moreno-García & Pimenta, 2002; (Linnaeus, 1758) Santo Cave Bicho et al., 2003; Pimenta, 2014 YES

Apodemus flavicollis Gruta da Aroeira (Almonda Jeannet, 2000; López-García et al., NO (Melchior, 1834) System), Figueira Brava 2018 Mus musculus Linnaeus, 1758 Guia, Figueira Brava Antunes et al., 1989; Jeannet, 2000 YES Rattus rattus Linnaeus, 1758 Figueira Brava, Bom Santo Cave Jeannet, 2000; Pimenta, 2014 YES Sciurus vulgaris Jeannet, 2000; Moreno-García & Linnaeus, 1758 Figueira Brava, Lagar Velho Pimenta, 2002 YES

Antunes et al., 1986b; Antunes et Goldra, Guia, Caldeirão, Figueira al., 1989; Póvoas et al., 1992; Eliomys quercinus Brava, Lapa do Suão, Bom Santo Jeannet, 2000; Valente & Haws, Linnaeus, 1766 Cave 2006; Pimenta, 2014 YES Glis glis (Linnaeus, 1766) Lapa do Picareiro Bicho et al., 2003 YES Allocricetus bursae Schaub., 1930 Caldeirão Póvoas et al., 1992 NO Mimomys sp. Forsyth-Major, NO 1902 Morgadinho Antunes et al., 1986a Pliomys episcopalis Galería Pesada and Gruta da Marks et al., 2002a; López-García et NO Méhely, 1914 Aroeira (Almonda) al., 2018 Myodes sp. Pallas, NO 1811 ? Nehring, 1899 Chionomys nivalis Póvoas et al., 1992; Bicho et al., Martins, 1842 Caldeirão, Lapa do Picareiro 2003 YES

Arvicola sapidus Caldeirão, Figueira Brava, Lagar Póvoas et al., 1992; Jeannet, 2000; Miller, 1908 Velho Moreno-García & Pimenta, 2002 YES Arvicola amphibious NO (Linnaeus, 1758) Lapa do Suão Valente & Haws, 2006 Arvicola cantiana Hinton, 1910 Figueira Brava Jeannet, 2000 NO Antunes et al., 1989; Póvoas et al., Microtus 1992; Moreno-García & Pimenta, lusitanicus (Gerbe, Caldeirão, Lapa do Suão, Lagar 2002; Valente & Haws, 2006; 1879) Velho, Guia, Bom Santo Cave Pimenta, 2014 YES

Microtus Antunes et al., 1986; Póvoas et al., duodecimcostatus 1992; Jeannet, 2000; Moreno-García (de Selys- Figueira Brava, Caldeirão, Lapa & Pimenta, 2002; Bicho et al., 2003; Longchamps, do Picareiro, Lagar Velho, Valente & Haws, 2006; Pimenta, 1839) Goldra, Bom Santo Cave 2014 YES

Microtus arvalis Figueira Brava, Caldeirão, Lagar Póvoas et al., 1992; Jeannet, 2000; (Pallas, 1778) Velho Moreno-García & Pimenta, 2002 YES

Microtus agrestis Caldeirão, Lapa do Picareiro, Póvoas et al., 1992; Moreno-García (Linnaeus, 1761) Lagar Velho & Pimenta, 2002; Bicho et al., 2003 YES Microtus Gruta da Aroeira (Almonda Antunes et al., 1989; Póvoas et al., brecciensis Forsyth System), Figueira Brava, 1992; Jeannet, 2000; López-García NO

Major, 1905 Caldeirão, Goldra et al., 2018 Microtus cabrerae (Thomas, 1906) Caldeirão, Bom Santo Cave Póvoas et al., 1992; Pimenta, 2014 Castor fiber Linnaeus, 1758 Caldeirão Antunes, 1989 NO Prolagus calpensis Major, 1905 Morgadinho Antunes et al., 1986a NO Lepus granatensis Caldeirão, Pego do Diabo, Lapa Rosenhauer, 1856 do Suão Davis, 2002; Valente, 2004 Oryctolagus lacosti (Pomel, 1853) Algoz Antunes et al., 1986c NO

Hockett & Bicho, 2000; Davis, Oryctolagus Pego do Diabo, Lapa do Suão, 2002; Moreno-García & Pimenta, cuniculus Lapa do Picareiro, Caldeirão, 2002; Valente & Haws, 2006; (Linnaeus, 1758) Vale Boi, Lagar Velho Manne et al., 2012 YES Macaca sylvanus (Linnaeus, 1758) Galería Pesada (Almonda) Marks et al., 2002b NO Paleoloxodon Meirinha, Algés, Condeixa, antiquus Furninha, Almonda, Carregado, (Falconer & Cautle Mealhada, Foz do Enxarrique, Antunes & Cardoso, 1992; y, 1847) Santa Cruz, Santo Antão do Tojal Figueiredo, 2012 NO Mammuthus primigenius Algar de João Ramos, Figueira Antunes & Cardoso, 1992; Cardoso NO Blumebach, 1799 Brava, Oeiras sea floor & Regala, 2002 Galería Pesada (Almonda), Serra dos Molianos, Furninha, Gruta Nova da Columbeira, Lapa da Stephanorhinus Rainha, Lorga de Dine, Correio- hemitoechus Mor, Figueira Brava, Pedreira Veiga Ferreira, 1975; Cardoso, NO (Falconer, 1859) das Salemas, Escoural 1993a, Marks et al., 2002a Equus hydruntinus Pedreira das Salemas, Caldeirão, Cardoso, 1995; Davis, 2002; Manne Regalia, 1905 Vale Boi et al., 2012 NO Algar de João Ramos, Gruta das Fontainhas, Pedreira das Salemas, Algar de Cascais, Escoural, Gruta Nova da Columbeira, Furninha, Lapa da Rainha, Figueira Brava, NO Equus ferus Caldeirão, Vale Boi, Foz do Cardoso, 1993a; Raposo, 1995; (Linnaeus, 1758) Enxarrique David, 2002; Manne et al., 2012 Delphinus delphis Linnaeus, 1758 Figueira Brava Antunes, 2000 Hippopotamus amphibious Linnaeus, 1758 Condeixa, Mealhada Choffat, 1895; Antunes et al., 1988 NO Hippopotamus antiquus Desmaret, 1822 Algoz Antunes et al., 1986b NO Fontainhas, Pedreira das Salemas, Gruta das Salemas, Antunes et al., 1989; Cardoso, Escoural, Lapa do Picareiro, 1993a; Davis, 2002; Moreno-García Sus scrofa Figueira Brava, Caldeirão, Vale & Pimenta, 2002; Hockett & Haws, Linnaeus, 1758 Boi, Goldra, Lagar Velho 2009; Manne et al., 2012 YES

Galería Pesada (Almonda), Mealhada, Furninha, Fontainhas, Caldeirão, Pedreira das Salemas, Escoural, Lorga de Dine, Casais Robustos, Algar de Cascais, Porto Covo (Cascais), Lapa da YES Cervus elaphus Rainha, Figueira Brava, Pego do Linnaeus, 1758 Diabo Cardoso, 1993a Capreolus Caldeirão, Columbeira, Lapa da Cardoso, 1993a; Moreno-García & capreolus Rainha, Picareiro, Porto Covo Pimenta, 2002; Hockett & Haws, Linnaeus, 1758 (Cascais), Lagar Velho 2009 YES Algar de João Ramos, Gruta YES Dama dama Nova da Columbeira, Pedreira Linnaeus, 1758 das Salemas Cardoso, 1989 Dama vallonnetensis De NO Lumley et al., 1988 Galería Pesada (Almonda) Marks et al., 2002a Eucladoceros sp. Falconer, 1868 Algoz Antunes et al., 1986c NO Lorga de Dine, Fujaca, Gruta Nova da Columbeira, Pedreira das Salemas, Algar de Cascais, Figueira Brava, Escoural, Lapa da Rainha, Mealhada, Foz do Antunes et al., 1989; Cardoso, Bos primigenius Enxarrique, Furninha, Vale Boi, 1993a; Moreno-García & Pimenta, NO Bojanus, 1827 Guia, Lagar Velho 2002; Manne et al., 2012 Fontainhas, Caldeirão, Casais Robustos, Pedreira das Salemas, Figueira Brava, Escoural, Pego Capra pyrenaica de Diablo, Algar of João Ramos, YES Schinz, 1838 Vale Boi Cardoso, 1993a; Manne et al., 2012 Rupicapra Gruta das Salemas, Caldeirão, pyrenaica Pego do Diabo, Lapa do NO Bonaparte, 1845 Picareiro Cardoso, 1993a; Bicho et al., 2000 Pusa hispida (Schereber, 1775) Figueira Brava Antunes, 2000 NO Monachus monachus NO (Hermann, 1779) Furninha This work Furninha, Fontainhas, Casa da Moura, Algar de João Ramos, Lapa da Rainha, Pego do Diabo, Gruta Nova da Columbeira, Pedreira das Salemas, Gruta das Canis lupus Salemas, Caldeirão, Escoural, Cardoso, 1993a; Moreno-García & YES Linnaeus, 1758 Lagar Velho, Vale Boi Pimenta, 2002; Manne et al., 2012 Cuon alpinus Pallas, 1811 Escoural Cardoso, 1993a NO Furninha, Casa da Moura, Vulpes vulpes Figueira Brava, Caldeirão, Cardoso, 1993a; Moreno-García & (Linnaeus, 1758) Escoural, Vale Boi, Lagar Velho Pimenta, 2002; Manne et al., 2012 YES Furninha, Fontainhas, Molianos, Ursus arctos Gruta das Salemas, Algar de Cardoso, 1993a; Marks et al., Linnaeus, 1758 Cascais, Pedreira das Salemas, 2002a; Davis, 2002

Figueira Brava, Galería Pesada(Almonda), Caldeirão, NO Lorga de Dine, Escoural Mustela sp. Brugal et al., 2012; Manne et al., Linnaeus, 1758 Vale Boi, Furninha 2012 YES Martes sp. Pinel, Brugal et al., 2012; Manne et al., 1792 Vale Boi, Furninha 2012 YES

Meles meles Furninha, Caldeirão, Lapa do Davis, 2002; Valente & Haws, 2006; Linnaeus, 1758 Suão Brugal et al., 2012 YES Hyaena hyaena (Linnaeus, 1758) Furninha Brugal et al., 2012 NO Lorga de Dine, Caldeirão, Fontainhas, Gruta Nova da Columbeira, Lapa da Rainha, Gruta das Salemas, Algar de Crocuta crocuta Cascais, Figueira Brava, NO (Erxleben, 1777) Escoural, Porto Covo Cardoso, 1996; Valente, 2004 Escoural, Figueira Brava, Pego do Diabo, Gruta das Salemas, Pedreira das Salemas, YES Fontainhas, Furninha, Caldeirão, Felis silvestris Lapa da Rainha, Gruta Nova da Schreber, 1777 Columbeira Villaluenga, 2016 Escoural, Pego do Diabo, Pedreira das Salemas, Gruta das Salemas, Furninha, Caldeirão, Lapa da Rainha, Gruta Nova da Columbeira, Casa da Moura, Lynx pardinus Algar de João Ramos, Algar de YES Temminck, 1827 Cascais, Casais Robustos Villaluenga, 2016, This work Escoural, Figueira Brava, Pego do Diabo, Pedreira das Salemas, Fontainhas, Furninha, Casa da Moura, Galería Pesada Panthera pardus (Almonda), Caldeirão, Lorga de Cardoso & Regala, 2006; NO Linnaeus, 1758 Dine, Algar da Manga Larga Villaluenga, 2016 Escoural, Figueira Brava, Panthera spelaea Pedreira das Salemas, Lorga de Manne et al., 2012; Villaluenga, NO Goldfuss, 1810 Dine, Vale Boi 2016 Homotherium latidens Owen, NO 1846 Mealhada Antunes, 1988

2.3 Results

In the following resume table (Table 2.6) we can see some general numbers for the different groups. It should be noted that only native species are counted in each case, not the ones introduced in recent years, because those could not have been detected in the fossil record and therefore distort the representativeness of it. The table has been arranged following the data provided for nowadays species by Loureiro et al., 2008 for amphibians and reptiles; Assírio & Alvim, 2008 for birds; Rainho et al., 1998 for bats and Bencatel et al., 2017 for the other mammals (The four Quaternary marine mammals detected, as well as the modern ones are excluded). An extra column is added with the big mammals (mammals that are bigger than Meles meles and Felis silvestris) in the Quaternary of Portugal, but those are listed within the mammal category already.

Table 2.6: Resume table showing the number, percentage and status of extant and fossil species of the Quaternary of Portugal.

Amphibians Reptiles Birds Mammals Total Big mammals Extant 17 28 221 66 332 10 Known in the fossil record 7 13 80 74 174 30 Extinct 0 0 3 17 20 9 Locally extinct 0 1 10 17 29 11 Fossil only locally 0 1 13 34 49 20 In the fossil record and extant locally 7 12 67 40 125 10 Not known as fossil 10 15 154 26 207 0 % Extinct in the fossil record 0,00% 0,00% 3,75% 22,97% 11,49% 30,00% % Locally extinct in the fossil record 0,00% 7,69% 12,50% 22,97% 16,67% 36,67% % Fossil only locally 0,00% 7,69% 16,25% 45,95% 28,16% 66,67% % of extant species detected in the fossil record 41,18% 42,86% 30,32% 60,61% 37,65% 100,00% % of extant species undetected in the fossil record 58,82% 57,14% 69,68% 39,39% 62,35% 0,00% % of each group in extant species 5,12% 8,43% 66,57% 19,88% 100,00% 3,01% % of each group in fossil species 4,02% 7,47% 45,98% 42,53% 100,00% 17,24%

Looking at the general numbers at table 2.5, we can see that 174 species of fossil tetrapods have been discovered in the Quaternary of Portugal. A total of 49 (28,16%) of them no longer appear in the area either by local or complete extinction. 125 of the fossil species (that represent the 71,84% of them), still live in Portugal nowadays, comprising the 37,65% of the 332 species of living tetrapods in the country. We can presume therefore that 207 (or 62,35%) of them have not been detected in the fossil record despite already living in Portugal.

We can compare the groups now. In figure 2.2 we can see graphically the percentage of extinct, locally extinct and extant species in the fossil record for each group. We can see also the percentage of extant species detected vs the undetected in each group in the fossil record and the percentage of the total fossil and extant species that each group comprises.

Figure 2.2: Percentages of different parameters about the paleodiversity of the Quaternary in Portugal.

Of special interest is the case of the big mammals. In figure 2.3 we can see the status of the different species known in the fossil record. It should be noted that 2/3 of them are extinct or locally extinct. All ten species of them living nowadays in Portugal appear in the fossil record. Interestingly they comprised 17,24% of the fossil species, meanwhile today they are only the 3,01% of the living tetrapod species in Portugal (Table 2.5).

33,33% 30,00% % Extinct in the fossil record % Locally extinct in the fossil record %Extant

36,67%

Figure 2.3: Status of the different big mammals that appear in the fossil record in Portugal.

Looking at the localities in the table 2.1 we can see that most of them are cave deposits; but precisely how important are they to the understanding of the Quaternary fauna? In table 2.7 the species and localities associated with cave deposits have been eliminated. The references are in the table 2.6.

Table 2.7: Table showing the tetrapods and localities known outside the cave deposits in the Quaternary of Portugal.

TAXA LOCALITIES Salamandra salamandra (Linnaeus, 1758) Lagar Velho Triturus marmoratus (Latreille, 1800) Lagar Velho Bufo bufo (Linnaeus, 1758) Guia Pelobates cultripes Cuvier, 1829 Guia Pleurodeles waltl (Michahelles, 1830) Guia Epidalea calamita Laurenti, 1758 Guia Pelophylax perezi (Seoane, 1885) Guia Timon lepidus Daudin, 1802 Guia Acanthodactylus erythrurus (Schinz, 1833) Guia Natrix sp. Laurenti, 1758 Guia Testudo hermanni Gmelin, 1789 Mealhada Mauremys leprosa (Schweiger, 1812) Mealhada Aquila chrysaetos Linnaeus, 1758 Vale Boi Falco tinnuncus Linnaeus, 1758 Lagar Velho Sturnus vulgaris Linnaeus, 1758 Lagar Velho Crocidura russula Hermann, 1780 Lagar Velho, Guia Sorex araneus Linnaeus, 1758 Goldra Sorex sp. Linnaeus, 1758 Lagar Velho Talpa europaea Linnaeus, 1758 Goldra Galemys kormosi (Shereuder, 1940) Morgadinho Apodemus sylvaticus (Linnaeus, 1758) Goldra, Guia, Lagar Velho

Mus musculus Linnaeus, 1758 Guia Sciurus vulgaris Linnaeus, 1758 Lagar Velho Eliomys quercinus Linnaeus, 1766 Goldra, Guia Mimomys sp. Forsyth-Major, 1902 Morgadinho Arvicola sapidus Miller, 1908 Lagar Velho Microtus lusitanicus (Gerbe, 1879) Lagar Velho, Guia Microtus duodecimcostatus (de Selys-Longchamps, 1839) Lagar Velho, Goldra Microtus arvalis (Pallas, 1778) Lagar Velho Microtus agrestis (Linnaeus, 1761) Lagar Velho Microtus brecciensis Forsyth Major, 1905 Goldra Prolagus calpensis Major, 1905 Morgadinho Oryctolagus lacosti (Pomel, 1853) Algoz Oryctolagus cuniculus (Linnaeus, 1758) Vale Boi, Lagar Velho Meirinha, Algés, Condeixa, Carregado, Mealhada, Foz do Enxarrique, Santa Cruz, Paleoloxodon antiquus (Falconer & Cautley, 1847) Santo Antão do Tojal Mammuthus primigenius Blumebach, 1799 Oeiras sea floor Equus hydruntinus Regalia, 1905 Vale Boi Equus ferus (Linnaeus, 1758) Vale Boi, Foz do Enxarrique Hippopotamus amphibious Linnaeus, 1758 Condeixa, Mealhada Hippopotamus antiquus Desmaret, 1822 Algoz Sus scrofa Linnaeus, 1758 Goldra, Lagar Velho Cervus elaphus Linnaeus, 1758 Mealhada, Casais Robustos Capreolus capreolus Linnaeus, 1758 Lagar Velho Eucladoceros sp. Falconer, 1868 Algoz Pampilhosa do Botão, Mealhada, Foz do Bos primigenius Bojanus, 1827 Enxarrique, Vale Boi, Guia, Lagar Velho Capra pyrenaica Schinz, 1838 Vale Boi Canis lupus Linnaeus, 1758 Lagar Velho, Vale Boi Vulpes vulpes (Linnaeus, 1758) Vale Boi, Lagar Velho Mustela sp. Linnaeus, 1758 Vale Boi Martes sp. Pinel, 1792 Vale Boi Panthera spelaea Goldfuss, 1810 Vale Boi Homotherium latidens Owen, 1846 Mealhada

We can see in the table 2.8 the new resume table for the data in table 2.7, with an extra column for big mammals.

Table 2.8: Resume table showing the number, percentage and status of extant and fossil species in the Quaternary of Portugal, excluding the cave deposits.

Big Amphibians Reptiles Birds Mammals Total mammals Extant 17 28 221 66 332 10 Known in the fossil record (No caves) 7 5 3 37 52 17 Extinct (No caves) 0 0 0 12 12 8 Locally extinct (No caves) 0 1 0 4 5 3 Fossil only locally (No caves) 0 1 0 16 17 11 In the fossil record and extant (No caves) 7 4 3 21 35 6 Not known as fossil (No caves) 10 24 219 45 297 4 % Extinct in the fossil record (No caves) 0,00% 0,00% 0,00% 32,43% 23,08% 47,06% % Locally extinct in the fossil record (No caves) 0,00% 20,00% 0,00% 10,81% 9,62% 17,65% % Fossil only locally (No caves) 0,00% 20,00% 0,00% 43,24% 32,69% 64,71% % of extant species detected in the fossil record (No caves) 41,18% 14,29% 1,36% 31,82% 10,54% 60,00% % of extant species undetected in the fossil record (No caves) 58,82% 85,71% 99,10% 68,18% 89,46% 40,00% % of each group in extant species 5,12% 8,43% 66,57% 19,88% 100,00% 3,01% % of each group in fossil species (No caves) 13,46% 9,62% 5,77% 71,15% 100,00% 32,69%

If we exclude the cave deposits, the percentage of species detected overall falls from the previous 37,65% (125 species) to only a 10,54% (35 species). Also the total number of species falls from 174 to 52, a fall of the 70%. The cases of the birds and amphibians are especially interesting (As we see in figure 2.3) the first nearly disappearing from the fossil record (falling from a 45,98% to a mere 5,77% of the species) meanwhile the amphibians pass from the 4,02% to the 13,46% of all detected species, which is not surprising since all the 7 species located appear outside the cave deposits. It is also noteworthy that big mammal are over-represented comprising 32,69% of all the species when with the cave deposits they are the 17,24% and nowadays only the 3,01%. This same phenomenon happens with the relative number of mammal species inversely to the bird ones.

Percentage of species from each group in different contexts % of each group in extant species % of each group in fossil species % of each group in fossil species (Without cave deposits)

66,57% 71,15% 45,98% 42,53%

13,46% 9,62% 19,88% 5,12% 4,02% 8,43% 7,47% 5,77%

Amphibians Reptiles Birds Mammals

Figure 2.4: Percentage of species from each group in extant, fossil and fossil species (without cave deposits).

2.4 Discussion

An interesting feature of the data is the similarity between the numbers of amphibians and the reptiles. Between both groups there is only one locally extinct species, Testudo hermanni, all other species appear nowadays in Portugal. This feature is congruent with the proposed behavior for the herpetofaunal communities across the Quaternary (Blain et al., 2016; Bisbal-Chinesta & Blain, 2018): At the beginning of the Pleistocene, with most of the mountain ranges as barriers going from East to West and the Mediterranean and the Black Sea in the South preventing movements towards warmer areas, the “exotic” (and most vulnerable to climate cooling) taxa of Europe died out relatively quick, for example the giant chelonians that appear since the Miocene in Europe (Pérez-García et al., 2017) therefore only the most resilient species to cold conditions survived this first impact, enduring again and again later glaciations without losing more species, although there are some exceptions, like the varanids that only perished in the Middle Pleistocene (Georgalis et al., 2017) or relict thermophilic species like Chioglossa lusitanica Bocage, 1864 that endured since the Neogene the cold moments isolated in refugia (Alexandrino et al., 2001). Being a higher metabolic rate related with the rate of evolution (Martin & Palumbi, 1993) the capacity of the herpetofauna for speciation events was quite low during this period, even more taking into account the impoverish starting number of species. Both the percentage of species compared to the overall number of tetrapods and the representativeness of the abundance of these groups are quite reliable measures of their diversity. The percentage of species compared to the total is close to the same number in extant, fossil species and fossil species without cave deposits as we see in figure 2.4. Also, with around 40% of the extant species represented in the fossil record their representativeness is close to the average of all tetrapod species combined.

Now talking about birds we have to bear in mind that their mobility make the number of species present in the Portuguese territory change constantly, either related to seasonality (As migratory species displace North to breed and South to winter) or to mere chance (for example with American species that cross the Atlantic). Therefore the number of nesting species seem the closest proxy to the number of “resident” birds that we can found. Their importance relative to the total number of species of tetrapods today is great; with 221 species, they comprise 2/3 of tetrapods. But this numbers are not reflected in the fossil record, with only 80 species that represent less than half of the fossil tetrapods. This is probably related to several factors, like the lack of interest of many researchers in their bones until relative recent years, their fragility given the pneumaticity of them and the difficulties to identify to the specific level (like for example a lot of Passeriformes Linnaeus, 1758, whose skeletons are remarkable similar).

If we remove the cave deposits the situation is even more pronounced. Only three species “survive” comprising just 5,77% of all tetrapods. This situation could be explained for the accentuation of all the after mentioned factors in the more difficult preservation environment out of the caves. It is interesting to note that this situation is quite similar to what we see in the Cenozoic fossil bird record in Portugal with only 6 sites with bird remains, most of them impossible to ascribe at specific or even generic level, although in the Eocene of Silveirinha appears Fluviatilavis antunesi Harrison, 1983; in the Miocene site of Olival da Susana appears Palaeoperdix media Milne-Edwards, 1871 (Figueiredo, 2010) and in Costa da Caparica, also with Miocene chronology, a large sternum assigned tentatively to Pelagornis miocaenus Lartet, 1857 (Mayr et al., 2008). Therefore is not risky to assume that there is a huge hidden diversity of birds across the Cenozoic, even in the Quaternary, given that the detected species in the fossil record that nowadays live in Portugal only reach about the 30% of the total.

Mammal data have the most complex interpretation of all; they are extremely overrepresented in the fossil record without cave deposits (More than 70%), if we include the later they still comprise about half of the total tetrapod species, but nowadays they only represent 19% of the species (as we see in figure 2.3). This fact could be explained with the intersection of three factors: More than the 45% of the fossil mammal fauna is extinct or locally extinct and this large extinction lowered the overall number of species today. Also between them we have the biggest land animals in the Quaternary of Portugal, with dense and large bones that are more likely to be found even in less favorable conditions outside the cave deposits, like in alluvial terraces. Finally they have been subject of great scientific interest for 150 years in the country, since the first excavations of Carlos Ribeiro and Nery Delgado. These factors could also explain why around 60% of the actual species have been detected in the fossil record; with only small carnivores, “insectivores” and mostly bats being missing; although in the realm of extinct species there is a very real possibility of finding some known megafauna in the Iberian Peninsula but not in Portugal yet, like bison or cave bear.

Finally, if we exclude the cave deposits we can find a good picture of the importance of those in our knowledge of the paleobiodiversity of the Quaternary. Without them we pass from 174 species to 52; a fall of around the 70% if we do not count with the fountain of fossils that is the karst. This is congruent with what we see in the older Cenozoic fossil record, for example in the Miocene pseudokarstic infillings of Cerro Batallones, Madrid that produce a rich fauna otherwise impossible to find (Morales & Baquedano, 2017). For contexts where we don´t have that type of deposits (for example the Jurassic of the Lusitanian Basin) our sampling can be quite reduced.

2.5 Conclusions

- Most of the vertebrate fossil record of the Quaternary of Portugal (with emphasis in the Pleistocene) is situated in the Lusitanian and Algarvian basins. Only 3 out of 30 sites are outside those areas and 20 out of 30 sites are situated in caves. - Quaternary fossil herpetofaunal communities in Portugal evolve across the period with reduced numbers and only one locally extinct species in the area. - ~70% of living bird species are undetected in the fossil record due to their fragile bones, the difficulties in their identification and the lack of scientific interest. - 46% of mammal species known in the Quaternary fossil record of Portugal became locally extinct in the area or went totally extinct. They are also the better known group of Quaternary fossil tetrapods in the country, with most of the living species (60%) appearing in the fossil record. - The importance of the cave deposits is capital in our understanding of the Quaternary vertebrate fauna, without it we would not known about most of tetrapod species, not to mention facts about their paleobiology or paleoecology (Like their denning behavior). The tetrapod living species detected in the fossil record would fall from 37,65% (125 species) to only a 10,54% (35 species) without them.

3.THE PALAEOLOXODON ANTIQUUS OF SANTO ANTÃO DO TOJAL

3.1 Introduction and methods

The proboscidean fossil record of the Quaternary in Portugal is poor if we compare it with the record of neighboring Spain (Ros-Montoya, 2010) and similar sized countries like Greece (Tsoukala et al., 2011) or Serbia (Lister et al., 2012), or even with the Miocene record of Portugal itself (Antunes, 1989b). As we have seen before, this group is represented in the Quaternary of continental Portugal by isolated dental remains and a handful of bones, none of them really well preserved (Antunes & Cardoso, 1992). Between this absence of articulated (or even associated) skeletons, the occurrence of Santo Antão do Tojal is by far the most complete of a proboscidean of this age in Portugal, including an almost complete femur, a partial tibia, phalanx and neural spine, assigned to Elephas (Now Palaeoloxodon) antiquus (Zbyszewski, 1943, 1977b). On this chapter, for first time on a Portuguese proboscidean, an allometric equation (Larramendi, 2015) will be used in order to give an estimation of the body mass of the animal when it was alive. Also those newly obtained measurements of this specimen will be compared with an extensive sample of 47 femurs and 29 tibias of other European proboscideans in order to investigate the metric differences in the hind limb of Palaeoloxodon antiquus and Mammuthus primigenius. In base of these two new sets of data some conclusions regarding this individual specimen and both species will be taken. This section has been partially published in the “VII Congresso de Jovens Invertigadores em Geociências” (Estraviz-López & Mateus, 2018d).

3.2 Geographical and geological settings

In February of 1941 during a visit to a recently dug channel in Loures, Georges Zbyszewski found several bone fragments. Two days later an excavation was carried on the margin of the channel and about 2/3 of a right femur, a partial tibia and neural spine of a proboscidean were dug; alongside several lithic tools, a horse tooth and a probable hyena coprolite (Zbyszewski, 1943). In 1977 after the cleaning of the vegetation of the channel, Zbyszewski returned to the site. He found a mutilated phalanx of proboscidean in the other margin of the channel that he considered belonging to the same animal (Zbyszewski, 1977b). All of them were ascribed by him to Palaeoloxodon antiquus. The exact location of the findings is relatively unknown, but by the description provided (“1080m South, 27° West of Santo Antão do Tojal Church”) the probable coordinates are 38° 50' 27" N, 9° 8' 38" W. In the figure 3.1 we can see the approximate location of the site, alongside the channel side.

Figure 3.1: The location of the Santo Antão do Tojal Palaeoloxodon antiquus is marked by the pushpin, in the Lisboa Peninsula (left) and at local level (right). Credit Google Earth.

According to Zbyszewski, (1943) the fluvial terrace where the specimen was found was subject to the regular floodings from the estuary of the Tagus River, until the construction of channels to drain the area. The log of the locality appearing in the same work is the following, from top to bottom. No measurements of each layer were given:

“4 — Pink sandy silts with in-situ Upper Paleolithic and reworked Mousterian industries . 3 — Pink or red sands with Mousterian industries.

2 — Alternation between greenish clays and yellow-orange sands with iron levels, Mousterian industries and bones, among which those from Elephas antiquus.

1— Gravel and reddish claystone (Oligocene layers reworked).”

The material of Santo Antão do Tojal was stored in the Geological Museum of Lisbon, but no catalog number was provided in any of the works referring to it.

In Zbyszewski (1943) several measurements of the femur were taken and then compared to six specimens ascribed to Mammuthus primigenius, one to Palaeoloxodon antiquus and two to Mammuthus meridionalis (Nesti, 1825). Meanwhile the tibia was also measured and compared with five specimens ascribed to Mammuthus primigenius, one to Palaeoloxodon antiquus and one to Mammuthus meridionalis. In base of these metric comparisons and the assumed age of the terrace (“…belongs undoubtedly to a lower grimaldian terrace (Riss-Würm interglacial). It shows evidence of local backfilling dating of this time and posterior to the digging contemporary from the rissian glacial epoch…”) these remains were ascribed to Palaeoloxodon antiquus. In Zbyszewski, (1977b) the mutilated phalanx was presented and its measurements given. In Cardoso, (1993a) new measurements of the femur were presented. Raposo, (1995) mentioned that the Palaeoloxodon antiquus remains of Santo Antão do Tojal were dated by the method of Uranium-Thorium or Uranium series, yielding a datation of 81.900 +4000 - 3800 YBP. Finally this proboscidean has been cited sometimes in reviews about the Quaternary Portuguese proboscideans like Antunes & Cardoso, (1992) or Figueiredo, (2012) that agreed with the previously given identification.

3.3 The material

As we have seen before all the materials (Femur, tibia, vertebral apophysis and phalanx) is part of the exhibit of the Geological Museum in Lisbon, where it is situated in the archeological part of the museum.

Right femur (MG 5788)

The proboscidean femur of Santo Antão do Tojal (Figure 3.2) is the most complete of all the bones of the site, being it relatively well preserved compared to the other material. Only its distal part, including the condyles, is missing. It is displayed in a standing position with the distal broken part integrated into stand and two metallic rings securing it, all of this preventing the specimen of being removed from the display structure. The trochanteric fossa is deep and well marked, character that point towards P. antiquus, given its robustness and marked muscular insertions in bones (Larramendi et al., 2017).

Figure 3.2: General view of the Palaeoloxodon antiquus femur of Santo Antão do Tojal (MG 5788) at display in the geological museum in posterior view. Scale bar 7 cm. The chalk mark signals a cut in the bone where a stone tool was found, being it regarded by Zbyszewski as a proof that the animal was butchered by Neanderthals.

Right tibia (MG 5793)

The tibia is in worse shape than the femur, being its distal end totally missing. The proximal one has been clearly restored with plaster and has lost fragments. The tibia is displayed the same way than the femur, with the distal and more damaged part embedded into the stand and a metal ring securing the bone in an upright position that makes it impossible to take the piece out. In figure 3.3 we can see the tibia in anterior view. Given the position of the medial articular surface (the best preserved feature of the proximal part) relative to the tibial depression (marking the anterior part of the bone); the tibia was a right exemplar (Shil et al., 2013). In Figure 3.4 we can see those features in the tibia.

Figure 3.3: Right tibia of the Palaeoloxodon antiquus of Santo Antão do Tojal (MG 5793) at display in the Geological Museum in anterior view. Scale bar 7 cm.

Figure 3.4: Proboscidean tibia of Santo Antão do Tojal (MG 5793) in left, anterior and right proximal views. The principal anatomical features are marked, in white the tibial depression and in red the medial articular surface. Scale bar 7 cm.

Vertebral spinal process (MG 8081)

The incomplete spinal process of a vertebra from the proboscidean is 390 mm tall, despite that the tip of it is broken. Meanwhile, the sides of the neural canal are better preserved, being it 165 mm wide. It is impossible to be totally sure without the whole vertebra, but given the height of this process the vertebra was a thoracic one (Larramendi et al., 2017). In figure 3.5, we can see it.

Figure 3.5: The spinal process of the proboscidean of Santo Antão do Tojal (MG 8081) in A) anterior and B) posterior views.

First phalanx (MG 8080)

The phalanx of the Proboscidean of Santo Antão do Tojal is lightly damaged, as we can see in figure 3.6. In the table 3.1 we can see its measurements. It is a first phalanx according to Zbyszewski, (1977b).

Figure 3.6: The phalanx from the proboscidean of Santo Antão do Tojal in A) dorsal; B) lateral; C) ventral and D) lateral views.

Table 3.1: Measurements of the phalanx of the proboscidean of Santo Antão do Tojal. “Thickness” refers to anteroposterior diameter of the bone and “width” to the mediolateral diameter of it. All the measurements are in millimeters.

Work Zbyszewski, 1977, this work Site Santo Antão do Tojal Specimen Santo Antão do Tojal Species attributed Palaeoloxodon antiquus Number First Maximum length 100 Proximal width 75 Proximal thickness 67 Distal width 50 Distal thickness 47

3.4 Measurements of femur and tibia

The measurements taken of the femur and tibia were the ones used by Larramendi et al. ( 2017); which are similar to the ones of Von den Driesch, (1976) for the hindlimb of mammals. On them “width” refers to medio-lateral diameter of the bone and “thickness” to the anteroposterior diameter of it. We can see the measurements graphically in the figures 3.7 and 3.8. In the tables 3.2 and 3.3 we have the measurements of the Santo Antão do Tojal material as well a collection of measurements that have been taken from literature for other femora and tibiae of Palaeoloxodon antiquus and Mammuthus primigenius.

Figure 3.7: Measurements of the femur of mammals by Von den Driesch, 1976.

Table 3.2: Measurements of 48 individual femurs of Palaeoloxodon antiquus and Mammuthus primigenius taken from the literature. The specimen of Santo Antão do Tojal is marked. All measurements are in millimeters.

Minimum Minimum Minimum Distal Species Maximum Proximal Caput Caput diaphysis diaphysis diaphysis Distal articular Work Site attributed length width width thickness width thickness circumference width width Larramendi Neumark- Palaeoloxodon et al., 2017 Nord 1 antiquus 1320 425 205 210 156 121 470 318 - Larramendi Neumark- Palaeoloxodon et al., 2017 Nord 1 antiquus 1398 - 216 205 - - - - - Larramendi Neumark- Palaeoloxodon et al., 2017 Nord 1 antiquus 1040 - 155 159 116 95 340 229 - Larramendi Neumark- Palaeoloxodon et al., 2017 Nord 1 antiquus 1070 338 146 159 121 86 332 244 192 Tsoukala et Palaeoloxodon al., 2011 Sotiras antiquus ------330 290 Tsoukala et Palaeoloxodon al., 2011 Xerias antiquus - - - - 124 289 - - - Kevrekidis Palaeoloxodon & Mol, 2016 Amyntaio antiquus - - 215 217 - - - - - Palaeoloxodon Mazo, 1998 Buelna antiquus ~ 1320 - - - 175 130 460 320 - Palaeoloxodon Mazo, 1998 Torralba antiquus 1210 - - - ~175 ~130 - ~320 - Konidaris et Marathousa Palaeoloxodon al., 2018 1 antiquus 1330 ------Marra, 2009; Beuval et Palaeoloxodon al., 1998 Gröbern antiquus 1402 - 224 - 173 - - 308 - Beauval et Palaeoloxodon al. 1998 Gröbern antiquus - - 179 - - - - 234 - Palaeoloxodon Marra, 2009 Crumstad antiquus 1040 - 155 - 116 - - 208 - Marra, 2009 Riano Palaeoloxodon 1290 - - - 155 - - 283 -

antiquus Palaeoloxodon Marra, 2009 Viterbo antiquus 1370 - 200 - 192 - - 247 - Marra, 2009; Larramendi Palaeoloxodon et al., 2015 Viterbo antiquus 1440 - 242 - 242 - - 277 - Andrews & Cooper, Palaeoloxodon 1928 Upnor antiquus ~1500 467 214 ------Jakubowski, Palaeoloxodon 1988 Konin antiquus 1429 470 - - 157 130 - 323 275 Melentis, Palaeoloxodon 1963 Megalopolis antiquus 1372 476 - - 170 117 478 298 - Jakubowski, Palaeoloxodon 1996 Grabsch tz antiquus 1130 ------Christiansen, Palaeoloxodon 2004 ? antiquus 1075 - - - 136 102 373 - - Christiansen, Palaeoloxodon 2004 ? antiquus 1169 - - - 128 94 348 - 224 Zbyszewski, Santo 1943; this Antão do Palaeoloxodon work Tojal antiquus - 400 180 190 165 125 470 - - Harington et Mammuthus al., 2012 Muirkirk primigenius 1002 - - - 111 - 313 280 - Harington et Mammuthus al., 2012 Muirkirk primigenius 1010 - - - 121 - 327 - 185 Kirillova et Mammuthus al., 2012 Taymir primigenius 880 - 132 132 108 70 - 202 - Kirillova et Mammuthus al., 2012 Taymir primigenius - - - - 110 74 - 209 - Maschenko Mammuthus et al., 2006 Svesk primigenius 790 - - - - - 158 - - Maschenko Mammuthus et al., 2006 Svesk primigenius 1010 - - - - - 210 - - Maschenko Mammuthus et al., 2006 Berelekh primigenius 850 - - - - - 100 - -

Maschenko Mammuthus et al., 2006 Berelekh primigenius 1130 - - - - - 150 - - Álvarez-Lao Mammuthus et al., 2009 Padul primigenius 1060 314 147 152 142 73 - - - Álvarez-Lao Mammuthus et al., 2009 North Sea primigenius 810 245 110 120 95 60 - 165 - Álvarez-Lao Mammuthus et al., 2009 North Sea primigenius 1400 408 190 183 176 109 - 275 - Christiansen, Mammuthus 2004 Hebior primigenius 1203 - - - 160 80 377 - 229 Christiansen, Mammuthus 2004 Brussels primigenius 1273 - - - 158 108 419 - 287 Christiansen, Mammuthus 2004 Paris primigenius 1071 - - - 128 88 339 - - Christiansen, Mammuthus 2004 FMNH primigenius 1111 - - - 126 88 337 - 195 Christiansen, FMNH Mammuthus 2004 PM26267 primigenius 1014 - - - 151 98 391 - 227 Larramendi, Mammuthus 2015 Siegsdorf primigenius 1330 - - - - - 435 - - Larramendi, Mammuthus 2015 Rottweil primigenius 945 - 140 - 126 - 323 - 98 Mammuthus Lister, 2009 Condover primigenius 1165 - 161 - - - - 234 211 Mammuthus Lister, 2009 Condover primigenius 1190 - 158 - - - - 242 207 Demay, Changis- Mammuthus 2014 sur-Marne primigenius 1100 349 150 - 142 - - 198 - Demay, Changis- Mammuthus 2014 sur-Marne primigenius 1200 320 156 - 122 - - - - Petrova, et Mammuthus al., 2017 Ahlen primigenius 1240 - 183 ------Petrova, et Mammuthus al., 2017 Khoma primigenius 970 - 150 - 125 - - 177 - Petrova, et Mammuthus al., 2017 Berezovka primigenius 1050 - 153 - 122 - - 226 -

Figure 3.8: Measurements of the tibia of mammals by Von den Driesch, 1976.

Table 3.3: Measurements of 30 individual tibias of Palaeoloxodon antiquus and Mammuthus primigenius taken from the literature. The specimen of Santo Antão do Tojal is marked. In order to fit the table, abbreviations have been used: ML, maximum length; AL, articular length; PW, proximal width; PT, proximal thickness; PAW, proximal articular width; PAT, proximal articular thickness; MDW, minimum diaphysis width; MDT, minimum diaphysis thickness; MDC, minimum diaphysis circumference; DW, distal width; DT, distal thickness; DAW, distal articular width; DAT, Distal articular thickness. All measurements are in millimeters.

Species Work attributed ML AL PW PT PAW PAT MDW MDT MDC DW DT DAW DAT Larramendi et al., Palaeoloxodon 2017 antiquus 795 775 278 - 263 163 151 136 440 232 180 231 62 Larramendi et al., Palaeoloxodon 2017 antiquus 638 625 217 171 210 152 94 80 280 183 130 169 52 Larramendi et al., Palaeoloxodon 2017 antiquus 842 820 286 234 274 183 140 139 430 252 182 215 67 Larramendi et al., Palaeoloxodon 2017 antiquus 914 794 312 - 274 164 152 - - 237 - 205 63 Larramendi et al., Palaeoloxodon 2017 antiquus 576 490 208 150 188 123 98 83 - 170 132 153 47 Larramendi et al., Palaeoloxodon 2017 antiquus 624 598 205 143 192 139 89 87 288 167 133 145 67 Kevrekidis & Palaeoloxodon Mol, 2016 antiquus ------114 105 335 - - - - Konidaris et al., Palaeoloxodon 2017 antiquus 792 ------360 - - - - Andrews & Forster-Cooper, Palaeoloxodon 1928 antiquus 1020 - 281 183 - - 150 - - 230 - - - Van Kolfschoten, Palaeoloxodon 1993 antiquus ~775 - ~265 183 245 - 115 113 - ~200 - - - Van Kolfschoten, Palaeoloxodon 1993 antiquus 602 - 208 ------Van Kolfschoten, Palaeoloxodon 600 - 198 ------

1993 antiquus Palaeoloxodon Larramendi, 2015 antiquus 880 ------435 - - - - Palaeoloxodon Athanassiou, 2001 antiquus ------108 85 - 177 136 - - Zbyszewski, Palaeoloxodon 1943; this work antiquus - - 270 190 - - 110 125 400 - - - - Christiansen, Mammuthus 2004 primigenius 546 - - - - - 80 66 230 - - - - Christiansen, Mammuthus 2004 primigenius 499 - - - - - 81 70 237 - - - - Christiansen, Mammuthus 2004 primigenius 548 - - - - - 87 74 252 - - - - Mammuthus Larramendi, 2015 primigenius 800 ------341 - - - - Mammuthus Larramendi, 2015 primigenius 540 ------248 - - - - Kirillova et al., Mammuthus 2012 primigenius 501 - 174 149 - - 78 70 - 139 117 - - Kirillova et al., Mammuthus 2012 primigenius 500 - 182 148 - - 78 69 - 141 117 - - Mammuthus Lister, 2009 primigenius 680 - 228 180 - - - - - 180 147 - - Mammuthus Lister, 2009 primigenius 685 - 227 181 - - - - - 182 141 - - Harington et al., Mammuthus 2012 primigenius 588 - 197 - - - 90 - 259 123 - - - Harington et al., Mammuthus 2012 primigenius 595 - - - - - 90 ------Mammuthus Adams, 1881 primigenius 597 - 266 - - - - - 266 - - - - Mammuthus Adams, 1881 primigenius 508 ------Mammuthus Adams, 1881 primigenius 558 ------272 - - - - Mammuthus Demay, 2014 primigenius 640 - 217 159 - - 102 - - 153 144 - -

3.5 Body mass estimation

The body mass of an organism is one of the most important parameters for understanding its biology, being it closely related to metabolic rate, reproductive strategy, development and many other traits. But reconstructing the body mass of dead animals, sometimes of extinct species is a challenge. Traditionally, allometric equations based on the living proboscideans have been used, but they are not reliable for inferring the masses of long gone species, with different proportions and much larger or smaller masses. Recently the volumetric methods have started to be recognized as more accurate for reconstructing the masses of such animals; although this method has its own shortcomings: Full body and precise restorations based on well preserved and complete specimens are needed to achieve reliable results (Larramendi, 2015; Jukar et al., 2018). Considering the incompleteness of the Santo Antão do Tojal proboscidean, the total length for both bones discovered was inferred from the equation developed by Christiansen (2007). The shoulder height of the animal was calculated in base of the specific ratios for Palaeoloxodon antiquus, developed by Larramendi (2015). Finally in base of a specific allometric equation developed from volumetric reconstruction of better preserved specimens presented in that work, a body mass is given.

To estimate the body mass of the Santo Antão do Tojal proboscidean the total length for both bones should be inferred first thanks to the equation developed by Christiansen, (2007):

log Y = α + β log X

“Y” represents the maximum length of the bone in mm, “X” its least circumference in mm and “ α ” and “β” represent two constants, empirically parameterized either for Elephantidae Gray, 1821 (femur) or for proboscideans in general (tibia). Considering the measurements of the least circumference for both bones (470 mm for the femur and 400 for the tibia), the above equation provides an estimation for the lengths of the femur (1388 mm) and tibia (865 mm). To test how accurate these estimations are we can look to table 3.4 were they are compared with other similarly sized specimens in the dataset (ratio of maximum length vs minimum diaphysis circumference of the hind limb bones). In table 3.5 we have the ration between femur and tibia total lengths.

Table 3.4: Maximum length vs minimum diaphysis circumference ratios of femur and tibia, for selected Palaeoloxodon antiquus specimens. All measurements in mm.”E” stands for “estimated”.

FEMUR Minimum diaphysis Maximum length vs diaphysis Work Specimen Maximum length circumference circumference ratio Larramendi et al., 2017 152-E9 1320 470 2,81 Mazo, 1998 ? 1320 460 2,87 Melentis, 1963 AMPG 1960/32 1372 478 2,87 Santo Antão do This work Tojal 1388 E 470 2,95 TIBIA Minimum diaphysis Maximum length vs diaphysis Work Specimen Maximum length circumference circumference ratio Larramendi et al., 2017 152-E9 795 440 1,81 Larramendi et al., 2017 167-E43A 842 430 1,96 MAR-1-940/675- Konidaris et al., 2017 50 792 360 2,20 Larramendi, 2016 Konin 880 435 2,02 Santo Antão do This work Tojal 865 E 400 2,16

Table 3.5: Femur vs tibia lengths for selected Palaeoloxodon antiquus specimens. All measurements in mm.”E” stands for “estimated”.

Maximum length Maximum length Maximum length of femur vs Maximum length of Work Specimen femur tibia tibia ratio Larramendi et al., 2017 152-E9 1320 795 1,66 Larramendi et al., 2017 167-E43A 1398 842 1,66 MAR-1-940/675- Konidaris et al., 2017 50 1330 792 1,68 Andrews & Cooper, 1928 Upnor 1500 E 1020 1,47 Larramendi, 2015 Konin 1429 880 1,62 Santo Antão do This work Tojal 1388 E 865 E 1,60

Looking to table 3.6 we can see that the estimated maximum length of femur is slightly larger than the other specimens (greater length for the same circumference) but does not fall far from them. The tibia is also in the upper part of the estimation but inside the variation of proportions of the species. If we look at table 3.7 we can see that the tibia is larger for the same length of the femur than other similarly sized animals, but again it falls in the variation of proportions for the species.

According to Larramendi, (2015) the shoulder height of a given individual Palaeoloxodon antiquus is about 2,56 times the femur length and 4,15 the tibia length, which yields 3553 mm and 3589 mm of skeleton shoulder height respectively. The average of this estimation gives a result of 3571 mm or 3,57 meters of shoulder height for the skeleton. According to Larramendi et al., (2017) for animals with similar proportions to the Santo Antão do Tojal specimen the shoulder height with flesh for the species should be about 20 cm larger than the skeletal one, therefore the shoulder height in life is estimated in 377 centimeters or 3,77 meters. This estimated value is used in the equation presented in the same work for the body mass of average specimens of Palaeoloxodon antiquus:

Body mass in kilograms = 3,63 × 10-4 × (Shoulder height in cm)2,903

Once solved, the equation yields a result of 10940 kg or about 11 metric tons. On table 3.6 we can see a comparison of these measurements with other specimens of Palaeoloxodon antiquus from different sites. It must be noted that NOT all the shoulder heights and body mass estimations in the table have been obtained by the same methods.

Table 3.6: Shoulder height and body masses of different Palaeoloxodon antiquus specimens. The specimens below to the Santo Antão do Tojal one had not been calculated using the same methods as presented. Possible males are marked in blue.

Work Site Specimen Shoulder height (cm) Body mass (metric tons) Larramendi et al., 2017 Neumark-Nord 1 152-E9 363 9,5/9,8 Larramendi et al., 2017 Neumark-Nord 1 167-E43A 385 11,4/11,6 Larramendi et al., 2017 Neumark-Nord 1 151-E8 264 3,9 Larramendi et al., 2017 Neumark-Nord 1 171-E34A 295 4,5/5,4 Marra, 2009; Larramendi et al., 2016 Viterbo Fontana Campanile 381 11,3 This work Santo Antão do Tojal Santo Antão do Tojal 377 11 Kevrekidis & Mol, 2016 Amyntaio PHP AME 046/006 355 9,2 Konidaris et al., 2017 Marathousa 1 MAR-1-940/675-50 370 9 Jakubowski, 1996 Grabsch tz Grabsch tz 301 5,7

The Santo Antão do Tojal specimen seems to be average sized for a big male of this species, only three centimeters below the expected average of 380-420 cm shoulder height (Larramendi, 2015). It is almost as big as the Fontana di Campanile specimen of Viterbo and the specimen 167-E43A from Neumark-Nord; being the age of the last one estimated in 47 years (Larramendi et al., 2017).

3.6 Analysis of metrical data

Completeness of the sample

There are two reasons why the sample is incomplete: First, some bones (like the Santo Antão do Tojal specimen) are incomplete themselves and second, not all measurements considered here are presented in the literature, even in the case of entire specimens. The completeness of the sample can be quite problematic for its representativeness. In the case of the femur (Table 3.2) of 384 possible individual measurements there are 198 represented, or 51,56% of them. For the tibia (Table 3.3) of 360 possible individual measurements there are 163, representing 45,27% of them. But not all measurements are equally represented, for example some (like the maximum length of the tibia) appear in 90% of the specimens, meanwhile others (like the proximal articular thickness of the tibia) appear only in 20% of the specimens. In the table 3.7 we can see the number of individual measurements of the bone and the % of specimens in which they appear.

Table 3.7: Number of individual measurements taken for proboscidean femurs and tibias of the considered sample. Percentage of specimens where the measurement has been taken compared to the total is also presented. The measurements that have been taken in 50% or more of the specimens are marked.

FEMUR % of specimens were the measurement has been Measurement Number of individual measurements taken Maximum length 42 87,50 Proximal width 11 22,92 Caput width 24 50 Caput thickness 10 20,83 Minimum diaphysis width 33 68,75 Minimum diaphysis thickness 21 43,75 Minimum diaphysis circumference 21 43,75 Distal width 24 50 Distal articular width 12 25 TIBIA

Maximum length 27 90 Articular length 6 20 Proximal width 18 60 Proximal thickness 12 40 Proximal articular width 7 23,33 Proximal articular thickness 6 20 Minimum diaphysis width 19 63,33 Minimum diaphysis thickness 14 46,67 Minimum diaphysis circumference 16 53,33 Distal width 15 50 Distal thickness 11 36,67 Distal articular width 6 20 Distal articular thickness 6 20

Univariate statistical analysis

After identifying the most promising measurements for analyzing the data, those were investigated in R, version 3.2.2 of 64 bits; running in a computer Toshiba Satellite L655-1JV of 64 bits with an Intel Core i5 and Windows 7 Home Premium as operative system. The measurements (table 3.4) are well distributed across the bone, with at least one measurement in the proximal and distal epiphysis and one in the diaphysis. In the table 3.8 we have the mean of each measurement and species for femur and tibia, with a % in difference between those averages.

FEMUR

Minimum diaphysis Maximum length Caput width width Distal width

Mammuthus primigenius 1.075,17 152,50 130,76 220,80

Palaeoloxodon antiquus 1.272,50 194,25 156,31 281,36

% Difference 15,51 21,49 16,34 21,52

% Overall difference 18,72

TIBIA

Minimum diaphysis Minimum diaphysis Maximum length Proximal width circumference width Distal width

Mammuthus primigenius 585,67 213 263,13 85,75 153

Palaeoloxodon antiquus 754,83 248 371 120,09 205,33

% Difference 22,41 14,11 29,08 28,60 25,49

% Overall difference 23,94

Looking at this table (table 3.5) we can see that on average the femur of Palaeoloxodon antiquus is 18,72% larger than the one of Mammuthus primigenius and the tibia is 23,94% larger. In figures 3.8 (Femur) and 3.9 (Tibia) there is the plot of means for each selected measurement according to the bone and species; with an interval of confidence of 95%.

Figure 3.9: Plot of means for the previously selected measurements of the femur of Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with an interval of confidence of 95%. All measurements are in millimeters.

In figure 3.9 we can see that there is no overlap in the 95% confidence intervals in all the selected measurements of the femur except in the minimum diaphysis width.

Figure 3.10: Plot of means for the previously selected measurements of the tibia of Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with an interval of confidence of 95%. All measurements are in millimeters.

As we see in figure 3.10, we can detect a reliable difference in all selected measurements of the tibia except in the proximal width of the bone. Therefore, we are left with measurements for each bone that are keen to distinguish between the two species. In the femur: maximum length, caput width and distal width. In the tibia: maximum length, minimum diaphysis circumference, minimum diaphysis width and distal width.

Of those seven measurements (Three in the femur and four in tibia) it is possible to record three in the Santo Antão do Tojal specimen; one in the femur (Caput width) and two in the tibia (Minimum diaphysis circumference and width). In figure 3.11 we can see the measurements of the Santo Antão do Tojal specimen compared with the means of both bones for those measurements, falling it into the variability of Palaeoloxodon antiquus in each case.

Figure 3.11: Measurements of Santo Antão do Tojal specimen (in red) compared to the means for them in Mammuthus primigenius (left) and Palaeoloxodon antiquus (right) with a 95% confidence interval.

In some cases it seems that the range of variability of Palaeoloxodon antiquus is greater than in Mammuthus primigenius, so now we will investigate it by using the density estimation of the two most numerous measurements in each bone.

Figure 3.12: Estimation of density for maximum length and minimum diaphysis width of the femur of Mammuthus primigenius and Palaeoloxodon antiquus.

For the femur (Figure 3.12) we can see that the maximum length for Mammuthus primigenius follows an almost perfect normal distribution around the mean of 1075 mm of length, meanwhile Palaeoloxodon antiquus follows a bimodal distribution with peaks at 1100 mm and 1400 mm of total length. Looking at the minimum diaphysis width we can see that in Mammuthus primigenius it does not have neither a perfect bimodal or normal distribution, but a grouping of measurements at 120 mm and 150 mm meanwhile in Palaeoloxodon antiquus the bimodal distribution is clearer, with one slope around the 125mm and another around 175mm.

Figure 3.13: Estimation of density for maximum length and minimum diaphysis width of the tibia of Mammuthus primigenius and Palaeoloxodon antiquus

For the tibia (Figure 3.13) we can see for the maximum length in Mammuthus primigenius a relatively weak bimodal distribution grouped mostly around 500 mm of total length, with fewer measurements somewhat around 600 mm. Meanwhile for Palaeoloxodon antiquus, the bimodal distribution is again clear, with the measurements grouped around 600 and 800 mm. Looking at the minimum diaphysis width, the pattern is more complex: Mammuthus primigenius has three peaks of density around 80, 90 and 100 mm, when the mean is 85 mm. Palaeoloxodon antiquus bimodal distribution has peaks at 110 and 150 mm with the mean at 120 mm.

This bimodal distribution is interpreted therefore as proof of the more marked sexual dimorphism in Palaeoloxodon antiquus.

Multivariate statistical analysis

To investigate the sample in multivariate statistic the program Past (version 3.14) was used, running in the same hardware as previously specified. This program allows running PCA (Principal Component Analysis) without all data by replacing the missing data with the mean of the column. The results reached by this technique should be therefore taken with caution.

First we will investigate the femur dataset. We can see that the three first principal components account for more than 92% of the variation, with the first one alone amassing 80% of it. Then we can see that the PC1 is related strongly with the maximum length of the bone and PC2 with its minimum diaphysis circumference, meanwhile PC3 is a mix of different measurements with little differences between them, as shown by the bootstrapping of N=100000 (Figure 3.14).

Figure 3.14: From upper to bottom: Percentage of variation accounted by the first three Principal Component of the femur; PC1, PC2 and PC3. In PC1 and PC2 the highest contribution comes from the maximum length and the minimum diaphysis circumference respectively.

Plotting the PC1 versus the PC2 (Figure 3.15) we can see that for the same values of PC1 the Palaeoloxodon antiquus specimens tend to score higher in PC2, therefore having greatest circumferences and overall robustness; meanwhile the lower PC1 scores of all are obtained by Mammuthus primigenius individuals (shortest length of the femur).

Figure 3.15: PC1 vs PC2 of the femur of selected proboscideans. The Santo Antão do Tojal specimen is marked with the red square, Mammuthus primigenius in blue and Palaeoloxodon antiquus in black.

In the tibia dataset we can see a fairly similar situation (Figure 3.16). PC1 and PC2 account for most of the variation. The measurements that contribute the most to them are respectively the same as in the femur: The maximum length for the PC1 and the minimum diaphysis circumference in the PC2. The PC3 is again a mix of different measurements without a clear meaning.

Figure 3.16: From upper to bottom: Percentage of variation accounted by the first three Principal Component of the tibia; PC1, PC2 and PC3. In PC1 and PC2 the highest contribution comes from the maximum length and the minimum diaphysis circumference respectively.

Plotting the PC1 vs the PC2 produces a situation relatively similar to the one in the femur (Figure 3.17). The Palaeloxodon antiquus specimens tend to have more robust tibia (higher PC2) for the same lengths (PC1). It is also interesting to note the greatest variance of the tibia of Palaeloxodon antiquus that could be reflecting the previously detected bimodal distribution in the univariate statistic.

Figure 3.17: PC1 vs PC2 of the tibia of selected proboscideans. The Santo Antão do Tojal specimen is marked with the red square, Mammuthus primigenius in blue and Palaeoloxodon antiquus in black.

3.7 Conclusions

- The remains of the Santo Antão do Tojal proboscidean (Palaeoloxodon antiquus) which is around 80.000 YBP old; comprise a first phalanx, a neural arch and a right femur and tibia. - The Santo Antão do Tojal specimen is estimated to have measured alive about 3,8 meters in the shoulder and weighted nearly 11 tons, being it in the average for a male Palaeoloxodon antiquus. Despite being one of the most recent and Southwestern occurrences of this species, it is nearly identical to specimens that lived several thousand years before in Central Europe or the Italian Peninsula. - The size difference between Mammuthus primigenius and Palaeoloxodon antiquus is 18% on average for the femur and 24% in the tibia. The best measurements for differentiating between both species are in the femur maximum length, caput width and distal width; and in the tibia maximum length, minimum diaphysis circumference, minimum diaphysis width and distal width. - We agree with previous works (Zbyszewski, 1943; Antunes & Cardoso, 1992) in that the Santo Antão do Tojal proboscidean is a Palaeoloxodon antiquus in base of the caput width of the femur and the minimum diaphysis circumference and width of the tibia, combined with a deep trochanteric fossa in the femur and the overall size of the bone and its robustness. - Palaeoloxodon antiquus measurements follow a clearer bimodal distribution than Mammuthus primigenius ones, indicating a more pronounced sexual dimorphism. - The hind limb anatomy of Palaeoloxodon antiquus reflects more robust bones than in Mammuthus primigenius, even for similar sized animals, as other works have shown (Larramendi et al., 2017).

4.ALGAR DO VALE DA PENA: A NEW FOSSIL BEAR SITE FROM PORTUGAL

4.1 Overview

On this chapter a new locality with Ursus remains will be presented. First we will talk in the introduction about the location, geological settings, history of the cave and its morphology and physical characteristics. Then we will talk about the works carried until this moment and the techniques used in the expeditions. In the third part of the chapter we will talk about the bear remains recovered in detail, including comments about the mode of preservation. Then comparisons with the literature (Morphological and metrical) will be done in order to characterize the population of the cave. Next, the claw marks found in the walls of the cave will be presented; being this the first time that such markings are described in Portugal. Finally, there will a discussion about the data and some conclusions about the bear population of the cave as well as some comments about further work to be done in the cavity and with the material recovered.

4.2 Introduction

Location

The Algar do Vale da Pena is located near the village of Moita do Poço, parish of Turquel, municipality of Alcobaça, district of Leiria. It is inside of the natural park “Parque Natural das Serras de Aire e Candeeiros”. In the figure 4.1 we can see its location graphically.

Figure 4.1: Location of the Algar do Vale da Pena. Upper left, at peninsular level, it is situated in the Portuguese Estremadura. Bottom, view at regional level, the closest city is Alcobaça. Upper right, view at local level. Credit Google Earth.

Geological settings

The cave is located in the Northwestern slope of the Candeeiros range, a mountain range that appear between Rio Maior in the Southwest and Porto de Mos in the Northeast; spanning for about 30 kilometers and reaching about 600 meters ASL.

This range is part of the system “Estrela-Montejunto”, a series of ranges that divide Portugal from Southwest to Northeast. It is formed in calcareous rocks of the Maciço Calcário Estremenho, limited between the thrust of Arrife in the East and the fault of Candeeiros in the West, which runs parallel to the Candeeiros range. The rocks that form the cave are part of the Candeeiros Formation, Middle Jurassic in age, concretely Bajocian as we can see in figure 4.2 (Carvalho, 1996; Crispim, 2008).These calcareous rocks are not fossiliferous and are either oolitic or micritic in texture (Figure 4.3). Finally, we should note that there are some detritic deposits of Pliocene age that are located in the lower part of the slope and correspond to a marine transgression (Zbyszewski & de Matos, 1959).

Figure 4.2: Litostratigraphic log of the “Maciço Calcário Estremenho”, with the Candeeiros Formation marked (Carvalho, 1996).

Figure 4.3: Location of the cave (red and orange dot) in the corresponding sheet (26 D, Caldas da Rainha) of the geologic map (1:50000) of Portugal, including legend (Zbyszewski & de Matos, 1959).

History of the cave

This version of the history of the cave has been put together according to personal communications by Manuel Pinto Soares, the first professional speleologist in Portugal, who was also one of the first people to enter the cave, Octávio Mateus, of the FCT-UNL who contacted the park authorities for the authorization to recover the bones and Sérgio Medeiros, president of the GPS (Grupo de Protecção Sicó).

The calcareous rock of the Candeeiros range is exploited by several enterprises in open air quarries inside the natural park since many decades ago (Carvalho, 1996); like the one already active that can be seen south of the site in the figure 4.1. In one of such quarries, at the end of the decade of 1980, there was a rock fall that connected the outside with the cave. The quarry workers thought that maybe they could fill the gap and continue working as nothing happened, so they proceeded to throw several tons of big rock blocks into the newly opened cavity; but after throwing as many blocks as they could into the hole it became evident that they could not fill it.

They contacted with Olimpio Martins, worker and speleologist of the natural park, who told to them to stop the exploitation and realized the first descend into the cave; then the beginning of the decade of 1990 (1991 or 1992) Manel Pinto Soares accompanied Olimpio and installed the nails in the walls of the cave to descend safely. Olimpio told him about the bones at the end of the cave (“bear number four” in this work, as we will discuss later) and after he took a picture, Olimpio said to him to keep the secret about them. In that visit Manuel did not notice any other bone remains despite he passed nearby the main bone assemblage (“bear” number one in this work).

As the years passed the difficult vertical entrance of the cave became popular within the speleological community, as well as the claw marks and bone remains that were inside the cave and earned it the nickname of “Algar dos Ursos” or “Cave of the bears”. Manuel even heard other people talk about “three bears” or “three bear skulls”. During this time the cave was visited by groups of firefighters doing rescue practices and a dog fell into the cave, where its bones can be seen nowadays.

In 2011 some cavernicolous oniscidea (a type of crustacean isopods) were recovered in the cave and later described (Reboleira et al., 2015). In 2012 the GEM (Grupo de Espeleología e Montanhismo) started the works in order to map the previously uncharted cave; this map was completed in august 2015 and will be used later in this work, they also put some security tape to prevent people from stepping over sand deposits and the fossil remains (Amendoeira et al., 2015). Around that time (2013) the scientist of the CAVE project visited the site; it was an international project aiming to understand the cave evolution from an integrative perspective and its use for paleoenvironmental reconstruction from the point of view of several disciplines like archaeology, isotopic studies and geomorphology; but no study centered on this cave came to light (http://caveportugal.wixsite.com/cave/homeengl).

Sérgio Medeiros contacted with the Museu da Lourinhã to make them know about the fossil remains and Octávio Mateus prepared the first of the visits to the cave that would lead to the work presented here.

Morphology of the cave

The cave is not fully mapped yet and we do not have a side view of it. The cartography that will be used was produced by members of the GEM (Grupo de Espeleología e Montanhismo) during 2014 and 2015. In the figure 4.4 we have marked in black numbers some landmarks of the cave like claw marks, small conducts or sand deposits. The location of the bone assemblages is marked with red numbers, they will be referred as individual “bears” in the text but the MNI will be discussed in the next section.

Figure 4.4: Topography of the Algar do Vale da Pena. In black, some landmarks of the cave are marked and in red, the bone findings (Modified from Amendoeira et al., 2015).

The entrance of the cave is not very clearly drawn in the topography, but we have some pictures of it in the figure 4.5. It includes a first ramp of about ten meters of different pending, then a short three meters vertical step and then another ramp of about the same length before the final vertical well of about fifteen meters.

Figure 4.5: A) Entrance of the cave showing the first ramp and the top of two meters step. B) Picture taken from the base of the two meters step, showing the second ramp. C) Final vertical well of the entrance from the bottom. D) Picture from the bottom of the cave showing the final well (white line), the second ramp (green line), the two meter step (red line) and part of the firs ramp (blue line). Pictures by Carlos Ferreira.

The final landing is made over a pile of blocks and dirt thrown by the quarry workers into the cave in their unsuccessful attempt to fill it (1 in the figure 4.4). Once properly in the cave, without the need of ropes anymore, we descend the pile of blocks until we see a bifurcation. If we go right we will reach quickly some bone remains (“bear” number 5 in the figure 4.4) and also a small conduct with claw marks on it and the recent dog skeleton at the end that we can see in the figure 4.6 (2 and 3 in map).

Figure 4.6: Claw marks (left) and dog skeleton (right) in the small conduct at the right of the cave.

Following the main right tunnel of the cave we can go down in the left part of it until we reach a lake or going up sticking to the right wall until we find more bone remains near a bottle neck and the beginning of a new conduct (“bear” number 2 in figure 4.4).

Meanwhile taking the left main tunnel we will pass by a big room with a natural “railing” formed by speleotems that pass to the left of a big depression and then starts climbing. As soon as we turn left from it we will find a panel with even more claw marks and following the easiest path another mount of clay with many claw marks on it (number 4 and 5 in the map). Then going down right from that mount we will find the main fossil findings and six meters at its left another bone assemblage (numbers 1 and 3 in the figure 4.4). We can see sands near these findings (Figure 4.7 and number 6 in the map). Finally after crossing another narrow “railing” we will find the best claw marks in the cave, with the number 7 (Figure 4.7) and the last skeletal remains in the cave (“bear” number 4).

Figure 4.7: Thick deposit of sands below the calcareous crust (left) and claw marks (right).

After this last bone remains the cave starts to climb until it ends in a “cul-de-sac”. The location of the original entrance of the cave is a mystery, and it might be impossible to find it if the path to it was buried by the pile of tons of rocks that opened the current entrance.

4.3 Methods, expeditions to the cave and preparation of the material

Up to the moment of delivering this thesis (March 2019) there has been three expeditions to the cave. In all of them the entrance has been possible thanks to the support of the GPS (Grupo de Protecção Sicó) that offered to the author and accompanists the material and technical support to descend and climb safely by means of vertical progression through the 30 meters depth entrance. To descend, two pieces of equipment called the “shunt” and the “descensor” (Figure 4.8) have to be used. The technique consists in letting the rope pass through the “descensor” controlling the descend velocity with the position of the hand, up to go slow, down to go faster. The “shunt” is a security measurement, if the weight of the person is too far from it, it will lock. Therefore it has to be close to the “descensor” being pulled down with two fingers at the same time that the body of the climber goes down. To climb, the “ventral croll” and the “climbing fist” (Figure 4.8) have to be used at the same time pulling the body up with the help of the later and letting the rope to pass by the first one, then it will lock in the new position, so the climb is done in “steps”. Either going down or up, when the end of a rope is found the climber has to secure himself in the nail of the wall and change the equipment to the new rope, keeping in all moments two elements of security. There are five of these changes in the way up and down of the cave.

Figure 4.8: Equipment needed to descend (in black) and climb out of the cave again (in red). Left; A) Shunt, B) Descensor, C) Climbing fist. Right; D) Ventral croll and security knot to stop descending, note how the fingers secure the shunt.

It has to be noted that since its opening the cave has become an important refuge for bats, so the visit is not allowed by the Natural Park authorities between November to April because the disturbance to hibernating bats can cause them to exhaust their energy reserves and die. Every visit to the cave has been communicated to the corresponding authorities and in April 2017 an agreement was reached between the municipality of Alcobaça, the Natural Park and the FCT-UNL to take out the fossils of the cave and deliver them prepared to the Alcobaça municipality three years after they are collected.

First expedition (15 of October, 2016)

During the first expedition (Figure 4.9) the most distinguishable claw marks were recognized, as well as the bones of the “bear” number four and the main finding (the “bear” number one) was started to be dug. Soon it became evident that it could not be taken out with the small tools brought in this first expedition that could not pierce through the thick layer of calcareous crust in the side of the block. Also it was impossible to take the skull alone without a serious risk of destroying it, so the block had to be taken entirely.

Figure 4.9: First excavation around the skull of “bear number one” in October 2016.

Second expedition (14 of October, 2017)

During the second expedition (Figure 4.10) more claw marks were located in several panels of the cave, besides the remains of the “bear number five” and the dog that fell into the cave. In this expedition it also became clear that in order to take out the block a big expedition should be assembled and many people specialized in cave speleology would be required. Also Manuel Pinto Soares was interviewed about his knowledge of the cave. Rui Guerra, a documentalist and camera started the elaboration of a documentary about the digging in the cave filming the bear site and the surroundings. Up until this moment no bone remains where taken out of the cave.

Figure 4.10: The main block as in October 2017, during the second expedition, no new diggings where carried out this time. The skull is near the scale of seven centimeters,

Third expedition (16 and 17 of June, 2018)

This time a big expedition was set with 13 participants in the first day. The jeep of the FCT-UNL as well as its powerline and circular saw specially bought for the task were used, also the drill and stone splitters of Octávio Mateus and the power generator and illumination of the GPS. First the clay and mud was dug around the block of the “bear number one” with relative ease (more about it in the next section). Then the bones were protected with a first layer of plasters and the block with the main bones was successfully cut thanks to the drill and the stone splitters where the source of light was in Figure 4.10. Then the block was put in a net. In order to move it some members of the team went out of the cave, cut an eucalyptus tree of about 10 meters tall, removed the branches and descended it safely into de cave to be used. Even thanks to the tree, the block weighed several hundred kilograms and could only be taken up to the first point where the path goes to the right, since is not possible for so many people to maneuver in the confined space of the “railing”. Finally it was conveniently plastered with several extra layers. In the figure 4.11 we have a series of pictures about that process.

At the same time that the block was receiving its final plastering Octávio Mateus moved alongside the walls of the cave counting all the claw marks that he could find and discovering the remains of the “bears” number two and three. As much bones as possible were collected, including three small fragments from “bear number four”. The remains of “bears” two and three that could not be taken away were covered with plaster. In the next sections we will discuss more precisely which remains were taken, which ones stay in the cave and what was the position and preservation of the elements, as well as the some notes about the taphonomy and a comparison of the remains with other bears.

The remains (a total of 37 items) were stored during summer in the Museu da Lourinhã and were prepared in the same institution in October of 2018. Use of glues (Paraloid B72, at 5% of concentration in acetone for consolidation and 40% for gluing) was reduced to the minimum possible and the preparation was carried out only in a mechanical manner with several air-scribes in order to improve the possibilities for hypothetical isotopic dating, ancient DNA and other isotopic diet studies. In the figure 4.12 we can see the “before and after” of some bones recovered.

Figure 4.12: Before ( left) and after( right) for some ursid remains recovered in Algar do Vale da Pena: a) distal radius, b) mandibular ramus and c) basicranium.

4.4 The bone assemblages

Bear number one

It appears as 1 in figure 4.4 and is the main finding in the cave.

Items still in the cave: Main block with a nearly complete skull, scapula and humerus (Figure 4.13). The process to remove this block and its current status appear in figure 4.11, also numerous bone fragment and smaller bones still in the mud.

Figure 4.13: The main block of “bear” one during the second expedition. Scapula and protruding long bones are marked with black arrows. In red, line where the block was cut, main skull with the scale of 7 cm and muddy area with numerous smaller bone remains.

Comments: This finding seems to have been subject to little if any transportation of the bones. The skull posses the articulated mandible and this is highly unlikely to be conserved if it was subject to the movement. The bones seem to have been covered first with thin layer of calcareous crust, and then the skull and the bones in the main block were covered (At least partially) with several thick layers of calcareous crust, meanwhile that the bones in the muddy area were covered by sands and mostly clays.

Bear number two

It appears as 2 in figure 4.4, was discovered in the third expedition and is the second best finding in the cave for now.

Items still in the cave: Plaster jacket in situ including at least some partial remains.

Figure 4.14: Remains of the “bear number two” before (left) and after (right) removal of the block that was hiding them.

Comments: This assemblage of bones seems to have been moved by subterranean waters from their original position and eroded in the process. In the figure 4.15 we can see how there is muddy sediment below the calcareous crust in the hemimandible during the preparation, meanwhile in the previous assemblage it was the opposite. Also the anterior part of the hemimandible and other remains show signs of breakage and erosion before being covered by the calcareous crust.

Figure 4.15: Muddy sediment below the calcareous crust, during the preparation of the hemimandible (ADVP 2.1). Scale 7 cm. Bear number three

This bone assemblage is marked with the number 3 in the figure 4.4. It is only situated six meters below the first finding, and it could not be ruled out yet that it belongs to the same individual as the first one.

Items still in the cave: Partial tibia with on isolated block plaster jacket and other remains (proximal femur and possible fibula) in plaster jacket in situ.

Comments: This assemblage of bones is the worst preserved of all. There is not a single intact bone, being all of them heavily eroded and quite fragile. This fact contributes to the hypothesis that these bones were transported from the first site by water.

Bear number four

This assemblage is situated at the end of the cave, the number 4 in the map of figure 4.4.

Items still in the cave: Humerus, femur, and part of a pelvis totally encrusted in calcareous matrix plus several smaller bones equally encased in the crust (figure 4.16).

Figure 4.16: Material of bear number four encased in the calcareous crust. Left, humerus; right, femur and pelvis.

Comments: Is quite difficult to know if these bones are as well preserved as it seems, because they are completely embedded into the calcareous matrix. The recovery of more material of this assemblage seems nearly impossible. Even cut in blocks with a circular saw and successfully transported outside the cave, the preparation has so many setbacks that the best option is probably to let them in the place.

Bear number five

This assemblage is near the bear number two; it appears with the number 5 in the map of figure 4.4

Items still in the cave: Distal humerus (figure 4.17), and possible tibia encased in calcareous crust.

Figure 4.17: Distal humerus of the bear number five.

Comments: Despite being also encased in the crust this bones could be recovered easier than the ones of the previous assemblage, since the crust is thinner here.

4.5 Results: The recovered bone remains

In this section we will list the 37 recovered items; each of them will have the field number that consists in the cave (ADVP stands for “Algar do Vale da Pena”) the number of the bear and the number of item in particular. They are stored at FCT-UNL. Then measurements and/or pictures of the most relevant items (marked with an asterisk) will be presented. These measurements will be given according to the system for ursid measurements provided by Tsoukala & Grandal-d’Anglade (2002). Then they will be compared with other brown and cave bears as well as with Asiatic black bear in some specimens.

Items recovered for “bear” one: - ADVP 1.1: Partial left scapula *. - ADVP 1.2: Scapula fragments. - ADVP 1.3: Right scapholunate bone*. - ADVP 1.4: Right hamate bone*. - ADVP 1.5: Highly damaged metatarsal. - ADVP 1.6: Second phalanx*. - ADVP 1.7: Proximal epiphysis of third right metatarsal *. - ADVP 1.8: Distal epiphysis of right radius*. - ADVP 1.9: Distal part of diaphysis of left ulna*. - ADVP 1.10: Fragments of rib. - ADVP 1.11: Undifferentiated bone fragments. - ADVP 1.12: Calcareous matrix pieces in contact with bone. - ADVP 1.13: Bone fragments recovered in the first excavation. - ADVP 1.14: Big calcareous crust.

Items recovered for the “bear” two: - ADVP 2.1: Posterior part of right hemimandible *. - ADVP 2.2: Parietal and occipital *. - ADVP 2.3: Temporal bone*. - ADVP 2.4: Cranial fragments. - ADVP 2.5: Heavily damaged canine *. - ADVP 2.6: Left upper fourth premolar *. - ADVP 2.7: Left upper first molar *. - ADVP 2.8: Right lower second molar *. - ADVP 2.9: Left lower third molar *. - ADVP 2.10: Partial third phalanx*. - ADVP 2.11: Bone fragments. - ADVP 2.12: Calcareous crust in contact with bones.

Items recovered for the “bear” three: - ADVP 3.1: Partial distal epiphysis of right femur*. - ADVP 3.2: Vertebral centrum *. - ADVP 3.3: Highly damaged vertebral centrum. - ADVP 3.4: Fragment of rib. - ADVP 3.5: Bone laminae (Part of the proximal femur or the tibia). - ADVP 3.6: Possible fragment of sacrum. - ADVP 3.7: Bone fragments. - ADVP 3.8: Calcareous matrix.

Items recovered for the “bear” four: - ADVP 4.1: Proximal epiphysis of second right metatarsal*. - ADVP 4.2: Damaged proximal epiphysis of fifth right metatarsal*. - ADVP 4.3: Bone fragments.

ADVP 1.1: Partial left scapula

This bone (Figure 4.18) can only be recognized as a scapula thanks to the calcareous crust that is holding it together. During preparation it was clear that removing the calcareous crust would damage irreversibly the already compromised bone.

Figure 4.18: Damaged scapula (ADVP 1.1) from “bear number one” (Ursus arctos) of Algar do Vale da Pena.

ADVP 1.3: Right scapholunate bone

This carpal bone (Figure 4.19) is nearly in pristine condition, and one of the best preserved in “bear one”. Its measurements are presented in the table 4.1. The orientation and angle of the apophysis rule out an Ursus thibetanus Cuvier, 1823 origin (Torres Pérez-Hidalgo, 1984).

Figure 4.19: Right scapholunate bone (ADVP 1.3) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) proximal; B) distal; C) anterior and D) exterior views.

Figure 4.20: Measurements for the scapholunate bone in ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). All of them could be recorded for the specimen.

Table 4.1: Measurements from the right scapholunate bone (ADVP 1.3) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to a set of measurements of other brown bears (García Vázquez, 2015) and cave bears (unpublished own data from Cova da Ceza population). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Species Specimen L DT DAP DTart.prox. DAPart.prox. DTart.dist. DAPart.dist. U. arctos? ADVP 1.3 24 43 44 39 24 22 37 U. arctos LCF-137 26,79 49,43 47,12 42,89 26,82 27,74 41,23 U. arctos SIPA-52 26,27 43,59 43,02 39,54 25,75 23,12 38,45 U. arctos SIPA-118 31,3 51,12 49,55 47,33 30,13 28,4 42,7 U. arctos PUR-80 27,69 49,08 47,78 46,86 28,81 29,4 44,2 U. arctos SH5-98-S28-089 33,14 50,86 54,39 49,46 30,75 30,13 48,41 U. arctos SH5-98-S28-100 33,23 51,51 53,1 50,66 31,8 30,93 47,45 U. arctos SH5-97-T29-25 25,83 42,62 41,15 37,37 24,87 23,67 33,31 U. arctos SH5-97-U29-079 24,34 42,35 40,64 37,05 24,87 23,95 33,17 U. arctos TA-Lu-c-34 26,65 44,33 42,01 39,28 23,71 23,04 36,33 U. arctos 42,42 45,89 41,56 27,61

U. arctos 42,81 44,57 43 33,02

U. arctos El Cuervo 25,31 39,95 37,11 39,24 24,97 25,23

U. spelaeus CEZ 42 36,49 58,8 57,67 57,27 36,34 32,48 56,13 U. spelaeus CEZ 171 25,45 48,24 51,17 47,17 29,05 23,68 42,37 U. spelaeus CEZ 729 28,9 49,84 50,19 48,75 30,39 27,22 46,22 U. spelaeus CEZ 730 28,42 49,8 49,88 47,24 30,3 28,64 45,97 U. spelaeus CEZ 731 40,36 64,61 68,71 65,44 42,99 38,34 61,69

According to García Vázquez, (2015) the transverse vs anteroposterior diameters of the scapholunar are the best measurements to distinguish between males and females of brown bear; so in figure 4.21 we have a XY plot of those measurements.

Figure 4.21: Transverse diameter vs anteroposterior diameter of the scapholunate of brown bears (García Vázquez, 2015) as brown dots and cave bears (unpublished own data from Cova da Ceza population) as green squares; plus the scapholunate ADVP 1.3, marked as a yellow star.

Looking at the graph we can say that the dimensions of the scapholunar are similar to the expected on brown bears so in figure 4.22 we exclude the cave bears.

Putative females

Putative males

Figure 4.22: Transverse diameter vs anteroposterior diameter of the scapholunate of brown bears (data from García Vázquez, (2015)) as brown dots; plus the scapholunate ADVP 1.3, marked as a yellow star.

So ADVP 1.3 is classified as the scapholunate of a female brown bear in base of the ratio between transversal and anteroposterior diameters plus their overall dimensions.

ADVP 1.4: Right hamate bone

This carpal bone (Figure 4.23) is only slightly damaged and is still one of the best preserved of this bear, but taking into account the morphology of this bone and the difficulties to measure it properly (many of the views looks similar) only one measurement has been recorded. This measurement is presented in the table 4.2 and we can graphically see them at figure 4.24.

Figure 4.23: Right hamate bone (ADVP 1.4) from “bear number one” (Ursus arctos) of Algar do Vale da Pena.

Figure 4.24: Measurements for the scapholunate ursid bone according to the system of Tsoukala & Grandal- d’Anglade (2002).The recorded measurement is marked.

Table 4.2: Length of the right hamate bone (ADVP 1.4) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the length of the same bone of other brown bears (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Individual Light ADVP 1.4 27 LCF-141 33,07 LCF-142 27,1 SIPA-109 33,1 SIPA-111 32,17 Pur-Lu-33 35,37 SH5-98-S28-064 35,54 SH5-97-T29-084 25,75 TA-Lu-c-33 30,74

36 Putative males 34

26 Putative females

Figure 4.25: Length of the right hamate bone (ADVP 1.4) from “bear number one” of Algar do Vale da Pena compared to the length of the same bone of other brown bears (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002). Two groups (males and females) are classified according to this measurement.

As we can see in figure 4.25, ADVP 1.4 has the second smallest length of the hamate in the sample of brown bears; therefore it is classified as belonging to a female or juvenile brown bear.

ADVP 1.6: Second phalanx

This phalanx (Figure 4.26) has been classified as a second phalanx due to its morphology compared to the work of Torres Pérez-Hidalgo, (1984). Some parts of it are slightly damaged, but overall the condition is good. In the table 4.3 we have the longitude and distal transversal diameter the of the second phalanx compared with the same measurements of the second phalanxes of a recent brown bear, from which we know if the bones are from hindlimb or forelimb (García Vázquez, 2015). In figure 4.27 we can see the measurements graphically. In figure 4.28 we have an XY plot elaborated with the data of table 4.3.

Figure 4.26: Second phalanx (ADVP 1.6) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) ventral; C) proximal; D) distal views.

Figure 4.27: Measurements for the second phalanx of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). All of them could be recorded.

Table 4.3: Length vs distal transverse diameter of the second phalanx (ADVP 1.6) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the length of the same bone of a recent brown bear (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Anterior/posterior Number L DTdist. ? ADVP 1.6 25 11 Anterior LE1 28,97 14,53 Anterior LE1 29 14,87 Anterior LE1 29,79 15,38 Anterior LE1 28,8 14,38 Anterior LE1 27,21 14,41 Anterior LE1 28,33 14,95 Posterior LE1 21,55 12,11 Posterior LE1 23,15 11,78 Posterior LE1 22,69 12,41 Posterior LE1 22,98 12,7

Figure 4.28: Length vs distal transverse diameter of the second phalanx (ADVP 1.6, marked as blue square) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the same ration of a recent brown bear, data from García Vázquez, (2015) . The 95% ellipses are drawn in the graphs. Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Therefore, according to the figure 4.28 the second phalanx is more likely to come from the pes of the bear.

ADVP 1.7: Proximal epiphysis of third right metatarsal

This bone has been classified as the proximal epiphysis of the third right metatarsal in base of its morphology (Figure 4.29). Only its proximal transverse and anteroposterior diameters could be recovered. They are presented in the table 4.4 alongside the same measurements of brown bears from NW Spain (García Vázquez, 2015). In Figure 4.30 we see the measurements. In Figure 4.31 we can see graphically the comparison of table 4.4.

Figure 4.29: Proximal epiphysis of the third right metatarsal (ADVP 1.7) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) exterior; B) dorsal; C) interior and D) ventral views.

Figure 4.30: Measurements for the third metatarsal of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002).The recorded measurements are marked.

Table 4.4: Proximal transverse diameter and proximal anteroposterior diameter of the third metatarsal (ADVP 1.7) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to same measurements of fossil brown bears from NW Spain (García Vázquez, 2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Dtprox. DAPprox. ADVP 1.7 11 21 LCF-129 13,05 21,44 LCF-130 17,13 24,68 CGLL-028 17,37 27,87 SIPA-6 18,66 26,45 SH5-97-T29-42 16,13 23,05 SH5-98-V30-016 16,04 26,57 SH5-98-R29-001 17,06 25,03 SH5-97-T29-59 14,8 23,73 SH5-98-S28-075 20,26 30,67 SH5-98-T31-002 16,71 25,57 TA-Lu-c-22 15,61 23,37

Figure 4.31: Proximal transverse diameter vs Proximal anteroposterior diameter of the third metatarsal (ADVP 1.6, marked as blue square) from “bear number one” (Ursus arctos) of Algar do Vale da Pena (As green triangle) compared to the same ration of fossil brown bears from NW Spain, data from García Vázquez, (2015) . Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

As we can see in figure 4.30, the metatarsal from Algar do Vale da Pena is the smallest from the entire sample, almost as it belonged to a juvenile, taking into account how slender it is (DT prox), although it has to be taken into account that it could be slightly wear down.

ADVP 1.8: Distal epiphysis of right radius

This distal epiphysis of the right radius, ADVP 1.8 (Figure 4.32) is the third most complete postcranial bone recovered until this point in the whole site. It shows a totally fused epiphysis; therefore it is interpreted as belonging to an adult or nearly adult bear. In table 4.5 we can see its measurements compared with brown bears from NW Spain (García Vázquez, 2015), as well as some measurements from Portuguese caves (only maximum and minimum transverse diameters of the bone) as presented in Cardoso, (1993). In figure 4.33 we have graphically the measurements of the radius according to Tsoukala & Grandal-d’Anglade (2002).

Figure 4.32: Distal epiphysis of the right radius (ADVP 1.8) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) anterior; B) exterior; C) Posterior; D) Interior and E) Distal views.

Figure 4.33: Measurements for the radius of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002). Recorded measurements are marked.

Table 4.5: Measurements from the radio (ADVP 1.8) from “bear number one” of Algar do Vale da Pena compared to same ratio of fossil brown bears from NW Spain (García Vázquez, 2015) as well as some of Portuguese caves. Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

DTdist. DAPdist. DTart.dist. DAPart.dist.

ADVP 1.8 52 32 33 27 CB-004 54,45 34,98

LCF-008-1996 66,13 41,14

LCF-009-1996 66,84 40,65

LCF-96-037 65,68 39,59

LCF-96-038 53,98 31,58

CGLL-043 55,91 35,42

SIPA-3 55,25 32,19

Pur-Lu-19 62,82 38,25

Pur-Lu-20 58,73 40,43

1996-SH5-001 52,76 32,56

SH5-97-V28-001 52,49 32,05

SH5-98-S28-035 65,81 41,75

SH5-98-S29-011 55,29 32,95

SH5-98-U30-012 57,61 33,88

TA-Lu-c-6 55,41 32,77

56,53 36,19

El Cuervo 50,34 32,48

Furninha max 62

Furninha min 56,2

Fontainhas max 63,5

Fontainhas min 58,7

Figure 4.34: Transverse diameter of distal radius from ADVP 1.8 and fossil and recent brown bears from NW Spain and Portugal (Cardoso, 1993; García Vázquez, 2015). Specimens with more than 60 mm are classified as male and less than 58 mm as females.

Figure 4.35: Transverse diameter vs anteroposterior diameter of distal radius from ADVP 1.8 (Green in the graph) and fossil and recent brown bears from NW Spain (García Vázquez, 2015). Two clusters of individuals are classified as females or males.

As we can see in figures 4.34 and 4.35 ADVP 1.8 can be classified according to its measurements as the distal part of the radius of a small female brown bear.

ADVP 1.9: Distal part of diaphysis of left ulna

This distal part of the ulnar diaphysis (Figure 4.36) has its epiphysis highly damaged; therefore only one measurement could be taken. In figure 4.37 we have graphically the measurements of the ulna according to Tsoukala & Grandal-d’Anglade (2002). In table 4.6 we have the measurement of ADVP 1.9 together with equivalent measurements of brown bears from NW Spain (García Vázquez, 2015).

Figure 4.36: Distal distal part of the ulnar diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena in A) anterior; B) exterior; C) Posterior and D) Interior views.

Figure 4.37: Measurements for the ulna of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002). Taken measurement is marked with a circle.

Table 4.6: Minimum anteroposterior diameter of the ulnar distal diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen DAPdia.min. ADVP 1.9 20 CB-001 22,37 CB-002 22,81 LCF-010- 1996 28,04 LCF-163 27,42 LCF-164 22,13 EIX-009 22,82 CGLL-053 26,44 SIPA-4 21,82 SIPA-63 21,57 Pur-Lu-18 27,72 SH5-97-T29- 11 20,81 SH5-97-T29- 27 20,67 TA-Lu-c-4 22,55 El Cuervo 23,32

Figure 4.38: Minimum anteroposterior diameter of the ulnar distal diaphysis (ADVP 1.9) from “bear number one” (Ursus arctos) of Algar do Vale da Pena compared graphically to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Suggested males and females are marked.

Acording to the data in table 4.6 and the figure 4.38 ADVP 1.9 is classified as the ulna of a small female brown bear.

ADVP 2.1: Posterior part of right hemimandible

This is the best preserved cranial remain recovered in the cave so far. As we can see in figure 4.39, it has been eroded in the anterior part and lost all its teeth; the mandibular condyle is also lightly damaged. In figure 4.40 we can see the measurements taken to it and in table 4.7 their values compared to other bears of NW Spain.

Figure 4.39: Posterior part of right hemimandible (ADVP 2.1) from “bear number two” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) posterior; C) exterior; D) anterior; E) interior and F) ventral views.

Figure 4.40: Measurements for the mandible of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurements are marked with a circle.

Table 4.7: Measurements from the partial hemimandible (ADVP 2.1) from “bear number two” (Ursus arctos) of Algar do Vale da Pena compared to same measurements of fossil brown bears from NW Spain (García Vázquez, 2015) as well as some cave bears from the same region (Unpublished own data from Cova de A Ceza). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Species Specimen L cond.-Mp H pro.ang.-Cor Hmd(M3)(la) Hcond DTmd(M2,M3) Hmd(max) ? ADVP 2.1 70 86 53 15 19 103 U. arctos LCF-003-1996 92,46 102,41 57,34 21,59 23,31 134,25 U. arctos LCF-002-1996 93,51 102 56,92 21,16 23,86 134,25 U. arctos LCF-155 44,98 19,47

U. arctos LCF-167 98,03 101,72 60,91 19,7 22,23 132,85 U. arctos CCV-003 96,02 102,8 57,9 25,05 25,03 141,25 U. arctos CCV-004 95,89 59,53 23,38 22,8 132,45

U. arctos GP-1 81,08 44,21 15,07 16,07

U. arctos CGLL-012 44,07 18,39

U. arctos CGLL-011 46,13 18,16

U. arctos SIPA-15 48,77 18,69

U. arctos SIPA-53 80,02 52,33 15,12

U. arctos SIPA-128 54,94 22,05

U. arctos PVR-013 89,79 95,87 54,77 22,65 20 129,85 U. arctos PVR-014 88,56 95,4 49,82 23,35 19,57 125,35 U. arctos SH5-97-U29-7 79,01 83,01 42,39 16,37 14,35 106,15 U. arctos SH5-97-U29-6 81,07 82,49 42,25 16,48 14,25 105,55 U. arctos SH5-98-V29-7 97,65 78,54 52,59 19,87 23,26 117,7 U. arctos SH5-98-S30-6 54,11 22,49

U. arctos SH5-97-N19-4 82,48 83,81 15,29 17,4 104,55

U. arctos SH5-98-U28-010 74,88 45,68 14,33 18,45 105,4

U. arctos TA-109 100,37 81,97 45,66 18,22 19,95 112 U. arctos TA-108 45,96 18,76

U. arctos Cráneo 001 57,23 68,66 36,25 16,16 85,91

U. arctos Cráneo 001 58,17 65,88 35,91 16,23 16,07 84,26 U. arctos Cráneo 002 67,08 60,55 14,83 14,4 84,45

U. arctos Cráneo 002 64,52 61,43 38,35 15,39 14,99 88,2 U. arctos ? 90,54 86,84 55,76 15,64 19,93 142,89 U. arctos ? 83,46 87,59 55,28 15,54 18,85 142,1 U. spelaeus 141 102,03 66,52 26,7 25,88

U. spelaeus 142 52,98 65,33 17,4 23,82 82,2

U. spelaeus 144 55,79 57,29 16,27 22,4 66,47

U. spelaeus 145 53,18 54,24 15,9 68,03

U. spelaeus 146 57,36 16,35 22,74 76,97

U. spelaeus 191 23,57

U. spelaeus 518 59,68 63,31 18,9 25,9 75,71

U. spelaeus 664 30,09

U. spelaeus 801 50,3 16,63 19,8 75,22

With this complete set of data a PCA was performed in the program Past (version 3.14), although the results were not informative, they pointed towards good measurements to use in biplot graphs, which can be seen in figures 4.41, 4.42 and 4.43.

Figure 4.41: Total height of the mandible vs length from the mandibular condyle to the third molar from ADVP 2.1 (blue square in the graph), fossil and recent brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as green triangles (unpublished data from Cova da Ceza). Two clusters of individuals are classified as females or males.

Figure 4.42: Total height of the mandible vs height of the mandibular condyle from ADVP 2.1 (blue square in the graph) and fossil and recent brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as red inverted triangles (unpublished data from Cova da Ceza).

In the figure 4.41 we can see that the Portuguese specimen falls in the variability of the female brown bears. In the figure 4.42 we can see that the cave bears tend to be more robust (taller) mandibular condyles even when the total height of their mandible is smaller, ADVP 2.1 falls into the smaller brown bears. In the figure 4.43 the measurements are centered in the mandibular corpus, the only cave bear included in the analysis poses a sturdier mandibular corpus than any of the 24 confirmed brown bears used.

This data suggest that the mandible belongs to a fairly average sized female brown bear.

ADVD: 2.2 Parietal and occipital

This item (Figure 4.44) comprises partial occipital and parietal bones. The foramen magnum and cranial condyles are totally gone and the remaining part of the cranial cavity is obscured by a layer of calcite that cannot be removed without compromising the stability of the whole bone. Interestingly, despite that the interparietal suture is pretty much fused; the nucal crest is almost nonexistent, as well as the external occipital protuberance and parietal crest, even taking into account that they might be slightly damaged. Not a single measurement according to the system of Tsoukala & Grandal- d’Anglade, (2002) could be recorded in the specimen.

Figure 4.44: Partial occipital and parietal bones (ADVP 2.2) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) dorsal; B) posterior and C) anterior views.

ADVP 2.3: Temporal bone

This specimen (Figure 4.45) comprises an almost complete temporal bone and part of a parietal one. The zygomatic apophysis (the start of the zygomatic arch in the temporal bone) is broken and partially covered with calcareous crust. The post-glenoid apophysis (the place where the mandibular condyle articulates with the cranium) is broken as well; being the broken part separated and covered by calcareous crust which prevents it to be glued back in place. The ootic channel and the petrosal part of the temporal are lightly damaged. Both the temporal crest (between the base of the zygomatic apophysis and the posterior part of the petrosal part of the temporal) and the lambda crest (that runs along the posterior margin of the temporal) are well preserved. The squamosal part of the temporal is intact and articulates with a partial parietal bone. In the part of the cranial cavity that remains we can see the foramina of several cranial nerves, and the inner ear canal. Again, not a single measurement according to the system of Tsoukala & Grandal-d’Anglade, (2002) could be recorded in the specimen.

Figure 4.45: Temporal and partial parietal bones (ADVP 2.3) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) exterior and B) interior views.

ADVP 2.5: Heavily damaged canine

This canine tooth is so heavily damaged that cannot be measured, since not only is eroded, but also partially covered in calcareous matrix, that would compromise heavily the bone if removed (Figure 4.46).

Figure 4.46: Partial canine teeth (ADVP 2.5) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in lateral view.

ADVP 2.6: Right upper fourth premolar

This last premolar of the dental series has been found in a nearly pristine condition (Figure 4.47). It has a marked metastyle near the metacone as well as a shallow cingulum near the paracone. In oclusal view it has a strong triangular shape, compared to the Furninha teeth depicted by Torres Pérez- Hidalgo, (1984), that are broader in the mesial part of the tooth. We can see its measurements graphically in figure 4.48 and numerically in table 4.8, as well as the breadth and length of several of these teeth of brown bears of NW Spain. If we look at the figure 4.49 that it is the longest tooth, but one of the narrowest. It is difficult to assign it to a possible age or sex in Ursus arctos.

Figure 4.47: Fourth upper premolar (ADVP 2.6) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views.

100

Table 4.8: Measurements from the fourth upper premolar (ADVP 2.6) from “bear number two” of Algar do Vale da Pena compared to broad an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Species L B Hpa Hme Hde Lpa-me Lpa-de Lme-de ADVP 2.6 ? 18 11 8 5 4 4 5 4 ARLU-38 U. arctos 16,9 13,62 10,37 8,81 5,29 7,02 7,61 7,26 ARLU-38 U. arctos 17,47 13,65 10,43 8,74 7,08 7,54 8,97 8,18 CB-010 U. arctos 17,61 14,55 12,38 8,85 8,61 5,36 7,62 6,17 CB-010 U. arctos 17,49 14,6 12,41 8,95 8,4 6,03 6,77 6,66 LCF-001-1996 U. arctos 15,5 13,31 8,27 7,24 7,72 7,48 8,07 6,75 LCF-001-1996 U. arctos 16,31 13,03 9,5 7,4 6,82 6,94 8,52 5,38 LCF-169 U. arctos 14,92 11,96 7,22 6,36 7,35 5,67 6,49 5,68 CGCH-015 U. arctos 16,7 14,37 11,83 7,37 6,6 5,54 7,85 5,73 CGCH-016 U. arctos 17,18 13,92 11,34 5,98 7 6,03 8,18 6,15 SIPA-16 U. arctos 14,8 12,22 9,29 6,15 4,93 6,14 6,88 5,81 SIPA-56 U. arctos 16,2 12,3 6,82 5,55 5,39 6,53 7,17 5,67 SIPA-57 U. arctos 16,27 12,92 7,94 4,98 5,41 6,59 7,07 6,29 SIPA-135 U. arctos 16,83 12,82 9,01 6,38 6,43 6,25 7,68 6,16 SIPA-140 U. arctos 17,57 13,39 9,49 7,52 6,04 6,1 7,46 6,04 RT-001 U. arctos 17,43 10,95 8,53 7,02 4 6,41 8,65 4,52 Pur-Lu-47 U. arctos 16,32 12,8 9,18 6,91 6,32 6,95 8,05 6,97 Pur-Lu-48 U. arctos 16,67 12,91 9,09 6,72 6,36 7,21 7,5 5,95 SAB-82 U. arctos 17,17 12,33 9,7 7,6 5,81 6,88 9,76 6,54 SH5-98-T28-013 U. arctos 15,02 9,45 10,05 6,45 6,6 4,86 7,16 3,83 SH5-97-N20-2 U. arctos 15,05 12,17 9,51 6,19 6,99 4,8 6,09 5,64 SH5-97-N20-2 U. arctos 15,33 11,81 8,87 7,59 6,9 4,41 7,82 6,51 SH5-98-U30-001 U. arctos 14,55 12,37 10 8,3 8,48 6,62 8,15 6,7 SH5-98-U30-001 U. arctos 14,34 13,12 10,57 8,86 8,22 6,23 7,38 7,42 ? U. arctos 14,98 13,65 8 5,07 4,94 6,87 9,64 6,84 ? U. arctos 14,1 12,59 8,14 5,35 6,19 6,58 9,22 5,89 - U. arctos 14,2 11,7 - U. arctos 14,9 11,8 MZB 82-7005 U. arctos 14,23 13,21 8,25 7,1 7,15 5,56 6,6 5,52 MZB 82-7005 U. arctos 13,15 12,64 8,54 6,47 6,94 5,74 6,61 5,03

101

Figure 4.48: Measurements for the fourth upper premolar of ursids according to the system of Tsoukala & Grandal-d’Anglade (2002).

Figure 4.49: Breadth vs length of the fourth upper premolar ADVP 2.6 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015).

ADVP 2.7: Partial left upper first premolar.

As we can see in the figure 4.50, this molar is partially eroded in its lingual side, therefore only the parastyle, paracone, metacone and metastyle are visible. The metastyle of Ursus thibetanus should protrude more from the tooth and a continuous more marked cutting ridge should run over paracone and metacone in a tooth of this species (Torres Pérez-Hidalgo, 1984). The space between the metacone and the metaconule (That can be partially seen) does not have any ornamentation, as in cave bears. We can see its measurements in figure 4.51 and table 4.9, as well as the breadth and length of several of these teeth of brown bears of NW Spain (García Vázquez, 2015) and cave bears from Liñares cave, Galicia, NW Spain (González & Martelli, 2003). If we look at the figure 4.52 we can see that even the smallest cave bears have more robust teeth than the bigger brown bears. ADVP 2.7 falls into the lower part of the variation of U. arctos.

102

Figure 4.50: First upper molar (ADVP 2.7) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views.

Figure 4.51: Measurements for the first upper molar according to the system of Tsoukala & Grandal-d’Anglade (2002). Taken measurements are marked with a circle.

103

Table 4.9: Length and breadth from the first upper molar (ADVP 2.7) from “bear number two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from Liñares cave, Galicia, NW Spain (González & Martelli, 2003). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Species L B ADVP 2.7 ? 21 15 273 U. spelaeus 28,41 19,85 272 U. spelaeus 26,88 19,46 273 U. spelaeus 27,96 20,31 272 U. spelaeus 27,07 20 271 U. spelaeus 26,86 19,87 1 U. spelaeus 28,16 20,06 1100 U. spelaeus 26,6 19,1 841 U. spelaeus 28,7 20,8 271 U. spelaeus 26,84 19,2 1 U. spelaeus 28,2 19,84 1100 U. spelaeus 27,2 18,3 841 U. spelaeus 28,6 20,8 ARLU-38 U. arctos 21,99 17,07 ARLU-38 U. arctos 21,67 17,23 CB-010 U. arctos 25,62 19,46 CB-010 U. arctos 25,14 19,81

LCF-001-1996 U. arctos 25,05 18,38 LCF-001-1996 U. arctos 25,07 18,18 LCF-169 U. arctos 22,54 16,48

CGCH-015 U. arctos 23,83 18,2 CGCH-016 U. arctos 23,56 18,52 Pur-Lu-49 U. arctos 22,61 17,23

RT-001 U. arctos 23,33 18,29 RT-001 U. arctos 23,09 18,12 SIPA-16 U. arctos 21,27 15,96

SIPA-56 U. arctos 23,26 16,43 SIPA-33 U. arctos 25,21 17,87 SIPA-136 U. arctos 23,85 17,71

SIPA-141 U. arctos 24,18 17,9 SH5-98-T28-013 U. arctos 20,43 15,01 SH5-97-N20-2 U. arctos 21,39 16,17

SH5-97-N20-2 U. arctos 21,68 15,94 SH5-98-U30-001 U. arctos 22,68 16,21 SH5-98-U30-001 U. arctos 22,57 16,12

SH5-97-AO34-39 U. arctos 21,19 15,08 SH5-97-AO34-36 U. arctos 20,79 14,68 TA-196 U. arctos 22,02 17,92

? U. arctos 23,11 17,25 ? U. arctos 22,98 17,25 - U. arctos 21,85 15,9

- U. arctos 21,5 16,2 MZB 82-7005 U. arctos 21,74 15,14

104

Figure 4.52: Breadth vs length of the first upper premolar ADVP 2.7 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015) and cave bears as orange triangles from Liñares cave, Galicia, NW Spain (González & Martelli, 2003).

ADVP 2.8: Right lower second molar

This molar (Figure 4.53) was slightly damaged in its posterior part around the entoconid, but overall it was well preserved. In the table 4.10, we can see its measurements of breadth and length together with those of several teeth of cave bears A Ceza cave, Galicia, NW Spain (Unpublished own data) and brown bears of NW Spain (García Vázquez, 2015) . In figure 5.54 we have the measurements taken and in figure 5.55 a biplot that shows that ADVP 2.8 falls into the variability of brown bears (between the sizes of modern and fossil brown bears) and that it is well outside the range of U. spelaeus.

Figure 4.53: Second lower molar (ADVP 2.8) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views.

105

Table 4.10: Breadth and length of the second lower molar (ADVP 2.8) from the “bear two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza cave, Galicia, NW Spain (Unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Species L B ADVP 2.8 ? 22 14 del 518 U. spelaeus 32,79 18,48 del 144 U. spelaeus 31,7 18,44 del 142 U. spelaeus 31,38 17,71 del 518 U. spelaeus 32,92 19,1 del 517 U. spelaeus 32,32 18,68 del 801 U. spelaeus 30,88 18,03 LCF-003-1996 U. arctos 27,26 15,72 LCF-002-1996 U. arctos 27,89 15,79 LCF-154 U. arctos 23,44 14,71 LCF-155 U. arctos 23,43 15,22 EIX-018 U. arctos 24,59 14,9 CGLL-011 U. arctos 24,74 14,59 CGLL-004 U. arctos 25,19 15,32 SIPA-15 U. arctos 22,67 14,11 SIPA-129 U. arctos 26,03 16,42 SIPA-133 U. arctos 25,93 16,64 Pur-Lu-52 U. arctos 25,53 15,91 Pur-Lu-53 U. arctos 25,09 15,17 SH5-97-U29-6 U. arctos 23,11 14,51 SH5-98-V29-7 U. arctos 26,15 15,62 SH5-98-S30-6 U. arctos 26,77 15,95 SH5-97-N19-4 U. arctos 23,56 13,62 SH5-98-U30- 014 U. arctos 23,23 13,85 SH5-98-U30- 025 U. arctos 23,36 13,98 TA-109 U. arctos 25,28 14,09 TA-192 U. arctos 24,31 16,15 TA-195 U. arctos 24,62 16,03 VA87/16F/37 U. arctos 25,37 15,42 ? U. arctos 22,7 15,2 ? U. arctos 22,71 16,01 QU U. arctos 20,8 13,1 QU U. arctos 21,5 12,75 MZB 82-7005 U. arctos 21,05 12,28 MZB 82-7005 U. arctos 21,2 12,91

106

Figure 4.54: Measurements for the second lower molar according to the system of Tsoukala & Grandal- d’Anglade (2002).

Figure 4.55: Breadth vs length in mm of the second lower molar ADVP 2.8 (blue square in the graph) and the same measurements of recent and fossil brown bears from NW Spain as green diamonds(García Vázquez, 2015) plus cave bears from A Ceza cave as black dots, Galicia, NW Spain (Unpublished own data).

107

ADVP 2.9: Left lower third premolar

This tooth (Figure 4.56) is the best preserved dental remain in the entire sample and has a fairly similar morphology to the teeth depicted for U. arctos from Furninha by Torres Pérez-Hidalgo, (1984). In figure 4.57 we have graphically its measurements and in table 4.11 the length and breadth of several of those teeth from U. arctos and U. spelaeus, in figure 4.58 we can see that those measurements that can tell apart the species in an almost perfect way. ADVP 2.9 falls into the variability of U. arctos. In figure 4.58 we can see the notch in the posterior part of the tooth that is typical from most U. spelaeus as depicted by Torres Pérez-Hidalgo, (1984).

Figure 4.56: Third lower molar (ADVP 2.9) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) oclusal, B) lingual and C) labial views.

Figure 4.57: Measurements for the third lower molar according to the system of Tsoukala & Grandal-d’Anglade (2002).

108

Table 4.11: Breadth and length of the third lower molar (ADVP 2.9) from the “bear two” of Algar do Vale da Pena compared to breadth an length of the same tooth of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal-d’Anglade, 1993; González & Martelli, 2003; unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Species L B Specimen Species L B ADVP 2.9 ? 18 15 995 U. spelaeus 26,2 18,9 LCF-003-1996 U.arctos 17,96 15,21 1310 U. spelaeus 24,2 19,2 LCF-002-1996 U.arctos 18,38 15,5 15 U. spelaeus 24 17 LCF-147 U.arctos 18,51 14,49 52 U. spelaeus 22,5 15 LCF-155 U.arctos 19,32 14,76 1588 U. spelaeus 24,7 19,2 EIX-018 U.arctos 19,25 14,6 1273 U. spelaeus 25,6 18,1 CGLL-006 U.arctos 20,52 15,68 526 U. spelaeus 25,5 19,4 CGLL-011 U.arctos 20,24 15,59 1580 U. spelaeus 25,4 17,5 SIPA-23 U.arctos 15,21 13,45 842 U. spelaeus 26,5 19,9 SIPA-53 U.arctos 19,59 15,68 1579 U. spelaeus 30 19,9 SIPA-17 U.arctos 21,8 16,1 297 U. spelaeus 28,5 19,8 SIPA-130 U.arctos 21,54 16,53 997 U. spelaeus 28,2 19,5 SH5-97-U29-7 U.arctos 19,42 14,39 843 U. spelaeus 30,3 20,6 SH5-97-U29-6 U.arctos 19,37 13,88 1589 U. spelaeus 30,6 21,1 TA-192 U.arctos 19,73 15,26 996 U. spelaeus 30,5 21,3 TA-200 U.arctos 19,52 14,68 1581 U. spelaeus 30,3 19,9 ? U.arctos 16,25 13,77 1108 U. spelaeus 26,8 20 ? U.arctos 16,71 13,93 1590 U. spelaeus 26,7 18,9 QU U.arctos 12,85 11,6 1169 U. spelaeus 26,5 19,6 QU U.arctos 13,2 11,4 1591 U. spelaeus 27 17,4 MZB 82-7005 U.arctos 16,99 14,35 1295 U. spelaeus 28,1 19,5 MZB 82-7005 U.arctos 17,27 14,36 841 U. spelaeus 27,4 19,3 153 U. spelaeus 25,31 17,91 19 U. spelaeus 27 20 163 U. spelaeus 23,87 17,8 270 U. spelaeus 29,96 20,4 del 146 U. spelaeus 26,4 20,3 269 U. spelaeus 30,27 20,14 del 801 U. spelaeus 22,17 17,1 268 U. spelaeus 29,09 21,95 del 144 U. spelaeus 25,42 19,44 267 U. spelaeus 28,36 22,28 del 518 U. spelaeus 28,78 20,66 2 U. spelaeus 27,11 18,74 489 U. spelaeus 26 18,6 3 U. spelaeus 27,12 18,64 1585 U. spelaeus 26 18 843 U. spelaeus 26,7 20,5 121 U. spelaeus 25,8 17,3 847 U. spelaeus 27,7 19,2 1102 U. spelaeus 26 20,3 848 U. spelaeus 25,4 18,7 994 U. spelaeus 26,4 18,5 1103 U. spelaeus 23,2 19,5 1133 U. spelaeus 26,2 21,1 1101 U. spelaeus 24,3 19,4

109

Figure 4.58: Breadth vs length in mm of the third lower molar ADVP 2.9 (black dot in the graph) and the same measurements of recent and fossil brown bears from NW Spain as blue diamonds (García Vázquez, 2015) plus cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal-d’Anglade, 1993; González & Martelli, 2003; unpublished own data) as purple triangles.

ADVP 2.10: Partial third phalanx

This partial claw (Figure 4.59) is the only postcranial remain recovered within the “bear 2” assemblage. It has a strongly marked base, a rectilinear profile and a relatively thin bone lamina compared to the base; all these features are closer to the claw morphology of brown bears than to cave bears (Figure 4.59 and figure 4.60) (Torres Pérez-Hidalgo, 1984). In figure 4.60 we can see the measurements graphically (on a cave bear claw) and in table 4.12 we can see the height and diameter of the base of the phalanx ADVP 2.10 compared to several of those phalanxes from U. arctos and U. spelaeus. We can see the biplot with those measurements at figure 4.61 where despite the great overlapping; the brown bear claws tend to have smaller heights for similar diameters, indicating broader claws at the cave bears.

Figure 4.59: Third phalanx (ADVP 2.10) from the “bear two” (Ursus arctos) of Algar do Vale da Pena in A) Distal; B) and D) Lateral; C) Proximal view.

110

Figure 4.60: Measurements for the third phalanx of ursids according to the system of Tsoukala & Grandal- d’Anglade (2002). Taken measurements are marked with a circle.

Table 4.12: Height and diameter of the third phalanx (ADVP 2.10) from the “bear two” of Algar do Vale da Pena compared to the height and diameter of the same phalanx of fossil and recent brown bears from NW Spain (García Vázquez, 2015) and cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal- d’Anglade, 1993; González & Martelli, 2003; unpublished own data). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen Species H DT Specimen Species H DT Specimen Species H DT

ADVP 2.10 U. arctos 18 11 SH5-98-S28 U. arctos 24,35 16,54 1194 U. spelaeus 29,3 15,0 CB-017 U. arctos 19,75 14,03 SH5-97-T29-30 U. arctos 19,84 13,02 1201 U. spelaeus 28,0 18,3

CB-018 U. arctos 21,22 15,24 SH5-98-U30-044 U. arctos 21,12 13,71 1329 U. spelaeus 29,0 17,5

CB-019 U. arctos 21,88 14,45 SH5-98-S28-105 U. arctos 21,69 14,12 1233 U. spelaeus 25,5 13,7

LCF-103 U. arctos 23,57 13,38 SH5-98-S28 U. arctos 26,48 16,35 1294 U. spelaeus 26,3 16,3

LCF-107 U. arctos 19,25 13,18 SH5-97-U29 U. arctos 19,74 13,01 390 U. spelaeus 22,1 12,6

SIPA-147 U. arctos 20,92 12,39 SH5-97-T29-20 U. arctos 19,01 11,87 145 U. spelaeus 22,4 13,8

SIPA-148 U. arctos 19,9 13,25 TA-Lu-c-17 U. arctos 17,87 13,05 89 U. spelaeus 22,5 14,2

SIPA-151 U. arctos 23,16 14,48 VA87/14D/74 U. arctos 26,09 16,21 188 U. spelaeus 21,5 13,0

SIPA-152 U. arctos 19,59 12,98 VA88/16E/231 U. arctos 24,23 15,39 1853 U. spelaeus 28,7 17,3

Pur-Lu-36 U. arctos 22,87 15,39 El Cuervo U. arctos 19,86 14,59 1854 U. spelaeus 26,1 16,2

Pur-Lu-37 U. arctos 24,94 16,26 El Cuervo U. arctos 23,33 14,43 1856 U. spelaeus 26,0 14,7

SH5-97-U29-24 U. arctos 21,83 12,73 El Cuervo U. arctos 22,73 15,33 1857 U. spelaeus 23,0 14,2

SH5-98-S28-079 U. arctos 22,76 14,56 El Cuervo U. arctos 19,55 14,64 1858 U. spelaeus 26,7 15,7

SH5-98-T30-003 U. arctos 22,58 13,27 1 U. spelaeus 17,1 17,92 1859 U. spelaeus 21,8 13,0

SH5-98-S28-103 U. arctos 26,3 16,54 2 U. spelaeus 14,79 15,18 1860 U. spelaeus 22,4 12,9

SH5-97-U29 U. arctos 21,37 13,3 743 U. spelaeus 13,38 14,52 1861 U. spelaeus 29,1 15,9

SH5-98-S28 U. arctos 21,15 15,31 651 U. spelaeus 23,0 14,5 1862 U. spelaeus 26,4 14,8

SH5-98-S28 U. arctos 21,16 15,3 652 U. spelaeus 28,4 15,7 1863 U. spelaeus 27,7 14,7

SH5-97-T29-16 U. arctos 21 12,64 805 U. spelaeus 25,3 14,2 1865 U. spelaeus 25,7 14,0

SH5-97-T29 U. arctos 22,19 13,17 806 U. spelaeus 25,9 14,7 1867 U. spelaeus 24,3 14,7

SH5-98-S28 U. arctos 21,91 15,35 807 U. spelaeus 24,3 14,1 1868 U. spelaeus 23,7 14,7

SH5-97-T29-19 U. arctos 22,3 13,18 901 U. spelaeus 26,3 14,7 1870 U. spelaeus 29,9 16,7

SH5-98-U28-012 U. arctos 19,95 13,92 1022 U. spelaeus 22,5 14,9 262 U. spelaeus 24,0 14,6

SH5-98-S28-045 U. arctos 22,8 15,02 1106 U. spelaeus 26,5 14,2 953 U. spelaeus 22,5 13,5

SH5-98-U30-009 U. arctos 23,75 14,09 1152 U. spelaeus 24,3 15,2 952 U. spelaeus 31,5 17,0

111

Figure 4.61: Height and diameter of the third phalanx (ADVP 2.10) (black dot in the graph) and the same measurements of recent and fossil brown bears from NW Spain as blue squares (García Vázquez, 2015) plus cave bears from A Ceza, Liñares and Eirós caves, Galicia, NW Spain (Grandal-d’Anglade, 1993; González & Martelli, 2003; unpublished own data) as crimson diamonds. 95% ellipses are drawn.

ADVP 2.10 is precisely in the limit for the ellipse for brown bears and is the smallest of all claws shown, possibly attributable to a juvenile.

ADVP 3.1: Partial distal epiphysis of right femur

The assemblage from whom this item comes (“bear 3”) is the worst preserved of all. Only one measurement according to the system of Tsoukala & Grandal-d’Anglade (2002) could be recorded. Only the most distal epiphysis is preserved as we see in figure 4.62. We can see the recorded measurement as well as a collection of the same measurements from other fossil and recent brown bears in table 4.13. They are presented graphically in figure 4.63, where we can see that it is difficult to classify ADVP 3.1 as belonging to a male or female, being it a nearly average individual between the biggest and the smallest in the sample.

Figure 4.62: Distal epiphysis of femur (ADVP 3.1) from the “bear three” of Algar do Vale da Pena in A) distal B) lateral views.

112

Table 4.13: Anteroposterior diameter of the distal epiphysis of the femur (ADVP 3.1) from the “bear three” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

Specimen DAPdist. ADVP 3.1 75 CB-008 80,55 LCF-006-1996 81,59 LCF-007-1996 82,74 LCF-173 69,97 LCF-174 72,15 EIX-010 65,36 CGLL-047 86,77 SIPA-2 99,07 SIPA-106 98,39 Pur-Lu-21 76,93 Pur-Lu-22 82,65 SH5-97-T28-12 106,3 SH5-98-T30-008 63,66 SH5-97-T29-58 62,18 SH5-98-S28-096 73,6 TA-Lu-c-7 69,57 TA-Lu-c-8 71,41 59,16

62,81

El Cuervo 62,22 El Cuervo 61,59

106,3 108 99,0798,39 98

86,77 88 82,74 82,65 80,5581,59 76,93 78 75 73,6 72,15 71,41 69,97 69,57 DAPdist. 68 65,36 63,66 62,18 62,8162,22 61,59 59,16

7 8

- -

21 22

58 12

c c

- -

106

010 047

008 173 174

008 096

- -

- - -

- -

1996 1996

Lu Lu

- -

Lu Lu

- -

SIPA

- -

T29 T28

EIX

- -

LCF LCF

S28

T30

SIPA

ADVP 3.1 ADVP

El Cuervo El Cuervo El

CGLL

TA TA

006 007

Pur Pur

97 97

- -

LCF LCF

SH5 SH5 SH5 SH5

Figure 4.63: Anteroposterior diameter of the distal epiphysis of the femur (ADVP 3.1) and the same measurement of recent and fossil brown bears from NW Spain (García Vázquez, 2015).

113

ADVP 3.2: Vertebral centrum

This vertebral centrum (Figure 4.64) is partially covered by calcareous crust, which prevented it to be heavily eroded like the other centrum recovered (ADVP 3.3). The neural arch is nearly gone and one the cranial/caudal sides of the vertebra is highly damaged. The better preserved part of the vertebra is heart-shaped centrum.

Figure 4.64: Vertebral centrum (ADVP 3.2) from the “bear three” of Algar do Vale da Pena in A) dorsal and B) anterior views.

ADVP 4.1: Proximal epiphysis of second right metatarsal

This metatarsal fragment (Figure 4.65) is slightly worn in its proximal part; we have its measurements graphically in figure 4.66 and numerically in table 4.14. The erosion apparently caused a similar phenomenon tan the one observed in ADVP 1.7; the metatarsal is smaller in size than the ones of fossil and recent brown bears from NW Spain (Figure 4.67), almost like a juvenile individual.

Figure 4.65: Proximal epiphysis of second right metatarsal (ADVP 4.1) from the “bear four” of Algar do Vale da Pena in A) Interior; B) Ventral; C) Exterior and D) Dorsal.

114

Figure 4.66: Measurements for the second metatarsal according to the system of Tsoukala & Grandal- d’Anglade (2002).Taken measurements are marked with a circle.

Table 4.14: Measurements of the second metatarsal (ADVP 4.1) from the “bear four” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

L DTprox. DAPprox. DTdia.min. DAPdia. DTdist. DAPdist. DTart.dist.

ADVP 4.1 9 20 9 8

SIPA-92 16,15 22,73 13,5 12,3

SIPA-95 16,38 13,25 12,18

SH5-98-97-S29-009 76,25 13,53 22,78 12,92 9,04 18,19 15,02 14,27 SH5-97-T29-7 71,99 11,31 20,1 12,3 8,14 17,11 14,04 14,46 SH5-98-S28-044 83,73 16,7 27,17 13,91 11,3 21,08 18,5 16,88 SH5-97-T29-088 71,84 13,32 20,64 12,13 8,28 17,51 14,31 14,12 SH5-98-S28-006 75,59 15,17 22,39 12,88 8,82 18,52 14,81 14,1 TA-Lu-c-23 69,28 13,06 21,63 11,01 7,75 18,07 15,25 10,95 TA-Lu-c-24 69,74 12,68 21,98 11,16 8,05 18,38 14,96 11,52

115

Figure 4.67: Anteroposterior vs transversal diameter of the proximal epiphysis (upper) and diaphysis (bottom) of the second metatarsal, ADVP 4.1 (blue square) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015).

ADVP 4.2: Damaged proximal epiphysis of fifth right metatarsal

This proximal part of the metatarsal (Figure 4.68) was reconstructed from several fragments and could not be entirely repaired; we have its measurements graphically in figure 4.69 and numerically in table 4.15. In figure 4.70 we can see a graphical comparison with brown bears from NW Spain, demonstrating a similar situation as with the other metapodial bones (ADVP 1.7 and ADVP 4.1), the Portuguese specimen is by far the less robust.

Figure 4.68: Proximal epiphysis of fifth right metatarsal (ADVP 4.2) from the “bear four” of Algar do Vale da Pena in A) Interior; B) Ventral; C) Exterior and D) Dorsal views.

116

Figure 4.69: Measurements for the fifth metatarsal according to the system of Tsoukala & Grandal-d’Anglade (2002). Taken measurements are marked with a circle.

Table 4.15: Measurements of the fifth metatarsal (ADVP 4.2) from the “bear four” of Algar do Vale da Pena compared to the same measurement of fossil and recent brown bears from NW Spain (García Vázquez, 2015). Measurements are in mm according to the system of Tsoukala & Grandal-d’Anglade (2002).

DAPprox. DTdia.min.

ADVP 4.2 21 10 LCF-136 24,29 11,42 SH5-97-T29-36 24,17 11,57 SH5-97-T29-087 24,36 11,82 SH5-98-S28-076 31,56 14,51 TA-Lu-c-20 24,58 11,18 TA-Lu-c-19 25,74 11,09 VA87/LIMPIEZAEXTERIOR/4 26,43 12,75 24,79 11,37

25,85 11,23

Figure 4.70: Anteroposterior proximal diameter vs transversal diameter of diaphysis of the fifth metatarsal, ADVP 4.2 (blue square) and the same measurements of recent and fossil brown bears from NW Spain as black dots (García Vázquez, 2015).

117

4.6 The claw marks

In several walls of the cave, where there is an accumulation of clays, we can find several parallel markings that appear mostly in groups of four or five grooves. Octávio Mateus has counted 189 of those groups alongside the walls of the cave. They are interpreted as claw marks left behind by bears that were navigating its way across the darkness of the cave or digging beds for hibernation, the location of the largest concentrations of them can be seen as the numbers 2, 4, 5 and 7 at the cave map of figure 4.4. In figure 4.71 we can see some of the best examples. They vary greatly in size, being mostly between five to ten centimeters in width. Without the calcareous flowstones that cover many clay areas of the cave the number of tracks would probably be in the hundreds. Given that U. arctos tend to not let behind nearly as many of this marks as U. spelaeus (Fosse et al., 2004); this numbers point towards the use of the cave repetitively during a long period of time. These are the first of this type of marks described in Portugal (Neto de Carvalho, 2018).

Figure 4.71: Claw marks in the walls of the Algar do Vale da Pena. Some of them are today partially covered by calcareous crusts. Scale bar 7 cm.

118

4.7 Discussion

The cave was sealed until recently, so we do not know its original entrance (or entrances), but the claw marks in the walls point towards that the bears entered the cave by their own means (probably for hibernation) and died of natural causes inside, so the cave did not acted like a natural trap. The bones of most assemblages also point towards this hypothesis, since they do not show signs of important transport and are still articulated or closely associated; with the exception of “bear 2” which shows clay below the calcareous crust and eroded bones and teeth, and “bear 3” whose bones are in the worst condition of all groups of remains. Those exceptions could be explained by water flowing and eroding in situ the bones or lightly transporting them. The number of tracks point to the use of the cave as den by bears, maybe during generations.

Talking about the characterization of the bear population, first we have to address that the minimal number of individual is two, in base of the recovered cranial parts and the articulated skull that is still in the cave. But as pointed earlier it is likely that the real number is between four and five, in base of the separation between assemblages, being it of several dozens of meters and in the absence of important transportation.

Now we will discuss each studied assemblage and bone in the sections 4.4 and 4.5:

“Bear” one: Does not present signs of transportation, with some articulated parts. It can be assumed to be a single animal.

- ADVP 1.1: Partial left scapula: Not measured. Its morphology is too distorted by conservation.

- ADVP 1.3: Right scapholunate bone: Morphology ruling out U. thibetanus, measurements are not congruent with U. spelaeus; the most likely affinity is to an average female U. arctos.

- ADVP 1.4: Right hamate bone: Only measurement is congruent with the hamate of a female U. arctos.

- ADVP 1.6: Second phalanx. It is likely a bone of hind limb. Not useful for specific or sex determination.

- ADVP 1.7: Proximal epiphysis of third right metatarsal. Smallest of all the considered bones, juvenile sized U. arctos.

- ADVP 1.8: Distal epiphysis of right radius. Slender morphology is ruling out U. spelaeus. The epiphysis is solidly fused, so it belongs to a mature individual. It is also less robust than the other Portuguese specimens; measurements are congruent with a small female U. arctos.

- ADVP 1.9: Distal part of diaphysis of left ulna. Slender morphology is ruling out U. spelaeus. Only recovered measurement is the smallest of the sample, congruent with a small or young female U. arctos.

Putting together all this lines of evidence we can attribute this assemblage to a small, mature female of U. arctos.

119

“Bear” two: Presents some signs of erosion and could have been subject to some minor transport. It could include the remains of more than one specimen (all are cranial elements except one third phalanx).

- ADVP 2.1: Posterior part of right hemimandible. The width of the mandibular ramus relative to its height is typical for a female U. arctos. The height of the mandibular condyle relative to height of the mandibular ramus is also typical of a below average U. arctos and different to the taller condyles in similarly sized ramus of U. spelaeus. The robustness of the mandibular corpus is above the average, but well below the values of U. spelaeus. Therefore this mandible could be attributed to a female U. arctos.

- ADVP 2.2: Parietal and occipital. The piece is not very diagnostic at specific level, but the grade of fusion of the sutures points towards an adult specimen. Although this, the nucal and parietal crests are not developed which points rule out a big mature male.

- ADVP 2.3: Temporal bone. Not diagnostic, although the calcareous crust overlaying the broken parts of the bone is one of the best examples of the taphonomical processes in the cave.

- ADVP 2.5: Heavily damaged canine. This piece is too fragile and damaged to be measured.

- ADVP 2.6: Left upper fourth premolar. Long, yet slender, compared to the sample this specimen is so particular that it is difficult to assign to a certain species.

- ADVP 2.7: Left upper first molar. Its morphology differs from U. thibetanus and U. spelaeus; therefore it can be classified confidently to U. arctos. The measurements point towards a small U. arctos.

- ADVP 2.8: Right lower second molar. Its dimensions are far from the ones of U. spelaeus, they are in the lower spectrum from U. arctos.

- ADVP 2.9: Left lower third molar. This piece has the typical dimensions of an average U. arctos, far from what we can expect from a cave bear.

- ADVP 2.10: Partial third phalanx. Its morphology and measurements are totally coincident with the claw of a small sized brown bear

This assemblage could belong to more than one specimen yet, it will be tentatively attributed to a mature and small to average female brown bear, similar to what we see in the first assemblage.

“Bear” three: Worst preserved assemblage of all, it could belong to weathered and moved remains of “bear” one.

- ADVP 3.1: Partial distal epiphysis of right femur. Only one measurement could be recorded, being it almost perfectly into the average for a brown bear.

- ADVP 3.2: Vertebral centrum. It is partially covered by calcareous crust not possible to classify.

120

“Bear” four: Most of the remains of this assemblage are closely associated and covered by thick layers of calcareous crust. It is unlikely that they suffered meaningful transportation.

- ADVP 4.1: Proximal epiphysis of second right metatarsal. This piece is noticeably smaller than any brown bear specimen sampled in NW Spain for any measurement considered.

- ADVP 4.2: Damaged proximal epiphysis of fifth right metatarsal. This piece is noticeably smaller than any brown bear specimen sampled in NW Spain for any measurement considered.

Both measured metatarsals of this assemblage are surprisingly small. The humerus, femur and hip bones that comprise the rest of it do not look proportionally smaller to other bears, but given that they have not been recovered from the cave, neither measured nor studied, this is only an appreciation. This assemblage cannot be attributed to any species with certainty according our current data.

“Bear” five: Is still encased in layers of calcareous crust and could not be studied for this work.

Therefore the fossil bear population in the cave is formed by brown bears of smaller size that the ones recorded in Northwestern Spain (García Vázquez, 2015) and most likely dominated by adult female animals.

It is curious to compare these facts with the works of Torres Pérez-Hidalgo, (1979) and Cardoso, (1993) that concluded that the Portuguese brown bears (mainly from Fontainhas, Furninha and some of Serra de Molianos) were more robust that the Spanish bears of comparable age.

Given the absence of a reliable stratigraphy for the cave, the age of the fossil remains could only be addressed thanks to absolute methods like isotopic or ESR (Electron Spin Resonance) datings.

4.8 Conclusions

- The Algar do Vale da Pena is a cave recently reopened and is a new fossil brown bear locality in Portuguese Estremadura. - 37 items have been recovered from the cave, including 20 informative fossil remains. - The fossil remains probably belong to five individuals considering its distribution in the cave, althought the minimal number of individuals in the cave is two, in base of the cranial remains. - The findings can be ascribed to Ursus arctos in base of morphology (A first upper molar with clearly individualized metacone-paracone and protocone-hypocone, but without a marked and big cingulum; mandible with a low condyle compared with the mandibular ramus, slender long bones without marked muscle insertions, occipital bone with a shallow nucal crest and straight third phalanxes with robust proximal parts) and metrical comparisons (In several measurements of 20 bones from the assemblage). - Two of them are congruent with the measurements and characteristics of mature, yet below the average sized, female bears. - The population is formed by relatively small animals (compared to their Spanish counterparts) in contrast with the observed in other Portuguese localities by Torres Pérez-Hidalgo, (1979) and Cardoso, (1993). - The claw marks in the walls of the cave are the first of its type documented in Portugal. - The great numbers of the claw marks, the morphology of the cave and the taphonomy of the remains point toward repetitive occupation of the cave by bears during a long time period. - The presence of Ursus arctos and the thickness of the calcareous mantles over the bones point towards a Pleistocene age, although the absence of informative stratigraphy prevents the precise dating of the assemblage for the moment.

121

4.9 Further work

Neither the works in the cave, nor the study of the fossil findings is over.

As presented in the section 4.4, there are two in situ jackets containing remains of the “bears” two and three, and two separated blocks including material of “bear” three and the main skull of “bear” one already in the cave. This material as well as some other samples is scheduled to be removed from the cave with the help of the GPS between May and November of 2019.

In December of 2018 samples of two rib fragments (ADVP 1.10 and ADVP 3.4) belonging to the assemblage were analyzed in the SAI (Servizos de Apoio a Investigación) of the University of A Coruña by mass spectrometry in order to determine the amount of Nitrogen present in them. If the amount of this element was above 0,05% it could be assumed that the samples had enough collagen or other proteins in them to be used in a 14C radiometric analysis; but as we can see in table 4.16, it was not the case.

Table 4.16: Nitrogen and Carbon percentages in the samples of ADVP 1.10 and ADVP 3.4 from Algar do Vale da Pena (Portugal).

Weight Sample N.º SAI (mg) % N % C AVDP 3.4 2018/45957 4,965 < 0,05 1,85 AVDP 1.10 2018/45958 5,096 < 0,05 3,24

Therefore the petrosal part of the temporal bone of ADVP 2.3 could be used next in order to prove if some proteins survived, or another absolute dating technique (ESR) could be used in other to date the fossils, although this second path of action would negate the possibility of isotopic diet studies or ancient DNA analysis.

Also, once all the material of the cave is recovered, a larger and more complete study of the Portuguese Ursus material (including at least the specimens of Fontainhas, Furninha and Serra de Molianos) with measurements taken according the system of Tsoukala & Grandal-d’Anglade (2002), would be useful for a better understanding of the fossil bear population from the Algar do Vale da Pena in its immediate geographical context.

122

5. SANTA MARGARIDA: A NEW MICROVERTEBRATE LOCALITY FROM ALGARVE

5.1 Overview

On this chapter a new and abundant microvertebrate assemblage from Algarve will be described for first time. Special emphasis will be given to the discovery and preparation methods, taking into account the peculiar mode of preservation of the remains (Bone breccias). Then a taxonomic identification will be provided for some of the specimens, alongside a list of groups present. Inferences will be made about the age and origin of the association. Finally some conclusions will be made as well as some remarks about further work that could be done in the locality.

5.2 Introduction: Location, discovery, geological settings and nature of the site

The site of Santa Margarida is located near the village of the same name in the freguesia of Alte, municipality of Loulé, Algarve region, South Portugal, 300 m above the sea level (Figure 5.1).

Figure 5.1: Location of the site of Santa Margarida. Figure by Cátia Ribeiro. Credit of the maps, google maps.

The discovery of the site happened following this series of events: During a conference in Algarve in 2016 Octávio Mateus was approached by a local photographer, Jorge Graça, who informed him of the presence of multitude of bones in the blocks of a wall near Santa Margarida. Despite the efforts for finding the site that year it was not possible. In 2017 the team went back again to search for the wall and this time they were able to find it, photograph several blocks and take other back to the Department of Earth Sciences at FCT-UNL (Figure 5.2) that where prepared during the next months to extract the microfossils. In May of 2018 a team of the FCT went back again to the place and discovered that the source of the blocks was just adjacent to the wall. The cemented terra rossa blocks with breccified bones and dental remains were found only in a span of 20 meters of the wall, with the point of maximum concentration being found in 2017, plus other blocks with bones scattered inside the area limited by it. More blocks were collected.

123

Figure 5.2: Maximum concentration point of fossil bearing blocks in the wall of the site as it was discovered in 2017. Note the difference of color between the bone bearing breccified blocks of cemented terra rossa and the grey tone of the calcareous rocks. Photo by Octávio Mateus.

The locality of Santa Margarida is situated in the karst of the Picavessa Formation (Figure 5.3), Early Jurassic in age and described as formed by calciclastic and micritic rocks, with rare arrecifal and oolitic limestones (Terrinha et al., 2006). As the successive expeditions to the area have shown, the Quaternary microvertebrate fossils are conserved in heavily breccified deposits of cemented terra rossa that is covered by a crystalline crust of calcareous material and a thin clayish soil (Figure 5.4). Both layers and the soil are overlaying the limestone rocks of the Picavessa Formation and vary greatly in thickness and quantity of fossils across the area, to fully understand its fine stratigraphy a much bigger excavation should be done.

Figure 5.3: Location of Santa Margarida site. In blue the calcareous Picavessa Formation. Modified by Cátia Ribeiro from the geologic map of Algarve by LNEG.

124

Figure 5.4: Left; comparison of freshly cut cemented terra rossa block and the underlying grayish calcareous rock of the Picavessa formation. Right; calcareous crust covering the “terra rossa”.

The wall where the fossils were found is a traditional algarvian wall. These walls were made with in situ materials (usually limestone) and depending of their orientation and the topography they have been traditionally used for a) forming terraces in steppe terrain, in order to reduce the soil erosion and retain the soil water for cultivation and b) To separate proprieties. The walls of the first type can reach three meters in height. They are perpendicular to the pending of the slope in order to retain the soil (Gonçalves et al., 2018). The wall where the discovery was made runs parallel to the pending of the slope, not perpendicular, making it a representative of the second type. These “property walls” are usually between 0,6 to 1,5 m wide and 0,6 to 1,2 m in height (Gonçalves et al., 2018)., which is congruent with the observed. In the terrain delimited by it there are some deteriorated and old fruit trees that are indicative of an abandoned traditional cultivation field in Algarve: Fruit trees, and between them legumes planted to fix Nitrogen in the soil (Gonçalves et al., 2018).

The abundance of small vertebrates in the blocks was clear. For every square centimeter of exposed surface, the richest blocks could have several fossils on it. Also, some blocks with larger bones of centimeters of diameter could be recovered (Figure 5.5).

Figure 5.5: A) Abundant microfossils in rich block; B) Rabbit humerus in situ; C) and D) larger long bones protruding from block.

125

Using the Dino-lite™ (A device designed for taking pictures of small fossils) it was possible to take some pictures to the microfossils in situ, before any preparation (Figure 5.6) which gave a first idea of the taxonomic groups that could be identified. Those groups included rodents, “insectivores”, lizards and even gastropods (not shown in the picture).

5.3 Preparation methods

In spite of the good condition in which several microfossils could be observed alongside the surface of the blocks it was necessary to extract them from the matrix to properly study them. First, the mechanical preparation was experimented in order to recover as many of the largest remains as possible. They were consolidated using PARALOID B72™ and extracted with the help of a air-scribe (Figure 5.7). Some arvicoline and squamate mandibles were extracted using this method, but it was not useful for the recovery of the small fossils found in the blocks (Most of them).

Figure 5.7: Left, consolidated arvicoline mandible with PARALOID B72™; center, extraction of the mandible using a air-scribe and right the extracted mandible.

126

Therefore, the chemical preparation was started. The first step was to choose the chemical product to be used. A test using five products was prepared, including: water (as control), acetic acid at 5% of concentration in water, formic acid at 5% of concentration in water, Hydrogen peroxide in a commercial concentration of 30% and a commercial anti-calcareous product (Water softener) at 50% in water. Samples of cemented terra rossa were weighed before and after 24 hours in the respective products (therefore discounting the dissolved and disaggregated rock) and two days of drying. The results can be seen in table 5.1.

Table 5.1: Results of the tests for determining the product to be use in the chemical preparation of Santa Margarida samples.

Test Initial weight (g) Final weight (g) % of weight lost water (control) 103,102 104,504 + 1% acetic acid at 5% 51,228 50,036 2% formic acid at 5% 38,846 34,872 10% Hydrogen peroxide at 30% 48,793 45,864 6% Commercial anti-calcareous product at 39,231 24,343 38% 50%

After this test the water softener was chosen. A second minor qualitative test was conducted to proof the best times for the product to be used without affecting too much the fossils, but maximizing the amount of dissolved and disaggregated rock. Samples were submerged in the product for 12, 24 and 48 hours. The best results (less fragile and dissolved bones observed, but with good amount of matrix dissolution) were the ones of the 24 hours test.

Therefore the processing of the maximum amount of sediment in the span of three months (September-December of 2017) was carried in the Museum of Lourinhã using the facilities of the lab. In the figure 5.9 we have a workflow of how the process was running. At every given time two kilograms of rocks were being processed, one of them in the product and other one being de-acidified in water, to prevent that product trapped in the pores of the rock and bones kept reacting all the time.

After the entire preparation process and the picking of the fossils, they were classified in broad groups (postcranial bones, “insectivores”, lizards, gastropods… etc) and then the most promising of them were further cleaned by ultrasound. Not all of them were subject of this preparation, because some were already too fragile. Finally the teeth were mounted in plastic boxes over thin strands of Blu- Tack™ in oclusal view and then photographed with the binocular lenses in the department of Earth Sciences of the FCT-UNL (Figure 5.8) for further identification. This was a complex task given the fragmentation of the sample.

Figure 5.8: Right, rodent teeth mounted in the plastic box and other fossils stored in similar settings (Scale 7cm); left, the binocular lenses used to take the pictures (Photo by Alexadre Guillaume).

127

Figure 5.9: Workflow followed in the preparation of the Santa Margarida material (Portugal) in the Museum of Lourinhã.

128

5. 4 Results

A total of around 0,5 kg of sediment was disaggregated from the blocks subject to treatment and picked for fossils. We can see the amount of recovered sediment in figure 5.10.

Figure 5.10: Amount of processed sediment from Santa Margarida (Portugal) at Museum of Lourinhã.

On this section a taxonomic identification of some of the microvertebrate and gastropod remains will be given, as well as the number of dental remains recovered for broader groups that we have in table 5.2. It is noteworthy that 90% of all remains belong to rodents.

Table 5.2: Number of specimens from broad groups recovered from the processed sediment of Santa Margarida.

Gastropoda Cuvier, 1797 1 complete shell, 3 smaller fragments, 1 bisected shell in situ. Lacertidae Gray, 1825 1 semi-complete mandible, 3 mandibular fragments Chiropera Blumenbach, 1779 4 unicusp teeth, 1 molar Soricidae Fischer, 1817 20 isolated teeth or mandible parts with attached teeth Rodentia Bowdich, 1821 273 tooth fragments, isolated teeth, and mandibles with teeth Lagomorpha Brandt, 1855 1 distal humerus.

In spite of this abundance of material (~0,6 fossils for gram of sediment) most of the remains became impossible or too difficult to identify at generic or specific level, so only the better preserved ones were subject to a more precise identification presented here.

129

Gastropoda Cuvier, 1797

In spite of the aggressive treatment with water softener, one nearly pristine snail shell was recovered during the picking (Figure 5.11), as well as other three fragments of similar sized shells. This shell was identified as belonging to the snail Paralaoma servilis (Shuttleworth, 1852) in base of the marked umbilicus, the presence of ribs and the overall geometry of the shell (Callapez, personal communication, 2019).

Figure 5.11: Paralaoma servilis from Santa Margarida in A) lateral, B) ventral and C) dorsal views.

Another gastropod was found and photographed in situ in one of the blocks in the field (Figure 5.12). This was identified by Pedro Callapez in base of the whorl morphology as Pomatia elegans (Müller, 1774).

Figure 5.12: Bisected shell of Pomatia elegans from Santa Margarida.

130

Lacertidae Gray, 1825

A relative big lizard mandible was separated mechanically from one of the blocks (Figure 5.13). It bears the classic lacertid features: pleurodont condition of the teeth, which are cylindrical and monocuspid; plus a Meckelian groove opened on its whole length (Blain et al., 2007). The fragility of the mandible prevented the usage of ultrasounds for cleaning it; therefore it was not possible to take reliable measurements of the piece that could have allowed further identification. Another three mandibular fragments that bear similar characteristics were picked from the disaggregated sediments. In figure 5.14 we can see the most complete of those fragments.

Figure 5.13: Lacertid mandible from Santa Margarida (Portugal) in lingual view.

Figure 5.14: Lacertid mandible from Santa Margarida (Portugal) in A) lingual and B) oclusal views.

The teeth of the mandible were measured with the help of the software ImageJ, following the method presented by Blain et al., 2007. The obtained measurements and ratios are presented in the table 5.3

Table 5.3: Measurements and ratios of the most complete lacertid mandible fragment picked from the disaggregated sediments of Santa Margarida, Portugal, following the methods of Blain et al., 2007: h, is height of the teeth; d, its diameter and a, the height of the teeth that protrudes over the dental crest. All measurements are in mm.

Teeth h d a d/h a/h 1 0,638 0,347 0,266 0,544 0,417 2 0,729 0,323 0,287 0,443 0,394 3 0,756 0,26 0,301 0,344 0,398 4 0,860 0,301 0,364 0,350 0,423 5 0,945 0,364 0,371 0,385 0,393 6 1,037 0,357 0,364 0,344 0,351

Average 0,828 0,325 0,326 0,393 0,393

131

Comparing these ratios with the ones presented by Blain et al., 2007 for the genera Podarcis Wagler, 1830 (d/h=0,269; a/h= 0,439); Acanthodactylus Daudin, 1802 (d/h=0,280; a/h= 0,330) and Psammodromus Hallowell, 1852 (d/h=0,301; a/h= 0,327) it is possible to conclude that the Santa Margarida lacertid (d/h=0,393; a/h= 0,393) does not belong to either of those genera.

Chiroptera Blumenbach, 1779

Five unicusp teeth and one molar assigned to bats have been recovered in the disaggregated sediments of Santa Margarida. The unicusp teeth are ascribed to Chiroptera in base of the marked cingulum (Hillson, 2005) (Figure 5.15).

Figure 5.15: Unicusp teeth ascribed to a bat from Santa Margarida (Portugal) in meso-lingual view (Image by Cátia Ribeiro).

The recovered molar (Figure 5.16) bears some typical characters of the upper molars of bats, like being dilambdadont, with tall “W” -shaped ridges and a low lingual shelf (Hillson, 2005). The parastyle is curved, which is typical for European bats and the parastyle/paracone length is shorter than the metacone/metastyle length. All the main cusps are quite robust. There is a thick lingual/distal cingulum. All these are congruent with a right M2; second upper molars are generally more elongated than first upper molars, which are more compressed mesio-distally (Piskoulis, personal communication, 2019). The morphology is similar to the M2 some species of Myotis Kaup, 1829 with a strong and marked postprotocrista (Gunnell et al., 2011).

Figure 5.16: Right upper second bat molar from Santa Margarida (Portugal) in oclusal view. The rectangle marks the w shaped labial ridge and the circle the lingually recurved hypocone shelf and the strong postprotocrista. Metastyle/paracone (1,3 mm) and metastyle/lingual (2 mm) lengths are marked. Drawing by Pavlos Piskoulis.

132

Soricidae Fischer, 1817

A total of 20 mandibular and dental remains are ascribed to this group. In the table 5.4 we can see the different groups in which they are classified.

Table 5.4: Classification of the different Soricidae remains recovered from Santa Margarida (Portugal).

Soricinae indet. 3 tooth fragments and 1 lower molar Sorex sp. 1 right upper incisive Crocidura sp. 5 mandible parts with attached teeth, 5 lower molars, 1 upper molar, 4 upper incisors

In the figure 5.15 we can see three upper incisors. The first one (Figure 5.17 A) can be attributed to Sorex Linnaeus, 1758, in base of the fissident (Having a small medial cupule) upper incisor with red pigmented enamel, plus the size of the specimen (Reumer, 1984). The second specimen (Figure 5.17 B) can be assigned to the genus Crocidura Wagler, 1832 in base of the size, the unpigmented enamel and the marked and slightly undulated cingulum in the posterior bucal part (Reumer, 1984). The third specimen (Figure 5.17 C) bears similar characteristics than the previous one, but with a less robust morphology. According to Raquel Moyá-Costa (Personal communication, 2019) it probably belongs to an older individual that could also be representative of a different species of the genus.

Figure 5.17: Soricidae upper incisors recovered from Santa Margarida (Portugal) in labial view. A) Sorex, B) and C) Crocidura.

In figure 5.18 we can see an upper molar that broke during the process of taking the pictures. Yet, it is possible to distinguish some features that allow it to be classified, like a hypocone situated in a postero-lingual position relative to the protocone and poorly individidualized of it. There is also a deep valley between the protocone and the metacone and a crest connecting it with the paracone. It´s trapezoidal shape is congruent with a second upper molar. Given the absence of enamel coloration, it´s size and the previously commented characters it is assigned to the genus Crocidura (Reumer, 1984).

Figure 5.18: Broken second upper molar of Crocidura recovered from Santa Margarida (Portugal) in oclusal view. Labial part to the right.

133

Some teeth recovered from the locality were still attached to parts of the mandible, but only one mandible part conserved with more than one tooth (Figure 5.19). It presents the P4, M1 and M2 (Moyá-Costa, Personal communication, 2019). The P4 is tetrahedral shaped; the M1 and M2 have the valley openings well above the cingulum in labial view. The cingulum is well marked and undulated in the labial part and weak in the lingual. The mandible also poses a small mandibular foramen situated in labial view in the transition of the P4 to the M1. All this characteristics allow us to attribute this mandible fragment to the genus Crocidura (Reumer, 1984).

Figure 5.19: Fragment of Crocidura mandible (bearing P4, M1 and M2) recovered from Santa Margarida (Portugal) in A) oclusal, B) lingual and C) labial views.

Rodentia Bowdich, 1821

The rodents comprise the immense majority of the fossil remains recovered in the locality of Santa Margarida (+90%). In the next table (Table 5.5) we have the breakage of the rodent dental remains from Santa Margarida.

Table 5.5: Classification of the different rodent dental remains recovered from Santa Margarida (Portugal).

Rodentia indet. 152 incisors or parts of incisors, including 41 incisor tips Allocricetus bursae 1 upper left first molar Eliomys quercinus 1 lower premolar, 1 second lower molar, 2 second upper molars, 1 indeterminate molar Apodemus cf. 6 third lower molars, 2 second lower molars, 15 first lower molars, 5 third upper molars, 9 sylvaticus second upper molars, 5 first upper molars, 8 indeterminate teeth Arvicolinae indet. 46 teeth and 22 fragments Victoriamys chalinei 1 right first lower molar Iberomys huescarensis 1 Left hemimandible with first and second molar Iberomys brecciensis 1 Left hemimandible with first molar and 1 right first lower molar Iberomys cabrerae 1 left first lower molar

134

The rodent chisel-like incisors are highly characteristic, only lagomorphs are similar in European Quaternary faunas and its size mostly prevents confusion. In Figure 5.20 we have an incisor of a rodent from Santa Margarida, 152 incisor parts (including 42 tips) have been recovered during the picking. These teeth are mostly similar in all rodent species and are not usually used in the fossil identifications when finding isolated (Hillson, 2005), although new techniques are developing quickly in that direction and they might not be considered nearly useless anymore soon (Paine et al., 2019).

Figure 5.20: Tip of a rodent incisor from Santa Margarida (Portugal).

The molars and premolars of the rodents are much more characteristic of their different species and allow its specific identification (Hillson, 2005). In the Santa Margarida sample a total of 51 teeth (6 third lower molars, 2 second lower molars, 15 first lower molars, 5 third upper molars, 9 second upper molars, 5 first upper molars and 8 indeterminate teeth) whose morphology is congruent with Apodemus Kaup, 1829 were recovered ( Hillson, 2005). Moreover, in the first lower molars we can see six low principal cusps, which is a characteristic of Apodemus (López-García, 2008). All the first upper molars have a developed tubercle 7 (t7) on them as well as a t7–t4 connection in almost all cases (Figure 5.21) (Piñero et al., 2015). In all the lower first molars the anterocentral cusp is connected to the anterolabial and anterolingual cusps, meanwhile in the Apodemus mystacinus (Danford and Alston, 1877) from the Quibas site (Spain) around 1/3 of them have this cusp totally isolated (Piñero et al., 2015). All these characteristics allow us to classify the teeth as belonging to Apodemus cf. sylvaticus.

Figure 5.21: A) third lower molar and B) first upper molar of Apodemus cf. sylvaticus from Santa Margarida (Portugal) in oclusal view.

135

Another group of rodents present in Santa Margarida are the dormice; Gliridae Muirhead, 1819. These rodents are characterized by their cheek teeth: “brachydont, low and table-like in most genera, with multiple ridges running buccolingually across the occlusal surface” (Hillson, 2005). There are 1 lower premolar, 1 second lower molar, 2 second upper molars and 1 indeterminate molar present in the sample. The teeth of Santa Margarida (Figure 5.22) present highly curved and continuous ridges without extra cusps besides the four main ones and strongly triangular shaped premolars, which allow us to ascribe them to Eliomys quercinus (Linnaeus, 1766) (Hillson, 2005; Piñero et al., 2015).

Figure 5.22: Left upper first or second molar of Eliomys quercinus from Santa Margarida (Portugal) in oclusal view.

There is at least another non arvicoline rodent present in the sample, a cricetine rodent represented by an upper first molar (Figure 5.23) that bears the characteristic low crown with six cusps aligned in two rows (Hillson, 2005). It has not anterostyle, the anterolophule is clearly bifurcated and the lingual side (judging for the position of the anterolophule relative to the protocone) is longer than the labial one. These characteristics are typical of the species Allocricetus bursae Schaub, 1930 according to Cuenca- Bescós, (2003).

Figure 5.23: Left upper first molar of Allocricetus bursae from Santa Margarida (Portugal) in oclusal view.

136

The arvicolines are the most diverse group of rodents in the sample of Santa Margarida with 73 mandibles, isolated teeth and teeth fragments. Their teeth are highly characteristic with multiple prismatic elements arranged in a very long oclusal surface with characteristics folds that allow its identification to specific level (Hillson, 2005). In the figure 5.24 we have a general schematic draw with the characteristics of the teeth of this group. In the Santa Margarita sample, most of the teeth were too broken to be useful for identification, but at least four species were recorded:

Figure 5.24: Nomenclature and measurements for an arvicoline first lower tooth according to López-García et al., (2015). ABBREVIATIONS: a: length of the anteroconid complex; ACC, anteroconid complex; AC, anterior cap; BRA, buccal reentrant angle; BSA, buccal salient angle; c: width of the opening of triangles T4-T5; L, length; La, mean width of T4; Li, mean width of T5; LRA, lingual re-entrant angle; LSA, lingual salient angle; PL, posterior lobe; TTC, trigonidtalonid complex, T1-T7, triangles 1-7; W, width.

The first tooth is a right first lower molar, whose frontal part has been worn down (Figure 5.25). Morphologically it posses three closed triangles (T1-T3) but the T4-T5 are very confluent, with a wide connection to the anterior cap, that translates in broad angles of the BRA3 and LRA4. The anterior cap itself is wide but short. The labial and distal triangles are similar in size. There is not LRA5. These characters are diagnostics of Victoriamys chalinei (Alcalde et al., 1981) (López-García et al., 2012; Martin, 2012).

Figure 5.25: Right lower first molar of Victoriamys chalinei from Santa Margarida (Portugal) in oclusal view.

137

The second tooth is a first lower molar (Figure 5.26), attached to a hemimandible which also bears the second molar and was mechanically recovered from the block. It posses lingual triangles that are larger than the labial ones. The T4-T5 connection is broad and evident but its connection to the anterior cap is narrow. The BRA4 is absent and so is the T6. The LRA5 is small, therefore the T7 is almost non-separated of the anterior cap, but both are present. The anterior cap is short and triangular. These characteristics are congruent with Iberomys huescarensis (Ruiz-Bustos, 1988) according to López-García et al., (2012, 2015).

Figure 5.26: Left lower first molar of Iberomys huescarensis from Santa Margarida (Portugal) in oclusal view.

The next two teeth (Figure 5. 27) are one left and one right first molars; being the left one attached to part of its hemimandible. They have similar characteristics to the previous one, but the lingual triangles are better developed, while the labial ones are smaller. The connection of the T4-T5 is narrow, as well as the connection to the anterior cap. The LRA5 is marked and separates a T7 from the anterior cap. The T6 is also visible in the left one. According to López-García et al., (2015) these characteristics are congruent with Iberomys brecciensis (Giebel, 1847).

Figure 5.27: Right and left (respectively) lower first molars of Iberomys brecciensis from Santa Margarida (Portugal) in oclusal view.

138

Finally the last recognizable arvicoline teeth in the sample (Figure 5.28) is a first lower left molar that has lost its posterior lobe. Despite that, it is possible to distinguish on it extremely elongated lingual triangles with acute angles at their tips. All the T1-T5 triangles are well individualized. The T7 is greatly developed, reaching the size of the lingual triangles. The LRA5 is greatly developed and well marked as well. There characters allow it to be referred to Iberomys cabrerae (Thomas, 1906) according to López-García & Cuenca-Bescós, (2012).

Figure 5.28: Left lower first molar (without the posterior lobe) of Iberomys cabrerae from Santa Margarida (Portugal) in oclusal view.

The measurements and ratios for all the identified arvicoline teeth were recorded (Using the program ImageJ) following the indications showed in the figure 5.22. We have them in the next table (table 5.6).

Table 5.6: Measurements and ratios of the arvicoline first lower molars from Santa Margarida

Allophaiomys Iberomys Iberomys Iberomys Iberomys chalinei huescarensis brecciensis 1 brecciensis 2 cabrerae L 2,701 2,958 2,516 2,330 W 1,004 1,041 1,050 0,972 0,955 TTC 1,446 1,427 1,214 1,064 ACC 1,286 1,514 1,349 1,223 1,419 a 1,286 1,514 1,349 1,223 1,419 c 0,260 0,271 0,129 0,056 0,043 La 0,380 0,407 0,384 0,417 0,304 Li 0,493 0,657 0,694 0,584 0,688

A/L 0,476 0,512 0,536 0,525 C/W 0,259 0,260 0,123 0,058 0,045 La/Li 0,771 0,619 0,553 0,714 0,442

A/L x100 47,612 51,183 53,617 52,489 C/Wx100 25,896 26,033 12,286 5,761 4,503 La/Lix100 77,079 61,948 55,331 71,404 44,186

139

These measurements and ratios where compared with those provided in López-García et al., (2012, 2015) for Allophaiomys lavocati Laplana & Cuenca-Bescós, 2000; Victoriamys chalinei, Iberomys huescarensis, Iberomys brecciensis and Iberomys cabrerae for a collection of localities.

Figure 5.29: A/L x100 vs C/W x100 ratios for a collection of arvicoline teeth. Victoriamys chalinei from El Chaparral locality appear as pink circles. Allophaiomys lavocati from El Chaparral appear as blue pluses symbols. Iberomys huescarensis, from a collection of Iberian and Italian sites appear as black dots. Data from López-García et al., (2012, 2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle.

In the figure 5.29 we can see a comparison between the ratios A/L (which measures how much of the total length of the tooth is comprised by the anteroconid complex) and C/W (which measures how much thick is the connection of the T4-T5 compared to its overall width) for the specimens of Santa Margarida and other localities. We can see how the Santa Margarida specimens ascribed to Victoriamys chalinei and Iberomys huescarensis fall into the variability of their respective species meanwhile both Iberomys brecciensis fall further from them with lower values of C/W (narrower connections between the T4 and T5) and greater values of A/L (Longer anteroconid complexes compared to the total length).

Figure 5.30: A/L x100 vs La/Li x100 ratios for a collection of arvicoline teeth. Victoriamys chalinei from El Chaparral locality appear as red inverted triangles. Allophaiomys lavocati from El Chaparral appear as blue squares. Iberomys huescarensis, from a collection of Iberian and Italian sites appear as black dots. Data from López-García et al., (2012, 2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle.

140

In the figure 5.30 we can see a comparison between the ratios A/L (which measures how much of the total length of the tooth is comprised by the anteroconid complex) and La/Li (which measures how long is the T4 compared to the T5) for the specimens of Santa Margarida and other localities. The Victoriamys chalinei of Santa Margarida falls between the variability of the V. chalinei and A. lavocati from El Chaparral. The I. huescarensis and the two I. brecciensis from Santa Margarida fall into the variability of the first. This indicates that the I. brecciensis of Santa Margarida do not have very asymmetrical teeth despite that the length of the anteroconid complex relative to the total length is greater than in most of I. huescarensis.

Figure 5.31: A/L x100 vs La/Li x100 ratios for Iberomys brecciensis (green squares) and I. chalinei (gold inverted triangles) from a collection of Iberian and Italian sites. Data from López-García et al., (2015). The specimens from Santa Margarida (Portugal) appear marked with a red circle.

In the figure 5.31 we can see a comparison between the ratios A/L (which measures how much of the total length of the tooth is comprised by the anteroconid complex) and La/Li (which measures how long is the T4 compared to the T5) for the specimens of Santa Margarida and other localities. Meanwhile one specimen of Iberomys brecciensis fall into the variability of its species the other have extremely similar sizes of the T4 and T5, almost at the level of the Victoriamys chalinei from Santa Margarida and surpassing the more archaic I. huescarensis.

Lagomorpha Brandt, 1855

One of the few relatively large sized bones recovered with an air-scribe from one of the blocks is a distal left humerus (Figure 5.32). It has a narrow trochlea that is inclined towards the medial line of the body; the coronoid fossa and the olecranon fossa are connected by a distinctive supratrochleal foramen; the anterior half of the diaphysis is mostly straight in lateral view meanwhile the posterior part is slightly curved caudally, making the proximal part of the diaphysis wider in lateral view. Compared with actual remains of rabbit (Oryctolagus cuniculus) it is easy to ascribe it to this species, given the mentioned characteristics. This is the animal species of largest size identified at this moment in Santa Margarida.

141

Figure 5.32: Left distal Oryctolagus cuniculus humerus from Santa Margarida (Portugal) in A) Anterior; B) Exterior; C) Posterior and D) Medial views.

Diverse postcranial elements

A wide array of postcranial bones (Figure 5.33) has been recovered in the sample, likely in the range of hundreds, but the great majority of them are too fragmentary to be identified to the species, or even family level.

Figure 5.33: Diverse postcranial elements of microvertebrates recovered from the sample of Santa Margarida. A) Distal femur in posterolateral view, B) Partial distal humerus? in anterior view and C) Distal radius in lateral view.

142

5.5 Discussion

The age

Starting with the gastropods, Pomatias elegans has been cited in Gelasian-Calabrian (2,5 – 0,781 million YBP) deposits from Italy (Bianchi et al., 2013) and it still lives today in Portugal (Platts et al., 2003), therefore it is not useful for precise dating of the site.

Much more interesting is the appearance of Paralaoma servilis. Its presence in New Zealand and in Australia dates from the Pleistocene and Holocene, but elsewhere this land snail is regarded as an invasive species that has arrived in modern times, although it should have departed from New Zealand soon after western colonization (Christensen et al., 2012). In spite of this, there is a record of this species coming from soil samples from a remote location of the island of Gávdos (Greece) that is regarded as unlikely to have been produced by a recent introduction (Welter-Schultes, 1998), therefore opening the possibility that this species is not an alien in South Europe. In this context the new cite of Santa Margarida can have two explanations: Either it belongs to a recent contamination (some of the blocks could have caveats that could house modern snails) or it proves the presence of this species as native from Southern Europe. Only further work and new shells (especially in situ in blocks) could prove this point.

The indeterminate lacertid remains are not of value for age determination, since this group has been present in Europe for dozens of millions of years until today (Bisbal-Chinesta & Blain, 2018).

Talking about bat remains, they are not age-informative; because the Myotis genus has been living in Europe at least since the beginning of the Pleistocene up to today (Gunnell et al., 2011). Something similar can be said about the shrews: The Crocidura shrews have been living in Europe for at least 5 million of years (Rofes & Cuenca-Bescós, 2011) and Sorex probably more (Reumer, 1984). Without a specific determination and it is not possible to give a more precise age.

Talking about the rodents, Eliomys quercinus , that today lives in the Iberian Peninsula, appears on it during the Early Pleistocene and overlaps with Eliomys intermedius Friant, 1953 during that time period (Piñero et al.; 2015, 2017). Allocricetus bursae is present in Iberia since the late Early Pleistocene until its extinction around 20.000 YBP, being specially common during the middle Pleistocene (López-García, 2008; Sesé et al., 2011). Apodemus sylvaticus is a species that appears in Iberia around the beginning of the Pleistocene (Sesé et al., 2011) but it becomes abundant only during the Late Pleistocene (López-García, 2008). The arvicolines are the most interesting taxa for age determination. Victoriamys chalinei is a typical species from the late Early Pleistocene of the Iberian Peninsula, being it restricted to the Arvicoline Zone 5 according to Martin, (2012) that ranges approximately between 1 million and 800.000 YBP. Iberomys huescarensis ranges broadly with the same time period (López-García et al., 2015). Iberomys brecciensis is a typical species from the Middle Pleistocene that spans between 800.000 and 120.000 YBP (López-García et al., 2015), meanwhile Iberomys cabrerae only appears after around 120 YBP and broadly correspond to the Late Pleistocene until today.

The three chronospecies of Iberomys (Figure 5.34) indicates that in Santa Margarida there are fossils separated between them for at least 680.000 years; which separates the last occurrence of I. huescarensis and A. chalinei from the first occurrence of I. cabrerae. Given this absence of overlapping between those species and the presence of I. brecciensis (That could be present either at the beginning or the end of the middle Pleistocene) a model with at least two separated time periods during the Pleistocene is proposed:

1) A first moment around 800.000 YBP that includes broadly the transition between the Early and Middle Pleistocene (End of the Calabrian). It is characterized by the presence of I.

143

huescarensis and V. chalinei and archaic forms of I. brecciensis. The A. bursae tooth could come from this phase too. 2) A second moment around 120.000 YBP that includes the end of Middle Pleistocene and the beginning of the Late Pleistocene. It is characterized by the appearance of late forms of I. brecciensis and the presence of I. cabrerae. The relatively abundant Apodemus teeth might come from this phase too.

The first chronology is similar to that of the oldest Quaternary sites known in Portugal: Morgadinho (Antunes et al., 1986a) and Algoz (Antunes et al., 1986b). In none of those sites I. huescarensis and A. chalinei have been cited, making Santa Margarida their first known occurrence in Portugal. The second chronology is similar to the site of Mealhada (Zbyszewski, 1977a) and slightly younger than some sites in the Almonda karstic system (Marks et al., 2002a; López-García et al., 2018).

Figure 5.34: Evolution of the first lower molar of the genus Iberomys with examples from the site of Santa Margarida (Portugal).

Stratigraphy and type of the site

All the works that have been carried out (see “methods” section) assumed that given the way of construction of a traditional algarvian wall (Gonçalves et al., 2018) and the similarity of the hand samples, all the blocks of cemented terra rossa belonged to the most superficial level of the site and therefore they were treated equally. As we have seen in the previous section the amount of time represented in the fossils contained in those superficial blocks leave us with two possible scenarios:

1) The fossils were mixed during the preparation. The blocks collected belonged to different layers (Figure 5.35 upper). 2) The fossils of the oldest chronology are reworked. The recovered fossils actually come from the same layer. (Figure 5.35 lower).

144

Figure 5.35: The two possible stratigraphy scenarios that explain the faunal assemblage recovered from the ex situ blocks from Santa Margarida (Portugal). Upper: Fossils from different but similar layers that became mixed during preparation. Lower: Reworked fossils situated in the same layer as the younger ones.

This second scenario is unlikely, given the fragility of microvertebrate fossils, but not impossible. It is not possible to test it with the fragmentation or erosion of the pieces, given the aggressive treatment with water softener that all specimens have been subject to.

About the nature of the deposit; given the calcareous crusts that cover the terra rossa (that could be interpreted as speleothems of flowstone type); the presence of bat remains and its geographical location in the slope of a hill, the site is interpreted as a cave deposit similar to the one known as “Cueva Des-Cubierta” in Pinilla del Valle, Madrid (Baquedano et al., 2016). The cave itself has been eroded away, but the cemented and breccified infilling is preserved, with the fossils that it contained. This could explain why the blocks were in the surface and there is not a trace of a cave nearby.

Finally the richness of microfossil vertebrates found in the locality has to be remarked. In the middle Pleistocene site of Gruta da Aroeira (Almonda System), 20 kg of sediment yielded 362 specimens (López-García et al., 2018); meanwhile in Santa Margarida ~0,5 kg yielded ~300 fossils, although they were much more fragmentary.

145

5.6 Conclusions

- Santa Margarida site is a new open-air microvertebrate locality from Algarve, municipality of Loulé. - It is known thanks to ex situ blocks of cemented terra rossa with fossils that were used in the construction of a traditional algarvian wall. Due to the hard cemented sediment of the blocks the preparation was difficult, but successful. - About 300 dental and mandibular remains of vertebrates and gastropod shells were collected from about 0,5 kg of disaggregated sediments, plus hundreds of bone fragments. - The number of dental remains recovered makes Santa Margarida a remarkably rich site for microvertebrates. - At least two invertebrate taxa (Pomatias elegans and Paralaoma servilis) and 12 vertebrates (Lacertidae indet., Chiroptera indet., Crocidura sp., Sorex sp., Eliomys quercinus, Allocricetus bursae, Apodemus cf. sylvaticus, Victoriamys chalinei, Iberomys huescarensis, Iberomys brecciensis, Iberomys cabrerae and Oryctolagus cuniculus) have been recognized. - This is the first reported occurrence of I. huescarensis and V. chalinei in Portugal. - There are least two time periods of the Pleistocene represented in Santa Margarida; the first one about 800.000 YBP and the second around 120.000 YBP. - Santa Margarida one of the three oldest Quaternary sites known from Portugal, together with Algoz and Morgadinho, also in Algarve. - The characteristics of the site are congruent with an eroded cave.

146

6. CONCLUSIONS

Here is presented the compilation of conclusions replicated from previous chapters.

Chapter 2: Paleobiodiversity of the Quaternary fossil tetrapods in continental Portugal

- Most of the vertebrate fossil record of the Quaternary of Portugal (with emphasis in the Pleistocene) is situated in the Lusitanian and Algarvian basins. Only 3 out of 30 sites are outside those areas and 20 out of 30 sites are situated in caves. - Quaternary fossil herpetofaunal communities in Portugal evolve across the period with reduced numbers and only one extinct species in the area. - 70% of living bird species are undetected in the fossil record due to their fragile bones, the difficulties in their identification and the lack of scientific interest. - 46% of mammal species known in the Quaternary fossil record of Portugal became locally extinct in the area or went totally extinct. They are also the better known group of Quaternary fossil tetrapods in the country, with most of the living species (60%) appearing in the fossil record. - The importance of the cave deposits is capital in our understanding of the Quaternary vertebrate fauna, without it we would not known about most of tetrapod species, not to mention facts about their paleobiology or paleoecology (Like their denning behavior). The tetrapod living species detected in the fossil record would fall from 37,65% (125 species) to only a 10,54% (35 species) without them.

Chapter 3: The proboscidean of Santo Antão do Tojal

- The remains of the Santo Antão do Tojal proboscidean (Palaeoloxodon antiquus), which is around 80.000 YBP old; comprise a first phalanx, a neural arch and a right femur and tibia. - The Santo Antão do Tojal specimen is estimated to have measured alive about 3,8 meters in the shoulder and weighted nearly 11 tons, being it in the average for a male Palaeoloxodon antiquus. Despite being one of the most recent and Southwestern occurrences of this species, it is nearly identical to specimens that lived several thousand years before in Central Europe or the Italian Peninsula. - The size difference between Mammuthus primigenius and Palaeoloxodon antiquus is 18% on average for the femur and 24% in the tibia. The best measurements for differentiating between both species are in the femur maximum length, caput width and distal width; and in the tibia maximum length, minimum diaphysis circumference, minimum diaphysis width and distal width. - We agree with previous works (Zbyszewski, 1943; Antunes & Cardoso, 1992) in that the Santo Antão do Tojal proboscidean is a Palaeoloxodon antiquus in base of the caput width of the femur and the minimum diaphysis circumference and width of the tibia, combined with a deep trochanteric fossa in the femur and the overall size of the bone and its robustness. - Palaeoloxodon antiquus measurements follow a clearer bimodal distribution than Mammuthus primigenius ones, indicating a more pronounced sexual dimorphism. - The hind limb anatomy of Palaeoloxodon antiquus reflects more robust bones than in Mammuthus primigenius, even for similar sized animals, as other works have shown (Larramendi et al., 2017).

147

Chapter 4: Algar do Vale Da Pena: A new fossil bear site from Portugal

Chapter 5: Santa Margarida: A new microvertebrate locality from Algarve.

148

8. REFERENCES

Adams, A. L. (1881). Monograph on the British fossil elephants. London: Printed for the

Palæontographical Society. Retrieved from https://catalog.hathitrust.org/Record/001669226

Alcalá, L., Mazo, A. V., & Morales, J. (1985). Mamíferos de las cuencas del Júcar y Cabriel (Mapa

geológico de España 1:50000 No. Hoja no 790). Madrid.

Alcover, J. A., Pieper, H., Pereira, F., & Rando, J. C. (2015). Five new extinct species of rails (Aves:

Gruiformes: Rallidae) from the Macaronesian Islands (North Atlantic Ocean). Zootaxa,

4057(2), 151–190. https://doi.org/10.11646/zootaxa.4057.2.1

Alexandrino, J., Froufe, E., Arntzen, J. W., & Ferrand, N. (2001). Genetic subdivision, glacial refugia

and postglacial recolonization in the golden-striped salamander, Chioglossa lusitanica

(Amphibia: Urodela). Molecular Ecology, 9(6), 771–781. https://doi.org/10.1046/j.1365-

294x.2000.00931.x

Álvarez-Lao, D. J., Kahlke, R.-D., García, N., & Mol, D. (2009). The Padul mammoth finds — On the

southernmost record of Mammuthus primigenius in Europe and its southern spread during the

Late Pleistocene. Palaeogeography, Palaeoclimatology, Palaeoecology, 278(1), 57–70.

https://doi.org/10.1016/j.palaeo.2009.04.011

Amendoeira, V., Lopes, P., Rodrigues, N., Rodrigues, P., Robalo, P., & Lopes, S. (2015). Cartografía

Algar do Vale da Pena (Algar dos Ursos). GEM (Grupo de espeleología e montanhismo).

Andrews, C. W., & Forster-Cooper, C. (1928). On a specimen of Elephas antiquus from Upnor.

London: British Museum(Natural History.) Department of Geology. Retrieved from

http://archive.org/details/onspecimenofelep1928andr

Antunes, M. T. (1986). Acerca de um osso do Plistocénico da Mealhada: presença de um «tigre dente

de sabre», Homotherium latidens (OWEN, 1846). Ciências da terra, 8, 43–54.

Antunes, M. T. (1989a). Castor fiber na gruta do Caldeirão: Existência, distribuição e extinção do

castor em Portugal. Ciências da terra, 10, 23–40.

Antunes, M. T. (1989b). The Proboscideans Data, Age and Paleogeography: Evidence from the

Miocene of Lisbon. In E. H. Lindsay, V. Fahlbusch, & P. Mein (Eds.), European Neogene

149

Mammal Chronology (pp. 253–262). Boston, MA: Springer US. https://doi.org/10.1007/978-

1-4899-2513-8_17

Antunes, M. T. (2000). Gruta da Figueira Brava: Pleistocene marine mammals. Memorias Da

Academia Das Ciencias de Lisboa, Classe de Ciencias, 38, 245–558.

Antunes, M. T. (2006). The Zebro (Equidae) and its extinction in Portugal, with an appendix on the

noun zebro and the modern zebra. Equids in Time and Space, 210–235.

Antunes, M. T., Azzaroli, A., Faure, M., Guérin, C., & Mein, P. (1986c). Mammifères pleistocènes de

Algoz, en Algarve: une revision. Ciências da terra, 8, 73–86.

Antunes, M. T., & Cardoso, J. L. (1986). O leão das cavernas, Pantera (Leo) Spelaea (Goldfuss, 1810)

em Portugal. Da Pré-história à história : homenagem a Octávio da Veiga Ferreira, 73–80.

Antunes, M. T., & Cardoso, J. L. (1992). Quaternary elephants in Portugal : new data. Ciências Da

Terra, 11, 17–37.

Antunes, M. T., Cardoso, J. L., & Faure, M. (1988). Présence de Hippopotamus incognitus au Portugal

et remarques sur les sites quaternaires de Mealhada. Comunicações Geológicas, 74, 165–172.

Antunes, M. T., Crespo, E., Mein, P., Pais, J., & Teixeira, J. P. (1989). Guia (Algarve), gisement de

vertébrés quaternaires à caractère saisonnier. Ciências da Terra, 10(97–106). Retrieved from

https://run.unl.pt/handle/10362/4427

Antunes, M. T., Manuppella, G., Mein, P., & Zbyszewski, G. (1986b). Goldra: premier gisement

karstique en Algarve, faune et industries. Ciências da terra, 8, 31–42.

Antunes, M. T., Mein, P., Nascimento, A., & Pais, J. (1986a). Le gisement pleistocène de Morgadinho,

en Algarve. Ciências da Terra, 8, 9–22.

Assírio, & Alvim. (2008). Atlas das aves nidificantes em Portugal. Lisboa: Instituto da Conservação

da Natureza e da Biodiversidade,. Retrieved from

http://www2.icnf.pt/portal/pn/biodiversidade/patrinatur/especies/aves/atlas-aves-nid-pt

Athanassiou, A. (2001). New findings of fossil large mammal remains in the Penios valley (Area of

Larissa, Thessaly, Greece). Bulletin of the Geological Society of Greece, 34(2), 533–539.

https://doi.org/10.12681/bgsg.17087

150

Baca, M., Popović, D., Panagiotopoulou, H., Marciszak, A., Krajcarz, M., Krajcarz, M. T., …

Nadachowski, A. (2018). Human-mediated dispersal of cats in the Neolithic Central Europe.

Heredity, 121(6), 557. https://doi.org/10.1038/s41437-018-0071-4

Baquedano, E., Arsuaga, J. L., Pérez-González, A., Márquez, B., Laplana, C., Ortega, M. ., …

Álvarez-Lao, D. (2016). The Des-Cubierta Cave (Pinilla del Valle, Comunidad de Madrid,

Spain): a Neanderthal site with a likely funerary/ritualistic connection. In Proceedings of the

European Society for the study of Human Evolution 5 (p. 41). Alcalá de Henares (Madrid):

European Society for the Study of Human Evolution.

Beauval, C., Michel, P., & Tastet, J. (1998). L’éléphant antique de Soulac (Gironde, France)

[Biometrical and morphological analysis of the Palaeoloxodon antiquus from Soulac (Gironde,

France)]. Quaternaire, 9(2), 91–100. https://doi.org/10.3406/quate.1998.1592

Bencatel, J., Álvares, F., Moura, A. E., & Barbosa, A. M. (2017). Atlas de Mamíferos de Portugal.

Universidade de Évora. Retrieved from http://atlas-mamiferos.uevora.pt/

Bianchi, V., Ghinassi, M., Aldinucci, M., Boscaini, N., Martini, I., Moscon, G., & Roner, M. (2013).

Geological map of Pliocene-Pleistocene deposits of the Ambra and Ombrone valleys

(Northern Siena Basin, Tuscany, Italy). Journal of Maps, 9(4), 573–583.

https://doi.org/10.1080/17445647.2013.829412

Bicho, N., Belcher, W., Haws, J. A., Markova, A., & Hockett, B. S. (2003). Paleoecologia e ocupação

humana da Lapa do Picareiro: resultados preliminares. Revista portuguesa de arqueologia,

6(2), 49–82.

Bicho, N., & Haws, J. (2008). At the land’s end: Marine resources and the importance of fluctuations

in the coastline in the prehistoric hunter–gatherer economy of Portugal. Quaternary Science

Reviews, 27(23), 2166–2175. https://doi.org/10.1016/j.quascirev.2008.08.011

Bicho, N., Hockett, B., Haws, J., & Belcher, W. (2000). Hunter–gatherer subsistence at the end of the

Pleistocene: preliminary results from Picareiro Cave, Central Portugal. Antiquity, 74(285),

500–506. https://doi.org/10.1017/S0003598X00059834

151

Bisbal-Chinesta, J. F., & Blain, H.-A. (2018). Long-term changes in composition and distribution

patterns in the Iberian herpetofaunal communities since the latest Pleistocene. Quaternary

Science Reviews, 184, 143–166. https://doi.org/10.1016/j.quascirev.2017.06.010

Blain, H.-A., Bailon, S., & Agustí, J. (2007). Anurans and squamate reptiles from the latest early

Pleistocene of Almenara-Casablanca-3 (Castellón, East of Spain). Systematic, climatic and

environmental considerations. Geodiversitas, 29(2), 269–295.

Blain, H.-A., Bailon, S., & Agustí, J. (2016). The geographical and chronological pattern of

herpetofaunal Pleistocene extinctions on the Iberian Peninsula. Comptes Rendus Palevol,

15(6), 731–744. https://doi.org/10.1016/j.crpv.2015.05.008

Boscaini, A., Alba, D. M., Beltrán, J. F., Moyà-Solà, S., & Madurell-Malapeira, J. (2016). Latest Early

Pleistocene remains of Lynx pardinus (Carnivora, Felidae) from the Iberian Peninsula:

Taxonomy and evolutionary implications. Quaternary Science Reviews, 143, 96–106.

https://doi.org/10.1016/j.quascirev.2016.05.015

Brugal, J.-P., Argant, J., Crispim, J. A., Figueiredo, S., Serra, A. M., & Palmqvist, P. (2012). The

Complex Carnivore-rich Assemblages from Furninha (Peniche, Portugal): a Multidisciplinary

Approach. Journal of Taphonomy, 10(3–4), 22.

Busack, S. D., Salvador, A., & Lawson, R. (2006). Two new species in the genus psammodromus

(reptilia: lacertidae) from the iberian peninsula. Annals of Carnegie Museum, 75(1), 1–10.

https://doi.org/10.2992/0097-4463(2006)75[1:TNSITG]2.0.CO;2

Cardoso, J. L. (1989). Le daim dans le Pléistocène du Portugal. Comunicações Geológicas, 75, 111–

118.

Cardoso, J. L. (1990). Presença de Rinoceronte: dicerorhinus hemitoechus (FALCONER, 1878) na

gruta do Escoural. Almansor, (8), 7–14.

Cardoso, J. L. (1992). Présence de Cuon alpinus europaeus BOURGUIGNAT, 1868 (Mammalia,

Carnivora) dans le Pléistocène du Portugal. Ciências da Terra, 11, 65–75.

Cardoso, J. L. (1993a). Contribuição para o conhecimento dos Grandes Mamíferos do Plistocénico

superior de Portugal. Oeiras: Cámara municipal de Oeiras.

152

Cardoso, J. L. (1993b). O estudo dos grandes mamíferos Plistocénicos de Portugal. Síntese Histórica.

In O Quaternário em Portugal: Balanço e perspectivas (pp. 97–103). Lisboa: Colibri.

Cardoso, J. L. (1994). Crocuta crocuta intermedia (M. de Serres, 1828) (Mammalia, Carnivora) no

Plistocénico de Portugal. Comunicações Geológicas, 80, 89–97.

Cardoso, J. L. (1995). Presença de Equus hydruntinus REGALIA, 1905 no Würm recente de Portugal.

Comunicações Geológicas, 81, 97–108.

Cardoso, J. L. (1996). Les grands Mammifères du Pléistocène supérieur du Portugal. Essai de

synthèse. Geobios, 29(2), 235–250. https://doi.org/10.1016/S0016-6995(96)80047-7

Cardoso, J. L. (2000). As investigações de Carlos Ribeiro e de Nery Delgado sobre o “Homem

Terciário”: Resultados e consequências na época e para além dela. Estudos Arqueológicos de

Oeiras, 8, 33–54.

Cardoso, J. L. (2002). Arqueofaunas: balanço da sua investigação em Portugal. Revista Da Associação

Dos Arqueólogos Portugueses, 54, 281–298.

Cardoso, J. L. (2008a). Joaquim Felipe Nery Delgado, arqueólogo. In Nery Delgado (1835-1908):

geólogo do reino (pp. 65–81). Lisboa: Instituto Nacional de Engenharia, Tecnologia e

Inovação : Centro de História e Filosofia da Ciência. Retrieved from

http://www.lneg.pt/servicos/103/1182/

Cardoso, J. L. (2008b). O. da Veiga Ferreira (1917-1997): sua vida e obra científica. Estudos

Arqueológicos de Oeiras, 16, 13–123.

Cardoso, J. L. (2013). Carlos Ribeiro e Oeiras. Razões de uma homenagem. Estudos Arqueológicos de

Oeiras., 20, 11–20.

Cardoso, J. L. (2015a). A investigação da antiguidade do Homem no Portugal de oitocentos : um

contributo para a história da ciência. Estudos Arqueológicos de Oeiras, 22, 9–42.

Cardoso, J. L. (2015b). Carlos Ribeiro (1813-1882), as formações quaternárias portuguesas e a

antiguidade do homem : um manuscrito desconhecido. Estudos Arqueológicos de Oeiras, 22,

43–92.

Cardoso, J. L., & Antunes, M. T. (1989). Rupicapra rupicapra (Mammalia) In the Late Pleistocene of

Portugal. Ciências Da Terra, 10, 81–96.

153

Cardoso, J. L., & Avila de Melo, A. (2001). Correspondência anotada de Carlos Ribeiro e de Nery

Delgado: contribuição para a história da Arqueologia em Portugal. Comunicações Geológicas,

88, 309–246.

Cardoso, J. L., & Eisenmann, V. (1989). Equus Caballus Antunesi, nouvelle sous-espece quaternaire

du Portugal. Palaeovertebrata, 19, 48–72.

Cardoso, J. L., & Regala, F. T. (2002). Sobre a presença de Mamute, Mammuthus Primigenius

(Blumembach, 1799) em Portugal: descoberta de uma lamela dentária em depósitos

Plistocénicos do fundo do estuário do Tejo (Cruz Quebrada, Oeiras). Estudos Arqueológicos

de Oeiras, 10, 39–47.

Cardoso, J. L., & Regala, F. T. (2006). O Leopardo, Panthera pardus (L., 1758), do Algar da Manga

Larga (Planalto de Santo António, Porto de Mós). Comunicações Geológicas, 93, 119–144.

Cardoso, J. L., Vilstrup, J. T., Eisenmann, V., & Orlando, L. (2013). First evidence of Equus asinus L.

in the Chalcolithic disputes the Phoenicians as the first to introduce donkeys into the Iberian

Peninsula. Journal of Archaeological Science, 40(12), 4483–4490.

https://doi.org/10.1016/j.jas.2013.07.010

Carneiro, A. (2007). Sharing common ground: Nery Delgado (1835–1908) in Spain in 1878.

Geological Society, London, Special Publications, 287(1), 119–134.

https://doi.org/10.1144/SP287.10

Carvalho, J. M. F. (1996). Área de Moleanos (Maciço Calcário Estremenho) : litologia temática e

sondagens (report). Retrieved from http://repositorio.lneg.pt/handle/10400.9/1881?locale=en

Choffat, P. (1895). Note sur les tufs de Condeixa et la découverte de l’hippopotame en Portugal.

Comunicações Geológicas, 3, 1–12.

Christensen, C., Yeung, N., & Hayes, K. (2012). First Records of Paralaoma servilis (Shuttleworth,

1852) (Gastropoda: Pulmonata: Punctidae) in the Hawaiian Islands. Bishop Museum

Occasional Papers: Records of the Hawaii Biological Survey for 2011, 112, 3–7.

Christiansen, P. (2004). Body size in proboscideans, with notes on elephant metabolism. Zoological

Journal of the Linnean Society, 140(4), 523–549. https://doi.org/10.1111/j.1096-

3642.2004.00113.x

154

Christiansen, P. (2007). Long-bone geometry in columnar-limbed animals: allometry of the

proboscidean appendicular skeleton. Zoological Journal of the Linnean Society, 149(3), 423–

436. https://doi.org/10.1111/j.1096-3642.2007.00249.x

Crespo, E. G. (2002). Paleoherpetofauna portuguesa. Revista Española de Herpetología, 16, 17–35.

Crespo, E. G., Patricio, G. A., & Antunes, M. T. (2000). Pleistocene Reptilia and Amphibia from

Gruta da Figueira Brava (Arrábida, Portugal). Memorias Da Academia Das Ciencias de

Lisboa, Classe de Ciencias, 38, 23–67.

Crispim, J. A. (2008). Património geológico da Serra de Montejunto (1st ed.). Lisboa: Sociedade

Portuguesa de Espeleologia.

Cuenca-Bescós, G. (2003). Allocricetus (Cricetidae, Rodentia, Mammalia) from the Pleistocene. A

phylogenetical approach. Coloquios de Paleontolgía, 1, 95–113.

Cunha, P. P., Martins, A. A., Buylaert, J.-P., Murray, A. S., Gouveia, M. P., Font, E., … Mota, R.

(2019). The Lowermost Tejo River Terrace at Foz do Enxarrique, Portugal: A

Palaeoenvironmental Archive from c. 60–35 ka and Its Implications for the Last Neanderthals

in Westernmost Iberia. Quaternary, 2(1), 3. https://doi.org/10.3390/quat2010003

Dantas da Gama, M. (1998). O galo-montês Tetrao urogallus (Linn) – regresso ao Gerês. Porto:

Canhões de Pedra.

Daveau, S. (2000). Fases históricas e perspectivas actuais do estudo do Quaternário em Portugal.

Estudos do Quaternário / Quaternary Studies, 3, 5–14.

Davis, S. (2002). The mammals and birds from the Gruta do Caldeirão, Portugal. Revista portuguesa

de arqueologia, 5(2), 29–98.

Davis, S., & MacKinnon, M. (2009). Did the Romans bring fallow deer to Portugal? Environmental

Archaeology, 14(1), 15–26. https://doi.org/10.1179/174963109X400646

Davis, S., & Simões, T. (2016). The velocity of ovis in prehistoric times: the sheep bones from Early

Neolithic Lameiras, Sintra, Portugal. In O Neolítico em Portugal antes do Horizonte 2020:

Perspectivas em debate (Vol. 2, pp. 51–66). Lisboa: Associação dos Arqueólogos

Portugueses. Retrieved from http://repositorio.ul.pt/handle/10451/25678

155

Dean, R. M., & Carvalho, A. F. (2011). Surf and Turf: The Use of Marine and Terrestrial Resources in

the Early Neolithic of Coastal Southern Portugal. In Trekking the Shore (pp. 291–302).

Springer, New York, NY. https://doi.org/10.1007/978-1-4419-8219-3_12

Delgado, J. F. N. (1867). Da existencia do homem no nosso solo em tempos mui remotos provada pelo

estudo das cavernas: Primeiro opusculo. Noticia ácerca das grutas da Cesareda, com a

versão em francez. Typ. da Academia Real das Sciencias.

Delgado, J. F. N. (1884). La grotte de Furninha a Peniche. In Compte-Rend de la 9ème Session

(Lisboa, 1880) (pp. 207–278). Lisboa.

Delgado, J. F. N. (1886). Estudo sobre os Bilobites e outros fósseis das quartzites da base do systema

Silurico de Portugal. Typ. da Academia real das sciencias.

Demay, L. (2014). Analyses archéozoologiques de l’assemblage d’ossements de mammouths

(Mammuthus primigenius) de Changis-sur-Marne : ostéométrie, âge, sexe (Île de France,

Seine-et-Marne). Internal reportl Institut de Paléontologie humaine, Fondation Albert Ier de

Monaco.

Detry, C. (2008). Vertebrates from Cabeço dos Morros: a Mesolithic shell midden near Salvaterra de

Magos, in the lower Tagus valley, Portugal. Promontoria, Revista Do Departamento de

História, Arqueologia e Património Da Universidade Do Algarve, 6, 51–71.

Ely, C. R., Fox, A. D., Alisauskas, R. T., Andreev, A., Bromley, R. G., Degtyarev, A. G., … Walsh,

A. (2005). Circumpolar variation in morphological characteristics of Greater White-fronted

Geese Anser albifrons. Bird Study, 52(2), 104–119.

https://doi.org/10.1080/00063650509461380

Estraviz-López, D., & Mateus, O. (2018a). Paleobiodiversity of Quaternary fossil tetrapods in

continental Portugal. In Abstract book of the 1st Paleontological Virtual Congress (p. 86).

Internet.

Estraviz-López, D., & Mateus, O. (2018b). The History of the Quaternary vertebrate paleontology in

Portugal. In Abstract book of the XVI Annual Meeting of the European Association of

Vertebrate Palaeontology (p. 65). Caparica, Portugal: Marzola, M., Mateus, O., Moreno-

Azanza, M. (eds.).

156

Estraviz-López, D., & Mateus, O. (2018c). The Quaternary mammals from the caves of Casais

Robustos and Cabeço de Morto, Alcanena (Portugal): The case of Lynx pardinus Temminck,

1827. In Life finds a way (pp. 159–162). Zarautz.

Estraviz-López, D., & Mateus, O. (2018d). The size and body mass estimations of the proboscidean

Paleoloxodon antiquus (Falconer & Cautley, 1847) from the Quaternary of Santo Antão do

Tojal, Portugal. In VIII Congresso de Jovens Investigadores em Geociências (pp. 70–73).

Estremoz, Portugal.

Figueiredo, S. M. D. (2010). A Avifauna Plistocénica de Portugal: especificidades evolutivas,

anatómicas e o seu contexto paleontológico, geológico e arqueológico (PhD thesis).

Universidad de Salamanca. Retrieved from https://gredos.usal.es/jspui/handle/10366/76422

Figueiredo, S. M. D. (2012). The pleistocene elephants of Portugal. In Elfenbeinstudien. 1, Marfil y

elefantes en la Península Ibérica y el Mediterráneo occidental: Actas del coloquio

internacional en Alicante el 26 y 27 de noviembre 2008, 2012, ISBN 978-3-8053-4424-1,

págs. 37-44 (pp. 37–44). Alicante: Deutsches Archäologisches Institut, Museo Arqueológico

de Alicante - MARQ. Retrieved from

https://dialnet.unirioja.es/servlet/articulo?codigo=5594548

Fosse, P., Besson, J.-P., Laborde, H., Thomas-Cantié, F., Cazenave, G., Delmasure, M.-C., … Quiles,

J. (2004). Denning behaviour of «modern» brown bear (Ursus arctos, L.) in caves : biological

and paleontological considerations from French Pyrenean sites. Publications Du Musée Des

Confluences, 2(1), 171–182.

García Vázquez, A. (2015). Caracterización del oso pardo (Ursus arctos L.) fósil en el NW de la

Península Ibérica: datos morfométricos y moleculares (PhD thesis). Universidad de A

Coruña. Retrieved from http://ruc.udc.es/dspace/handle/2183/15810

Georgalis, G. L., Villa, A., & Delfino, M. (2017). The last European varanid: demise and extinction of

monitor lizards (Squamata, Varanidae) from Europe. Journal of Vertebrate Paleontology,

37(2), e1301946. https://doi.org/10.1080/02724634.2017.1301946

Gonçalves, M. M., Prates, G., & Rosendahl, S. (2018). Renewing Terraces and Drystone Walls of

Algarvian Barrocal: Cultural and Touristic Values. In A. Mortal, J. Aníbal, J. Monteiro, C.

157

Sequeira, J. Semião, M. Moreira da Silva, & M. Oliveira (Eds.), INCREaSE (pp. 13–31).

Springer International Publishing.

González, F. L., & Martelli, A. (2003). Paleontology and taphonomy of Pleistocene macromammals of

Galicia (NW Iberian Peninsula) (Vol. 22). O Castro. Laboratorio Xeolóxico de Laxe. Area de

Xeoloxía e Minería.

Grandal-d’Anglade, A. (1993). El oso de las cavernas en Galicia: El yacimiento de Cova Eirós (Vol.

8). O Castro. Laboratorio Xeolóxico de Laxe. Area de Xeoloxía e Minería.

Grandal-d’Anglade, A., Vidal-Romani, J. R., Vaqueiro-Gonzalez, M., & Mendez, E. (2006). O

Rebolal: a new cave in Galicia (NW Spain) with rencent cave bear remains: preliminary

report. Scientific Annals, School of Geology Aristotle University of Thessaloniki (AUTH), 98,

173–178.

Gunnell, G. ., Eiting, T. ., & Geraads, D. (2011). New Late Pliocene bats (Chiroptera) from Ahl al

Oughlam, Morocco. Neues Jahrbuch Für Geologie Und Paläontologie - Abhandlungen,

260(1), 55–71.

Harington, C. R., Mol, D., & van der Plicht, J. (2012). The Muirkirk Mammoth: A Late Pleistocene

woolly mammoth (Mammuthus primigenius) skeleton from southern Ontario, Canada.

Quaternary International, 255, 106–113. https://doi.org/10.1016/j.quaint.2011.05.038

Harlé, E. (1910). Les mammifères et oiseaux quaternaires connus jusqu’ici en Portugal: mémoire suivi

d’une liste générale de ceux de la Péninsule Ibérique. Comunicações Geológicas, 8, 22–85.

Harrison, C. J. O., & Cowles, G. S. (1977). The extinct large cranes of the North-West palaearctic.

Journal of Archaeological Science, 4(1), 25–27. https://doi.org/10.1016/0305-4403(77)90110-

Hillson, S. (2005). Teeth (2nd ed.). Cambridge, UK: Cambridge University Press. Retrieved from

http://discovery.ucl.ac.uk/11267/

Hockett, B., & Haws, J. (2009). Continuity in animal resource diversity in the Late Pleistocene human

diet of Central Portugal. Before Farming, 2009(2), 1–14.

https://doi.org/10.3828/bfarm.2009.2.2

158

Hockett, B. S., & Bicho, N. F. (2000). The Rabbits of Picareiro Cave: Small Mammal Hunting During

the Late Upper Palaeolithic in the Portuguese Estremadura. Journal of Archaeological

Science, 27(8), 715–723. https://doi.org/10.1006/jasc.1999.0496

Jakubowski, G. (1988). Stanowisko s\lonia leśnego–Palaeoloxodon antiquus (Falconer & Cautley,

1847) w górnym plejstocenie odkrzwki Jóźwin kopalni węgla brunatnego “Konin.” Zeszyty

Muzealne, Muzeum Okręgowe w Koninie, 2, 14–87.

Jakubowski, G. (1996). Forest elephant Palaeoloxodon antiquus (Falconer & Cautley, 1847) from

Poland. Prace Muzeum Ziemi, 43, 85–109.

Jeannet, M. (2000). Gruta da Figueira Brava: les rongeurs. Memorias Da Academia Das Ciencias de

Lisboa, Classe de Ciencias, 38, 179–243.

Jiménez Fuentes, E., Cardoso, J. L., & Crespo, E. G. (1998). Presencia de Agrionemys (=Testudo)

hermanni (Gmelin, 1789) en el Paleolítico Medio de la Gruta Nova da Columbeira

(Bombarral, provincia de Estremadura, Portugal). Studia geologica salmanticensia, 34, 123–

139.

Jukar, A. M., Lyons, S. K., & Uhen, M. D. (2018). A cranial correlate of body mass in proboscideans.

Zoological Journal of the Linnean Society, 184(3), 919–931.

https://doi.org/10.1093/zoolinnean/zlx108

Kevrekidis, C., & Mol, D. (2016). A new partial skeleton of Elephas (Palaeoloxodon) antiquus

Falconer and Cautley, 1847 (Proboscidea, Elephantidae) from Amyntaio, Macedonia, Greece.

Quaternary International, 406, 35–56. https://doi.org/10.1016/j.quaint.2015.11.110

Kirillova, I. V., Shidlovskiy, F. K., & Titov, V. V. (2012). Kastykhtakh mammoth from Taimyr

(Russia). Quaternary International, 276–277, 269–277.

https://doi.org/10.1016/j.quaint.2011.11.022

Konidaris, G. E., Athanassiou, A., Tourloukis, V., Thompson, N., Giusti, D., Panagopoulou, E., &

Harvati, K. (2018). The skeleton of a straight-tusked elephant (Palaeoloxodon antiquus) and

other large mammals from the Middle Pleistocene butchering locality Marathousa 1

(Megalopolis Basin, Greece): preliminary results. Quaternary International, 497, 65–84.

https://doi.org/10.1016/j.quaint.2017.12.001

159

Krafft, B. A., Kovacs, K. M., & Lydersen, C. (2007). Distribution of sex and age groups of ringed

seals Pusa hispida in the fast-ice breeding habitat of Kongsfjorden, Svalbard. Marine Ecology

Progress Series, 335, 199–206.

Kullberg, J. C., da Rocha, R. B., Soares, A. F., Duarte, L. V., & Marques, J. F. (2014).

Palaeogeographical Evolution of the Lusitanian Basin (Portugal) During the Jurassic. Part I:

The Tectonic Constraints and Sedimentary Response. In R. Rocha, J. Pais, J. C. Kullberg, &

S. Finney (Eds.), STRATI 2013 (pp. 665–672). Springer International Publishing.

Lapparent de Broin, F. de, & Antunes, M. T. (2000). Pleistocene Chelonians from Gruta da Figueira

Brava (Arrábida, Portugal). Memorias Da Academia Das Ciencias de Lisboa, Classe de

Ciencias, 38, 105–116.

Larramendi, A. (2015). Shoulder Height, Body Mass, and Shape of Proboscideans. Acta

Palaeontologica Polonica, 61(3), 537–574. https://doi.org/10.4202/app.00136.2014

Larramendi, A., Palombo, M. R., & Marano, F. (2017). Reconstructing the life appearance of a

Pleistocene giant: size, shape, sexual dimorphism and ontogeny of Palaeoloxodon antiquus

(Proboscidea: Elephantidae) from Neumark-Nord 1 (Germany). Bollettino Della Società

Paleontologica Italiana, 56(3), 299–317.

Lister, A. M. (2009). Late-glacial mammoth skeletons (Mammuthusprimigenius) from Condover

(Shropshire, UK): anatomy, pathology, taphonomy and chronological significance. Geological

Journal, 44(4), 447–479. https://doi.org/10.1002/gj.1162

Lister, A. M., Dimitrijević, V., Marković, Z., Knežević, S., & Mol, D. (2012). A skeleton of ‘steppe’

mammoth (Mammuthus trogontherii (Pohlig)) from Drmno, near Kostolac, Serbia.

Quaternary International, 276–277, 129–144. https://doi.org/10.1016/j.quaint.2012.03.021

Lister, A. M., Parfitt, S. A., Owen, F. J., Collinge, S. E., & Breda, M. (2010). Metric analysis of

ungulate mammals in the early Middle Pleistocene of Britain, in relation to taxonomy and

biostratigraphy: II: Cervidae, Equidae and Suidae. Quaternary International, 228(1), 157–179.

https://doi.org/10.1016/j.quaint.2010.05.014

López-García, J. M. (2008). Evolución de la diversidad taxonómica de los micromamíferos en la

península ibérica y cambios paleoambientales durante el Pleistoceno Superior. (PhD thesis).

160

Universitat Rovira i Virgili. Retrieved from

https://dialnet.unirioja.es/servlet/tesis?codigo=80124

López-García, J. M., Berto, C., Luzi, E., Valle, C. D., Bañuls-Cardona, S., & Sala, B. (2015). The

genus Iberomys (Rodentia, Arvicolinae, Mammalia) in the Pleistocene of Italy. Italian Journal

of Geosciences, 134(1), 162–169. https://doi.org/10.3301/IJG.2014.48

López-García, J. M., Blain, H.-A., Sanz, M., Daura, J., & Zilhão, J. (2018). Refining the

environmental and climatic background of the Middle Pleistocene human cranium from Gruta

da Aroeira (Torres Novas, Portugal). Quaternary Science Reviews, 200, 367–375.

https://doi.org/10.1016/j.quascirev.2018.10.003

López-García, J. M., & Cuenca-Bescós, G. (2012). Changes in the geographical distribution of

Microtus (Iberomys) cabrerae (Thomas, 1906) from the Late Pleistocene to the Holocene.

Revista Española de Paleontología, 27(2), 117–124.

López-García, J. M., Cuenca-Bescós, G., Blain, H.-A., Cáceres, I., García, N., Made, J. V. D., …

Pacheco, F. G. (2012). Biochronological data inferred from the early Pleistocene Arvicolinae

(Mammalia, Rodentia) of the El Chaparral site (Sierra del Chaparral, Cádiz, southwestern

Spain). Journal of Vertebrate Paleontology, 32(5), 1149–1156.

https://doi.org/10.1080/02724634.2012.676584

Loureiro, A., Ferrand de Almeida, N., Carretero, M. A., & Paulo, O. S. (2008). Atlas dos Anfíbios e

Répteis de Portugal (1o). Lisboa: Instituto da Conservação da Natureza e da Biodiversidade.

Retrieved from http://www2.icnf.pt/portal/pn/biodiversidade/patrinatur/atlas-anfi-rept/inicio

Magri, D., Rita, F. D., & Palombo, M. R. (2010). An Early Pleistocene interglacial record from an

intermontane basin of central Italy (Scoppito, L’Aquila). Quaternary International, 225(1),

106–113. https://doi.org/10.1016/j.quaint.2009.04.005

Manne, T., Cascalheira, J., Évora, M., Marreiros, J., & Bicho, N. (2012). Intensive subsistence

practices at Vale Boi, an Upper Paleolithic site in southwestern Portugal. Quaternary

International, 264, 83–99. https://doi.org/10.1016/j.quaint.2012.02.026

Manne, T. H., Stiner, M. C., & Bicho, N. (2006). Evidence for bone grease rendering during the Upper

Paleolithic at Vale Boi (Algarve, Portugal). Promontoria Monografica, 3, 145–158.

161

Marks, A. E., Brugal, J.-P., Chabai, V. P., Monigal, K., Goldberg, P., Hockett, B., … Mallol, C.

(2002). Le gisement pléistocène moyen de Galeria Pesada (Estrémadure, Portugal) : premiers

résultats. PALEO. Revue d’archéologie préhistorique, 14, 77–100.

Marks, A. E., Monigal, K., Chabai, V. P., Brugal, J. P., Goldberg, P., Hockett, B., … Mallol, C.

(2002). Excavations at the Middle Pleistocene cave site of Galeria Pesada, Portuguese

Estremadura: 1997–1999. O Arqueólogo Português, 20, 7–39.

Marra, A. C. (2009). Pleistocene mammal faunas of Calabria (Southern Italy): biochronology and

palaeobiogeography. Boll Soc Paleontol Ital, 48(2), 113–122.

Martin, A. P., & Palumbi, S. R. (1993). Body size, metabolic rate, generation time, and the molecular

clock. Proceedings of the National Academy of Sciences, 90(9), 4087–4091.

https://doi.org/10.1073/pnas.90.9.4087

Martin, R. A. (2012). Victoriamys, a new generic name for Chaline’s vole from the Pleistocene of

Western Europe. Geobios, 45(5), 445–450. https://doi.org/10.1016/j.geobios.2012.01.001

Martínez Nistal, M. del C., Cordero del Campillo, M., Díez Baños, P., & Nuñez Gutiérrez, M. C.

(1986). La perdiz Perdix Perdix Hispaniensis, Reichenow,1892 (aves: galliformes), nuevo

hospedador de Goniocotes obscurus, Giebel, 1874 (Mallophaga: goniodidae). Anales de la

Facultad de Veterinaria de León, 32, 85–94.

Maschenko, E. N., Gablina, S. S., Tesakov, A. S., & Simakova, A. N. (2006). The Sevsk woolly

mammoth (Mammuthus primigenius) site in Russia: Taphonomic, biological and behavioral

interpretations. Quaternary International, 142–143, 147–165.

https://doi.org/10.1016/j.quaint.2005.03.013

Mayr, G., Hazevoet, C. J., Dantas, P., & Cachão, M. (2008). A Sternum of a Very Large Bony-

Toothed Bird (Pelagornithidae) from the Miocene of Portugal. Journal of Vertebrate

Paleontology, 28(3), 762–769. https://doi.org/10.1671/0272-

4634(2008)28[762:ASOAVL]2.0.CO;2

Mazo, A. V. (1998). El elefante fósil de Buelna (Asturias). Nota preliminar. Estudios Geológicos,

54(5–6), 273–274. https://doi.org/10.3989/egeol.98545-6223

162

Mein, P., & Antunes, M. T. (2000). Gruta da Figueira Brava: petits mammifères–Insectívora,

Chiroptera, Lagomorpha. Memorias Da Academia Das Ciencias de Lisboa, Classe de

Ciencias, 38, 163–177.

Melentis, J. K. (1963). Studien über fossile Vertebraten Griechenlands. 3. Die Osteologie der

pleistozänen Proboscidier des Beckens von Megalopolis im Peloponnes (Griechenland).

Hellenic Journal of Geosciences, 14, 1–107.

Moço, G., Guerreiro, M., Ferreira, A. F., Rebelo, A., Loureiro, A., Petrucci-Fonseca, F., & Pérez, J.

M. (2006). The ibex Capra pyrenaica returns to its former Portuguese range. Oryx, 40(3), 351–

354. https://doi.org/10.1017/S0030605306000718

Monteiro, M. S. G. (2016). As microfaunas do Abrigo da Pena d’Água (Torres Novas) : proposta de

metodologia para o estudo de microfaunas em contextos arqueológicos. Master thesis.

Retrieved from https://run.unl.pt/handle/10362/18098

Morales A., & Rodríguez J. (2009). Black rats (Rattus rattus) from medieval Mertola (Baixo Alentejo,

Portugal). Journal of Zoology, 241(4), 623–642. https://doi.org/10.1111/j.1469-

7998.1997.tb05737.x

Morales, J., & Baquedano, E. (2017). La colina de los Tigres Dientes de Sable: Los yacimientos

miocenos del Cerro de los Batallones (Torrejón de Velasco, Comunidad de Madrid). Obra

Social La Caixa, Consejo Superior de Investigaciones Científicas, CSIC. Museo Nacional de

Ciencias Naturales, Museo Arqueológico Regional, Comunidad de Madrid. Retrieved from

https://dialnet.unirioja.es/servlet/libro?codigo=706152

Morales Pérez, J. V., & Serra, A. S. (2009). The Quaternary fossil record of the genus Testudo in the

Iberian Peninsula. Archaeological implications and diachronic distribution in the western

Mediterranean. Journal of Archaeological Science, 36(5), 1152–1162.

https://doi.org/10.1016/j.jas.2008.12.019

Morales-Muñíz, A., Albertini, D., Blasco-Sancho, F., Cardoso, J. L., Castaños-Ugarte, P., Liesau von

Lettow-Vorbeck, C., … Riquelme Cantal, J. A. (1998). A preliminary catalogue of Holocene

equids from the Iberian Peninsula. Proceedings of The XIII International Congress of

Prehistoric and Protohistoric Sciences, 6, 65–82.

163

Moreno-García. M., & Pimenta, C. M. (2002). The Paleofaunal Context. In Portrait of the artist as a

child: the Gravettian human skeleton from the Abrigo do Lagar Velho—its archaeological

context (Vol. 22). Lisboa: Instituto Português de Arqueologia.

Mourer-Chauviré, C., & Antunes, M. T. (1991). Présence du grand pingouin,Pinguinus impennis

(aves, charadriiformes) dans le Pléistocène du Portugal. Geobios, 24(2), 201–205.

https://doi.org/10.1016/S0016-6995(91)80007-M

Mourer-Chauviré, C., & Antunes, M. T. (2000). L’Avifaune Pléistocène et Holocene de Gruta da

Figueira Brava. Memorias Da Academia Das Ciencias de Lisboa, Classe de Ciencias, 38,

129–162.

Nabais M., & Rodrigues F. (2017). Animal Bones from the Late Neolithic Site of Ponte da Azambuja

2 (Évora, Portugal). International Journal of Osteoarchaeology, 27(3), 500–507.

https://doi.org/10.1002/oa.2551

Nehring, P. A. (1899). On Myodes lemmus crassidens, var. nov. foss., from Portugal. Annals and

Magazine of Natural History, 4(22), 292–300. https://doi.org/10.1080/00222939908678201

Neto de Carvalho, C. (2010). Vertebrate tracksites from the Mid-Late Pleistocene eolianites of

Portugal: the first record of elephant tracks in Europe. Geological Quarterly, 53(4), 407–414.

Neto de Carvalho, C. (2014). Pegadas de lince-ibérico (Lynx pardinus) no Plistocénico Superior da

Ilha do Pessegueiro (Portugal) Footprints of Iberian lynx (Lynx pardinus) in the Upper

Pleistocene from Pessegueiro Island (Portugal). Comunicações Geológicas, 101, 513–516.

Neto de Carvalho, C. (2018). Bear tracks in the Pleistocene at north of Praia do Cavalo, Odemira

(Portugal). In Abstract Book of the XVI Annual Meeting of the European Association of

Vertebrate Palaeontology (p. 138). Caparica, Portugal: Marzola, M., Mateus, O., Moreno-

Azanza, M. (eds.).

Northcote, E., & Mourer-Chauviré, C. (1985). The distinction between the extinct Pleistocene

European Crane,Grus primigenia, and the extant Asian Sarus Crane, G. antigone. Geobios,

18(6), 877–881. https://doi.org/10.1016/S0016-6995(85)80041-3

Northcote, E., & Mourer-Chauviré, C. (1988). The extinct crane Grus primigenia Milne-Edwards in

Majorca (Spain). Geobios, 21(2), 201–208. https://doi.org/10.1016/S0016-6995(88)80018-4

164

Paine, O. C. C., Leichliter, J. N., Avenant, N., Codron, D., Lawrence, A., & Sponheimer, M. (2019).

The ecomorphology of southern African rodent incisors: Potential applications to the hominin

fossil record. PLOS ONE, 14(2), e0205476. https://doi.org/10.1371/journal.pone.0205476

Pereira, V., Soares, J., & Silva, C. T. da. (2017). Understanding the First Chalcolithic Communities of

Estremadura: Zooarchaeology of Castro de Chibanes, Portugal. Preliminary Results. Papers

from the Institute of Archaeology, 27(1). https://doi.org/10.5334/pia-483

Pérez, T., Albornoz, J., & Dom nguez, A. (2002). Phylogeography of chamois (Rupicapra spp.)

inferred from microsatellites. Molecular Phylogenetics and Evolution, 25(3), 524–534.

https://doi.org/10.1016/S1055-7903(02)00296-8

Pérez-García, A., Vlachos, E., & Arribas, A. (2017). The last giant continental tortoise of Europe: A

survivor in the Spanish Pleistocene site of Fonelas P-1. Palaeogeography, Palaeoclimatology,

Palaeoecology, 470, 30–39. https://doi.org/10.1016/j.palaeo.2017.01.011

Petronio, C. (1995). Note on the taxonomy of Pleistocene Hippopotamuses. Journal of Mountain

Ecology, 3, 53–55.

Petrova, E. A., Masutin, V. V., & Zhuykova, I. A. (2017). Two incomplete skeletons of woolly

mammoth (Mammuthus primigenius) from the late Pleistocene in the Kirov Region, European

Russia. Russian Journal of Theriology, 16(2), 157–175.

https://doi.org/10.15298/rusjtheriol.16.2.05

Pimenta, C. M. (2014). Microvertebrates. In Bom Santo Cave (Lisbon) and the Middle Neolithic

Societies of Southern Portugal (pp. 61–79). Faro: Universidade do Algarve.

Pimenta, C. M., Figueiredo, S., & García, M. M. (2008). Novo registo de Pinguim (“Pinguinus

impennis”) no Plistocénico de Portugal. Revista portuguesa de arqueologia, 11(2), 361–370.

Pimenta, C. M., Moreno García, M., & Lourenço, A. (2015). O registo ornito-arqueológico em

Portugal: inventários, comentários e mapas. Revista Portuguesa de Arqueologia, 18, 289–312.

Piñero, P., Agustí, J., Oms, O., Blain, H.-A., Laplana, C., Ros-Montoya, S., & Martínez-Navarro, B.

(2017). Rodents from Baza-1 (Guadix-Baza Basin, southeast Spain): filling the gap of the

early Pliocene succession in the Betics. Journal of Vertebrate Paleontology, 37(4).

https://doi.org/10.1080/02724634.2017.1338294

165

Piñero, P., Furió, M., Agustí, J., Blain, H.-A., & Laplana, C. (2015). Biochronological data for the

Early Pleistocene site of Quibas (SE Spain) inferred from rodent assemblage. Geologica acta,

13, 229–241. https://doi.org/10.1344/GeologicaActa2015.13.3.5

Pires, F., Cardoso, J. L., & Petrucci-Fonseca, F. (2002). Estudo arqueozoológico dos carnívoros do

povoado pré-histórico de Leceia (Oeiras). Estudos Arqueológicos de Oeiras, 10, 183–247.

Platts, E., Bailey, S., McGrath, D., & McGeough, G. (2003). A Survey of the Land Winkle Pomatias

elegans (Müller, 1774) in the Burren, Co. Clare. Biology and Environment: Proceedings of the

Royal Irish Academy, 103B(3), 197–201.

Poole, K. G., & Boag, D. A. (1988). Ecology of gyrfalcons, Falco rusticolus, in the central Canadian

Arctic: diet and feeding behaviour. Canadian Journal of Zoology, 66(2), 334–344.

https://doi.org/10.1139/z88-050

Povoas, L., Zilhão, J., Chaline, J., & Brunet-Lecomte, P. (1992). La faune de rongeurs du Pléistocene

supérieur de la grotte de Caldeirao (Tomar, Portugal). Quaternaire, 3(1), 40–47.

https://doi.org/10.3406/quate.1992.1971

Queirós, J., Acevedo, P., Santos, J. P. V., Barasona, J., Beltran-Beck, B., González-Barrio, D., …

Alves, P. C. (2019). Red deer in Iberia: Molecular ecological studies in a southern refugium

and inferences on European postglacial colonization history. PLOS ONE, 14(1), e0210282.

https://doi.org/10.1371/journal.pone.0210282

Rainho, A., Rodrigues, L., Bicho, S., Franco, C., & Palmeirim, J. M. (1998). Morcegos das Áreas

Protegidas Portuguesas (1st ed., Vol. 1). Setúbal: Instituto da Conservação da Natureza e das

Florestas. Retrieved from

http://www2.icnf.pt/portal/pn/biodiversidade/patrinatur/resource/docs/Mam/morc/rainho1998-

morcegos-ap/view

Rando, J. C., & Alcover, J. A. (2010). On the extinction of the Dune Shearwater (Puffinus holeae)

from the Canary Islands. Journal of Ornithology, 151(2), 365–369.

https://doi.org/10.1007/s10336-009-0463-6

Raposo, L. (1995). Ambientes, territorios y subsistencia en el Paleolítico Medio de Portuga.

Complutum, 6, 57–77. https://doi.org/10.5209/CMPL.31071

166

Reboleira, A. S., Goncalves, F., Oromí, P., & Taiti, S. (2015). The cavernicolous Oniscidea

(Crustacea: Isopoda) of Portugal. European Journal of Taxonomy, (161), 1–61.

Reumer, J.W.F. (1984). Ruscinian and early Pleistocene Soricidae (Insectivora, Mammalia) from

Tegelen (The Netherlands) and Hungary. Scripta Geologica, 73, 1–173.

Ribeiro, C. (1871). Descripção de alguns silex e quartizites lascados encontrados nas camadas dos

terrenos Terciario e Quaternario das bacias do Tejo e Sado ... Typ. da Academia.

Ripoll, M. P., Morales Pérez, J. V., Sanchis Serra, A., Aura Tortosa, J. E., & Montañana, I. S. (2010).

Presence of the genus Cuon in upper Pleistocene and initial Holocene sites of the Iberian

Peninsula: new remains identified in archaeological contexts of the Mediterranean region.

Journal of Archaeological Science, 37(3), 437–450. https://doi.org/10.1016/j.jas.2009.10.008

Rodríguez-Varela, R., García, N., Nores, C., Álvarez-Lao, D., Barnett, R., Arsuaga, J. L., &

Valdiosera, C. (2016). Ancient DNA reveals past existence of Eurasian lynx in Spain. Journal

of Zoology, 298(2), 94–102. https://doi.org/10.1111/jzo.12289

Rofes, J., & Cuenca-Bescós, G. (2011). Evolutionary history and biogeography of the genus Crocidura

(Mammalia, Soricidae) in Europe, with emphasis on Crocidura kornfeldi. Mammalian

Biology, 76(1), 64–78. https://doi.org/10.1016/j.mambio.2009.12.001

Ros-Montoya, S. (2010). Los proboscídeos del plio-pleistoceno de las cuencas de Guadix-Baza y

Granada (PhD thesis). Universidad de Granada. Retrieved from

https://dialnet.unirioja.es/servlet/tesis?codigo=63978

Sáez-Royuela, C., & Tellería, J. L. (1986). The increased population of the Wild Boar (Sus scrofa L.)

in Europe. Mammal Review, 16(2), 97–101. https://doi.org/10.1111/j.1365-

2907.1986.tb00027.x

Schuhmacher, T. X., Cardoso, J. L., & Banerjee, A. (2009). Sourcing African ivory in Chalcolithic

Portugal. Antiquity, 83(322), 983–997. https://doi.org/10.1017/S0003598X00099294

Sesé, C., Panera, J., Rubio-Jara, S., Pérez-González, A. (2011). Micromamíferos del Pleistoceno

Medio y Pleistoceno Superior en el Valle del Jarama: yacimientos de Valdocarros y HAT

(Madrid, España). Estudios geológicos, 67, 131–151.

167

Shil, S. K., Quasem, M. A., Rahman, M. L., Kibria, A. S. M. G., Uddin, M., & Ahasan, A. S. M. L.

(2013). Macroanatomy of the Bones of Pelvis and Hind Limb of an Asian Elephant (Elephas

maximus). International Journal of Morphology, 31(4), 1473–1478.

https://doi.org/10.4067/S0717-95022013000400052

Smith, S. E., Gregory, R. D., Anderson, B. J., & Thomas, C. D. (2013). The past, present and potential

future distributions of cold‐adapted bird species. Diversity and Distributions, 19(3), 352–362.

https://doi.org/10.1111/ddi.12025

Soares de Carvalho, G., & Cardoso, J. L. (1999). O Quaternarista Georges Zbyszewski (1909-1999).

Estudos Do Quaternário / Quaternary Studies, 2, 3–6.

Swennen, C., Nehls, G., & Laursen, K. (1989). Numbers and distribution of eiders somateria

mollissima in the Wadden Sea. Netherlands Journal of Sea Research, 24(1), 83–92.

https://doi.org/10.1016/0077-7579(89)90172-5

Teixeira, A., Venâncio, R., & Brito, C. (2014). Archaeological Remains Accounting for the Presence

and Exploitation of the North Atlantic Right Whale Eubalaena glacialis on the Portuguese

Coast (Peniche, West Iberia), 16th to 17th Century. PLOS ONE, 9(2), e85971.

https://doi.org/10.1371/journal.pone.0085971

Teixeira, C. (1979). Biografia de Georges Zbyszewski. Retrieved January 25, 2018, from

http://www.lneg.pt/CienciaParaTodos/edicoes_online/biografias/zbyszewski

Terrinha, P., Rocha, R., Rey, J., Cachão, M., Moura, D., Roque, C., … Bensalah, M. K. (2006). A

Bacia do Algarve: estratigrafia, paleogeografia e tectónica. Geologia de Portugal no contexto

da Ibéria, 1–138.

Torres Pérez-Hidalgo, T. (1984). Úrsidos del pleistoceno-holoceno de la península ibérica (PhD

thesis). Universidad Politécnica de Madrid. Retrieved from

https://dialnet.unirioja.es/servlet/tesis?codigo=2780

Torres Pérez-Hidalgo, Trinidad. (1979). Osos pardos fósiles encontrados en Portugal. Cuadernos de

Espeleologia, Santander, 9–10, 157–167.

Tsoukala, E., & Grandal-d’Anglade, A. (2002). Système de mesures du squelette des Ursidés. Études

et Recherches Archéologiques de l’Université de Liège (ERAUL), 100, 265–287.

168

Tsoukala, Evangelia, Mol, D., Pappa, S., Vlachos, E., van Logchem, W., Vaxevanopoulos, M., &

Reumer, J. (2011). Elephas antiquus in Greece: New finds and a reappraisal of older material

(Mammalia, Proboscidea, Elephantidae). Quaternary International, 245(2), 339–349.

https://doi.org/10.1016/j.quaint.2010.10.008

Valente, M. J. (2004). Humans and carnivores in the early upper Paleolithic in Portugal: data from

Pego do Diabo cave. Revue de Paléobiologie, 23(2), 611–626.

Valente, M. J., & Haws, J. (2006). Animal carcass utilization during the Late Upper Paleolithic

occupation of Lapa do Suão (Portugal). In Paleolithic Zooarchaeology in Practice (pp. 29–

38). BAR INTERNATIONAL SERIES.

Van Kolfschoten, T. (1993). The mammal Fauna from the Interglacial deposits at Maastricht-

Belvédère. Mededelingen Rijks Geologische Dienst, 47, 51–60.

Veiga Ferreira, O. (1975). Os rinocerontes quaternarios encontrados em Portugal. Comunicações

Geológicas, 59, 15–25.

Villaluenga, A. (2016). Presencia de felinos (Felis, Lynx y Panthera) en el registro arqueológico de la

Península Ibérica durante el Pleistoceno Superior. Archaeofauna, 25, 185–204.

Von den Driesch, A. (1976). A guide to the measurement of animal bones from archaeological sites:

as developed by the Institut für Palaeoanatomie, Domestikationsforschung und Geschichte der

Tiermedizin of the University of Munich. Peabody Museum Press.

Welter-Schultes, F. W. (1998). Die Landschnecken der griechischen Insel Gávdos, der südlichsten

Insel Europas. Haus der Natur.

Zbyszewski, G. (1943). Les eléphants quaternaires du Portugal. Comunicações Geológicas, 24, 71–89.

Zbyszewski, G. (1977a). Nova contribuição para o conhecimento da jazida Quaternária da Mealhada.

Memórias e Noticias, 84, 1–37.

Zbyszewski, G. (1977b). Três ossos de vertebrados quaternários. Comunicações Geológicas, 61, 191–

194.

Zbyszewski, G., & de Matos, R. (1959). Folha 26-D (Caldas da Rainha): Carta Geológica de Portugal,

na escala de 1:50.000. LNEG.

169

Zbyszewski, G., & Veiga Ferreira, O. (1990). Les Vertébrés Quaternaires Portugais à la Lueur des

Études Recentes. Revista da Universidade Coimbra, 35, 155–162.

Zilhão, J. (1987). A Gruta do Caldeirão (Pedreira, Tomar): Balanço de sete anos de escavações

arqueológicas (1979-1985). Algar (Bol. Soc. Portuguesa Espeleologia), 1, 29–38.

Zilhão, J. (1993). As origens da arqueología Paleolítica em Portugal e a obra metodologicamente

precursora de J. F. Nery Delgado. Arqueologia e Historia, 3, 111–127.

Zilhão, J. (1997). O Paleolítico superior da Estremadura portuguesa. Colibri.

Zilhão, J., Maurício, J., & Souto, P. (1991). A arqueología da Gruta do Almonda (Torres Novas).

Resultados das escavações de 1988-89. In Actas das IV Jornadas Arqueológicas (Lisboa

1990) (pp. 161–171). Lisboa: Associação dos Arqueólogos Portugueses.

170