Unexpected Climate Revealed by a Middle Holocene Avian Assemblage from Fuerteventura (Canary Islands)

Abstract

A group of avian species, mostly small passerines, allows us to reconstruct the landscape and general climate of an area of Fuerteventura prior to the arrival of the first humans. Many of the bird species are typical of forest environments and the edges of bodies of water, conditions incompatible with the current hot and arid climate. The record of a high number of quail as well as small flying passerines surely implies the concurrence of two types of diurnal birds of prey, hunters on the ground and in flight, respectively. No trace of the abundant Puffinus holeae has been found, which evidently occupied a habitat very different from those in the north and interior of the island.

1. Introduction

In recent years, much progress has been made in understanding the birds that arrived and lived in the Canary Islands. Most of the information has been obtained over the last thirty years. Until now, no ornithological association formed during the period we are presenting had been studied. Not all, but a substantial part of the bibliography can be found in [1]. Previously, avian skeletal remains found in Quaternary sediments on the different islands did not point to climatic conditions clearly different from those of today. This paper presents a study of bird bones from the site Cueva del Llano (UTM coordinates: 607,237.98–3,170,403.41), located in northern Fuerteventura, in the municipality of La Oliva (Figure 1A). This cave is part of a volcanic tube located in the north of the island of Fuerteventura, a short distance from the village of Villaverde. The tube has several branches. The sedimentary levels reported here are found in the “Ramal Nuevo” (New Branch) of the cavity, approximately 25 m long, 1.5 to 3 m high, and an average of 2 m in width. Volcanic tubes are highly relevant for the preservation of paleontological information on volcanic and oceanic islands such as the Canary Islands [2,3,4,5,6].

El Llano Cave is the only known volcanic tube in Fuerteventura with a stratified sedimentary fill (Figure 1B) which includes Pleistocene and Holocene materials. Five fill phases are recognized, subdivided into nine distinguishable sedimentary units (CLL1 to CLL 9 from oldest to youngest). With the exception of Phase I, all the others contain fossil remains of vertebrates (reptiles, birds and mammals) and invertebrates (mainly land snails) [7,8]. However, the highest concentration of fossil remains corresponds to the last two fill phases (Phases IV and V). Taphonomic analysis of reptile and micromammal fossils from the Holocene sedimentary units (CLL7, CLL8, and CLL9) reveals that the high concentration of vertebrate fossil remains was due to the action of predators [8]. A fragment of the posterior skull of Haliaeetus sp. has also been found in this cave [9].

The bird bones reported here were collected near the entrance of the tube, in levels CLL 7 through CLL 9. CLL 7 was formed during phase IV. Levels CLL 8 and CLL 9 are more recent. They were formed during phase V. Previous uncalibrated radiocarbon dates [7,8] appear calibrated in

Table 1. Radiocarbon datings in Cueva del Llano (modified from [7]. a. Caracollina lenticula; b. Hemicycla sarcostoma; c. Monilearia monilifera; e, Cryptella auriculata; f. Rumina decollata; g. Theba cf. geminata; h. Theba sp.

Material	Stratigraphic Unit/Infilling Phase	Uncal. yBP	Cal. yBP (2 σ)	Lab Code
Land snails (a,b,c,e,g)	Unit CLL2/Phase II	16,830 +/− 900	18,551 +/− 1940	Gd-9730
Land snails (b,c,h)	Unit CLL3/Phase II	16,430 +/− 1000	18,072 +/− 2330	Gd-9728
Land snails (b,g)	Unit CLL-7/Phase IV	9280 +/− 370	8521 +/− 938	Gd-9734
Land snails (a,b,c,e,f,h)	Unit CLL-9/Phase V	6880 +/− 270	5786 +/− 479	Gd-9718

(Figure 1C). Thus, both dates circumscribe the time span of the palaeornithological sample reported here. For the first time, a large set of species that lived in the Canary Islands prior to the arrival of humans [10] are presented.

2. Materials and Methods

A total of 209 fossil bones from different layers of Cueva del Llano have been studied and identified (

Table 2. Avian bones found in Cueva del Llano, according to taxa and stratigraphical layers.

	Phase IV	Phase V		Total
	Layer CLL 7	Layer CLL 8	Layer CLL 9
Coturnix coturnix	64	84	1	149
Anatidae indet.	1			1
Calonectris diomedea/borealis	1			1
Chlamydotis undulata	1			1
Tyto alba	2			2
Jynx torquilla		1		1
Alaudidae indet.		2	5	7
Anthus cf. spinoletta			1	1
Phoenicurus phoenicurus		2		2
Luscinia luscinia	1			1
Saxicola dacotiae	2	9		11
Turdus merula	3	3		6
Sylvia cf. atricapilla			1	1
Sylvia cf. melanocephala		1		1
Regulus sp.	2	1		3
Cettia cetti		1		1
Aegithalos caudatus	1			1
Cisticola juncidis	2	11	1	14
Serinus canaria	2	3		5
Total	82	118	9	209

). Three quail bones found on the surface are noted, but not included in the total. All bones come from stratigraphic layers (CLL) 7 to 9. that is, from the layers formed during the last two phases of sedimentation in the cavity [7,8]. The material will be stored at the Faculty of Sciences (University of La Laguna), with the inventory codes: PCCRULL n.

The taxonomic identification of the bone remains was carried out using the comparative collection of Arturo Morales, at the Archaeozoology Laboratory of the Autonomous University of Madrid. No images or bibliographic descriptions were used for the final taxonomic identifications. For the identification of genera and species of passeriforms, only morphological and metric characteristics of the humeri have been examined, since the other bones do not have clearly diagnostic features.

Coturnix gomerae, Turdus sp., Sylvia sp., Bucanetes githaginea, and Haliaeetus sp. have been identified in the Llano cave [8]: footnote of Table 2. The corresponding skeletal remains have not been studied for the present report. The reference for the presence of Haliaeetus sp. is Rando (1995) [9].

For some specialists, the Atlantic shearwater, Calonectris borealis, should be treated as a species, while for others, it is a subspecies of Cory’s shearwater, Calonectris diomedea. Both forms are not distinguished from each other in the morphology or dimensions of their bones, so here we give the name Calonectris diomedea/borealis to the bone remains that we assign to this shearwater.

For the phenetic status and current geographical distribution of avian species, Martin & Lorenzo (2001) [11] and García del Rey (2011) [12] have been followed.

The bone nomenclature is that of Baumel & Witmer (1993) [13].

Conventional radiocarbon ages were converted to calendar ages using the IntCal20 calibration curve [14]. Calibrated ages are presented in Table 1 at the 2-sigma confidence level (95.4%).

3. Results

Galliformes (Temminck, 1820)

Phasianidae (Vigors, 1825)

Coturnix coturnix (Jaume et al., 1993)

Common quail

Surface: two skulls and one sternum (PCCRULL 1984 to 1986). Items from the surface are not included in Table 2.

Layer CLL 7: one skull, two left humeri, two right humeri, one proximal end and shaft of a right humerus, two right ulnae, one left carpometacarpus, one carpometacarpus right, four synsacra, four right femora, one left femur, three proximal ends of a right femora, one left tibiotarsus, one right tibiotarsus, one proximal end of a left tibiotarsus, one proximal end of a right tibiotarsus, two distal ends of right tibiotarsi, one left tarsometatarsus, one right tarsometatarsus, one distal end of a left tarsometatarsus, and one distal end of a right tarsometatarsus (PCCRULL 1987 to 2025). Juvenile individuals: two right coracoidea, four right humeri, one left humerus, one right ulna, one left ulna, two right femora, one left femur, one proximal end of a right femur, two right tibiotarsi, three left tibiotarsi, one proximal end of a right tibiotarsus, one distal end of a right tibiotarsus, two right tarsometatarsi, two left tarsometatarsi, and one distal end of a left tarsometatarsus (PCCRULL 2026 to 2050).

Layer CLL 8: two left scapula, three cranial ends of a right scapula, one cranial end of a left scapula, one right scapula, one right coracoideum, two furculae, three right humeri, four left humeri, three right ulnae, one left ulna, two right radii, two left radii, three right carpometacarpi, five left carpometacarpi, five synsacra, one pelvis fragment, five right femora, three left femora (Figure 2A), four proximal ends of left femora, five right tibiotarsi, five left tibiotarsi, one distal end of a right tibiotarsus, five left tarsometatarsi, six right tarsometatarsi, and one distal end of a right tarsometatarsus (PCCRULL 2051 to 2123). Juvenile individuals: one right coracoideum, one right ulna, one right femur, three right tibiotarsi, one left tibiotarsus, three left tarsometatarsi, and one right tarsometatarsus (PCCRULL 2124 to 2134).

Layer CLL 9: one right coracoideum (PCCRULL 2135).

Anseriformes Wagler, 1831

Anatidae indet.

Layer CLL 7: one distal left ulna with part of the shaft of a juvenile individual (PCCRULL 2136). The bone is eroded.

Procellariiformes Fürbringer, 1888

Procellariidae Vigors, 1825

Calonectris diomedea/borealis (Scopoli, 1769)

Cory’s shearwater.

Layer CLL 7: one left scapula (PCCRULL 2137).

In Puffinus, the cranial area of the scapula has a generally triangular shape and is wider than in Calonectris. In this genus, the cranial end is rounded. In Calonectris diomedea, the tuberculum coracoideum protrudes cranially; while in Puffinus, such a tubercle is absent.

Gruiformes Bonaparte, 1854

Otididae Rafinesque, 1815

Chlamydotis undulata (Jacquin, 1784)

Houbara bustard.

Layer CLL 7: one distal end of one left ulna (PCCRULL 2138) that preserves most of its diaphysis (Figure 2B).

At the distal end of the ulna, the most distinctive feature in Otididae is that the condylus dorsalis ulnae forms a step in its continuation with the diaphysis. Chlamydotis undulata is intermediate in size between Otis tarda and Tetrax tetrax.

Strigiformes Wagler, 1830

Tytonidae Ridgway, 1914

Tyto alba Scopoli, 1769

Barn owl.

Layer CLL 7: one right coracoideum and one cranial end of a right coracoideum (PCCRULL 2139 and 2140).

The area of the facies articularis clavicularis and the processus acrocoracoideus is more developed in Asio otus than in Tyto alba. The foramen n. supracoracoidei is larger in A. otus. The processus procoracoideus ascends cranially in T. alba, and runs more anteriorly in Asio otus, Asio flammeus, and Strix aluco. In Athene noctua, it runs cranially but is much smaller.

Piciformes (Meyer & Wolf 1810)

Picidae Leach 1818

Jynx torquilla Linnaeus, 1758

Wryneck.

Layer CLL 8: one left humerus juvenile (PCCRULL 2141) (Figure 2C).

It has a small crista bicipitalis, a developed ventral tuberculum, a triangular crista deltopectoralis, very small ventral and dorsal condyles, a very prominent flexor process, and a very deep insertion area supracondylare ventrale. The humeral head is flat and broad as in the family. The humerus of Picus is much larger. The only Dendrocopus with a humerus the size of the fossil is D. minor. The humerus of Jynx torquilla has a muscle insertion depression above the processus flexorius, which is absent in Picus and Dendrocopus.

Passeriformes Linnaeus, 1758

Alaudidae Vigors, 1825

Alaudidae indet.

Lark.

Layer CLL 8: one right humerus and one left humerus juvenile (PCCRULL 2142 and 2143).

Layer CLL 9: two right humeri and three left humeri (PCCRULL 2144 to 2148) (Figure 2D).

These humeri have been compared with those of Galerida cristata, G. theklae, Lullula arborea, Alauda arvensis, Melanocorypha calandra, Eremophila alpestris, and Calandrella brachydactyla. The humeri found in the Llano Cave are noticeably smaller than in the other species. The dorsal part of the fossa pneumotricipitalis is very shallow in C. brachydactyla, A. arvensis, and L. arborea. In fossils, it is deeper, as seen in E. alpestris. The humeri of Galerida and M. calandra are much larger and the dorsal part of the fossa pneumotricipitalis is deep. These findings could not be compared with those of Calandrella rufescens. Those of Calandrella brachydactyla are clearly different in morphology and size, but it cannot be ruled out that they belong to C. rufescens, or even to another alaudid.

The humeri are the size and morphology of those of E. alpestris. However, E. alpestris generally lives in mountainous areas, above the tree line. These environmental conditions do not appear to have existed in Fuerteventura, not even in the past. Taxonomic identifications must be made taking into account only intrinsic characteristics of the fossils; in this type of study, these are morphological features. However, in this case, as an exception, an extrinsic characteristic—environmental requirements of E. alpestris that are probably impossible in Fuerteventura—suggests that E. alpestris should be ruled out.

Motacillidae Horsfield, 1821

Anthus cf. spinoletta (Vieillot, 1820)

Water pipit.

Layer CLL 9: one left humerus (PCCRULL 2149) (Figure 2E).

The pneumotricipital fossa is wide and deep; is not separated by the ventral tuberculum. The crista bicipitalis is small. The humerus of Anthus pratensis and A. trivialis are larger. The fossil could not be compared with those of other Anthus species. It is similar to that of A. spinoletta.

Muscicapidae Vigors, 1825

Phoenicurus phoenicurus (L., 1758)

Redstart.

Layer CLL8: one left humerus left (Figure 2F) and one right humerus (PCCRULL 2150 and 2151).

The crista deltopectoralis is relatively long, longer than in Luscinia and much longer than in Saxicola. The fossa pneumotricipitalis is wider and deeper in Oenanthe. Erithacus rubecula and Phoenicurus ochruros are quite morphologically similar to Phoenicurus phoenicurus, but they are larger in size.

Luscinia luscinia (L., 1758)/megarhynchos (Brehm, 1831)

Thrush nightingale/Nightingale.

Layer CLL 7: one right humerus (PCCRULL 2152).

The fossa pneumotricipitalis is shallow. It is deeper in Oenanthe, Phoenicurus, Saxicola and Erithacus. The humerus of L. svecica is more robust in its proximal end.

Saxicola dacotiae (Meade-Waldo, 1889)

Canarian Stonechat.

Layer CLL 7: one right and one left humeri (PCCRULL 2153 and 2154).

Layer CLL 8: five right (Figure 2G) and four left humeri (PCCRULL 2155 to 2163).

The humerus of Saxicola dacotiae is quite similar to that of Certhia, but the latter has a noticeably longer pectoral crest and a larger bicipital crest. Certhia is smaller than S. dacotiae. The bone morphology is like that of Saxicola torquata and Saxicola rubetra, but it is smaller in size.

Turdidae Rafinesque, 1815

Turdus merula Linnaeus, 1758

Blackbird.

Layer CLL 7: two left humeri, one right humerus (PCCRULL 2164 to 2166).

Layer CLL 8: three left humeri (PCCRULL 2167 to 2169).

In size and general shape, the humerus of Turdus is similar to that of Sturnus. In Sturnus, the crista deltopectoralis is relatively shorter than in Turdus. The fossa pneumotricipitalis is smaller, rounder, and deeper in Sturnus than in Turdus. Turdus humeri do not always allow identification of the species in question, but in this case it was possible. Those of Turdus viscivorus and T. pilaris are longer and considerably more robust than those of T. merula. Those of T. torquatus and T. philomelos are somewhat longer but considerably more robust. The humeri of T. iliacus are smaller than those of T. merula. The fossils match those of T. merula in size and morphological details.

Sylviidae Vigors, 1825

Sylvia cf. atricapilla (L., 1758)

Blackcap.

Layer CLL 9: one right humerus juvenile (PCCRULL 2170).

All Sylvia species have a very robust, but short, crista deltopectoralis. Sylvia borin and Sylvia atricapilla have a very developed caput humeri, more so than in the other species of the genus. The fossa pneumotricipitalis is deeper in S. atricapilla than in S. borin. S. atricapilla is smaller in size than S. borin and S. communis, and larger than S. undata, S. curruca, and S. melanocephala. It has not been possible to compare it with S. conspicillata or S. nana.

Sylvia cf. melanocephala (Gmelin, 1789)

Sardinian warbler

Layer CLL 8: one right humerus (PCCRULL 2171).

The humeri of Sylvia borin and S. atricapilla are larger and more robust than that of S. melanocephala. That of S. comunis is longer. In S. undata, it is noticeably smaller. That of S. cantillans is smaller, but has a relatively more robust crista deltopectoralis. This crista is longer in S. melanocephala than in S. curruca, but in the latter it is more robust. The humerus found in the cave is indistinguishable from that of S. melanocephala, but it has not been possible to compare it with that of S. conspicillata or S. nana.

Regulus (L., 1758)

Regulus sp.

Goldcrest.

Layer CLL 7: two left humeri (PCCRULL 2172 and 2173) (Figure 2H).

Layer CLL 8: one left humerus (PCCRULL 2174).

Regulus regulus and R. ignicapillus (the latter today present in Moroccan forests) individuals have the same sizes, and their respective humeri are mutually indistinguishable. The humerus of Regulus is shorter than in Aegithalos caudatus. The crista deltopectoralis is relatively short in that taxon, and the crista bicipitalis is well developed. The humerus of Remiz pendulinus has a distinct shape, with a very short crista deltopectoralis and fused crus dorsale and ventrale fossae, with its crus dorsale notably bigger than the crus ventrale.

The measurements of the fossil humeri from El Llano cave reach higher values than Regulus regulus and R. ignicapillus (

Table 3. Measurements in mm of Regulus humeri. (*) [15].

	Maximum Length	Proximal Width	Diaphysis Width
Regulus regulus *	9.4–10.0	3.3–3.5	0.9–1.0
Regulus ignicapillus *	9.9–10.1	3.3–3.4	1.0–1.1
Regulus sp. (Cueva del Llano)	10.3	3.2	1.2
Regulus sp. (Cueva del Llano)	10.4	3.4	1.0
Regulus sp. (Cueva del Llano)	10.4	3.5	1.1

). For this reason, these bones are attributed here to Regulus sp. The measurements of the fossils could not be compared with those of the Canarian Regulus regulus.

Regulus regulus is currently the smallest bird in the Canary Islands. There are two non-migratory subspecies, Regulus regulus teneriffae in Tenerife and La Gomera and Regulus regulus ellenthalerae in El Hierro and La Palma. They live in relict forests of the Canarian laurel forest.

Cettia cetti (Temminck, 1820)

Cetti’s warbler.

Layer CLL 8: one left humerus (PCCRULL 2175) (Figure 2I).

The humeri of Phylloscopus trochilus and Ph. collybita are morphologically similar to that of Cettia, but smaller in size. Regulus is smaller in size, but more robust, and its deltopectoral crest is longer. In Troglodytes, the dorsal pneumotricipital fossa does not exist; it is a convex surface. In Certhia, the humerus is much larger and has a very different morphology. The fossil humerus has been compared with Cettia forticeps, C. diphone and C. cetti. The proximal end of the Cettia cetti’s humerus has an overall rounded shape. The crista bicipitalis is more protruding in Sylvia than in Cettia cetti, Phylloscopus, and Acrocephalus. The crus dorsale of the fossa pneumotricipitalis is narrower in Cettia cetti than in Sylvia, Acrocephalus, and Phylloscopus.

Aegithalidae Reichenbach, 1850

Aegithalos caudatus (Linnaeus, 1758)

Long-tailed tit

Layer CLL 7: one right humerus (PCCRULL 2176) (Figure 2J).

This species shows morphological resemblances with Regulus. However, the crista deltopectoralis is absolutely and relatively larger in Aegithalos caudatus than in Regulus. The crus dorsale and crus ventrale fossae are likewise more separated than in Regulus.

Cisticolidae Sundevall, 1872

Cisticola juncidis (Rafinesque, 1810)

Fan-tailed warbler.

Layer CLL 7: one left humerus and one right humerus (PCCRULL 2177 and 2178) (Figure 2K).

Layer CLL 8: seven right humeri and four left humeri (PCCRULL 2179 to 2189).

Layer CLL 9: one proximal end of right humerus (PCCRULL 2190).

The humerus is more robust in Cisticola than in the genera Phylloscopus, Hippolais and Acrocephalus. The fossa pneumotricipitalis dorsalis is shallower in Cisticola than in Sylvia and Acrocephalus. This fossa does not exist in Locustella and Hippolais. In Regulus and Cettia the humerus is quite small in size. Such a bone is larger and more robust in most of the species within Sylvia. The humerus in Sylvia conspicillata and S. melanocephala is equal in size and robustness to that in Cisticola juncidis. They are very similar in morphology; however, the fossae pneumotricipitalis ventrale and dorsale are wider in Cisticola than in the above Sylvia species.

Fringillidae Vigors, 1825

Serinus canaria (Linnaeus, 1758)

Canary.

Layer CLL 7: two left humeri (PCCRULL 2191 and 2192).

Layer CLL 8: three left humeri (PCCRULL 2193 to 2195) (Figure 2L).

The crista deltopectoralis is short and with little development. The crus dorsale fossae is short. The crista bicipitalis also has little development. The humerus of Serinus canaria bears morphological similarity with those of several species. Petronia petronia, Plectrophenax nivalis, Pyrrhula pyrrhula, Fringilla teydea, F. coelebs, and F. montifringilla are noticeably larger. Carduelis chloris is longer and more robust. Carduelis cannabina is similar in size, but much more robust in build. Serinus cucullata is much smaller. The humerus in Serinus serinus is of the same morphology, although smaller. Carduelis carduelis and Serinus canaria are the same size. The crista deltopectoralis is wider and more prominent in Carduelis carduelis than in Serinus canaria. The ventral tuberculum is also more robust in Carduelis carduelis. In this species, the crista bicipitalis projects ventrally more than in Serinus canaria.

4. Discussion

The Canary Islands are located near the Tropic of Cancer, placing them within the latitude range of subtropical climates. Fuerteventura is one of the western islands, situated just 10 km from Africa. The Azores High pace and the Trade Winds have a decisive influence on the climate of the Canary archipelago. This, combined with its low-lying topography, contributes to very low annual rainfall, resulting in a hot and arid climate according to the Köppen classification. The Greenland ice cover has increased and decreased over the Quaternary as a consequence of climate fluctuations. Recent ice core analysis from northern Greenland reveals that the highest Holocene temperatures would occur between 10 and 7 ka BP, 3 to 7 °C warmer than today [16]. This suggests that the animals studied here died in warmer conditions than those prevailing today.

Most of the findings, 149 bone remains out of 209, have been assigned to the common quail. This volcanic tube likely housed the nests of some diurnal raptor for a considerable period of time, or it was a resting place. The discovery of a considerable number of small mammal bones [8] also implies that nocturnal birds of prey used the cavity. The significant proportion of passerine bones allows for the identification of many bird species. However, it may be convenient to emphasize that the paleontological samples are considerably biased in a stochastic sense, as the only identifiable bones are the humeri, particularly their proximal ends.

Until very recently, fossil and sub-fossil bones of quail found in the Canary Islands were attributed to Coturnix gomerae, considered an extinct endemic species of the Canaries. Recently, it has been indicated that C. gomerae is not a valid species and that all quail bones found in the Canary Islands correspond to the extant C. coturnix [1]. There may have been variations in the size of Canary Island quails over time, as has happened with other birds in other places [1] (and bibliography therein), but no differences are seen in the proportions of the front and rear limbs.

No marks of fire or other possible indications of human manipulation have been seen on the bones. Many of the birds found in this volcanic tube are Passerida, and therefore diurnal, small, and fliers. Falcons were probably responsible for hunting and gathering them into the cavity. The presence of quail remains requires a diurnal raptor that hunts on the ground, perhaps a medium-sized accipitrid.

The avian assemblage found in Cueva del Llano is in essence a sample of a past ornithofauna from the surroundings of the cavity, collected by predators that used the cave for shelter. The simplest hypothesis is that such a sample does not contain all or almost all of the species that were present. The record for several species consists of a single bone, which denotes that their discovery in Cueva del Llano was unlikely. This low probability of being recorded probably extends to other taxa, with individuals that were not hunted, brought into the cave, and fossilized. The high proportion of quail bones in the sample could be interpreted as due to an abundance of these animals in the cave’s surroundings because it was close to or surrounded by extensive grasslands. However, it could also be a sample bias determined by the identity of the predators and their hunting habits, i.e., predators that specialize in hunting quail, like kites, buzzards, or goshawks. Peregrine falcon could capture these birds in flight, near the coast. However, this falcon is unlikely to have hunted in an environment with the characteristics indicated by the species found in the volcanic tube.

Table 4. Taxa found in Cueva del Llano grouped in their characteristic habitats.

Sands: Beaches, Dunes	Wooded Areas and Undergrowth	Bodies of Water and Riverside Shrubs	Grasslands
C. diomedea/borealis	Jynx torquilla	Anatidae indet.	Coturnix coturnix
	Ph. phoenicurus	Cisticola juncidis	Chlamydotis undulata
	Luscinia luscinia	Regulus sp.
	Saxicola dacotiae	Cettia cetti
	Turdus merula	Anthus cf. spinoletta
	Sylvia cf. atricapilla
	S. cf. melanocephala
	Aegithalos caudatus
	Serinus canaria

shows the habitats that the identified birds would have required and provides an overview of the hunting grounds of the predators that brought their prey into the cave. The number of species appearing in each habitat does not reflect the approximate size of those habitats in the area surrounding Cueva del Llano.

A considerable proportion of the species identified in Cueva del Llano do not currently live in Fuerteventura, or even in the Canary Islands. Three species that live in forests or prefer vegetation on riverbanks and lagoon shores are not found in the Canary Islands today: Aegithalos caudatus, Cettia cetti and Cisticola juncidis. Some birds are found on other islands, but not in Fuerteventura: Luscinia luscinia, Regulus sp., Turdus merula, and Serinus canaria, which are all typical of wooded areas. Calonectris diomedea/borealis is an oceanic species that comes ashore to excavate a burrow and breed. The soil must be very loose, such as slightly consolidated sand.

It is worth mentioning that no skeletal remains of Puffinus holeae or Puffinus olsoni, two extinct shearwaters endemic to Fuerteventura and Lanzarote, have been found. Both shearwaters disappeared due to hunting pressure from native humans in recent times, shortly after the arrival of Europeans in the Canary Islands [1]. Hole’s shearwater nested in large colonies in sandy areas and paleodunes of Fuerteventura (particularly in Jandía) and Lanzarote, where it dug warrens in large numbers. Their bones are abundant in the Cueva de Villaverde, another volcanic tube located in the same municipality as La Oliva. This cavity was occupied by indigenous humans. It seems clear that the absence of P. holeae in the Cueva del Llano is due to the different accumulating agents involved in the formation of both sites.

It seems likely that there was a lagoon or pond near the cave, around which large areas of riparian vegetation developed. Likewise, wooded areas with undergrowth, where there were even wrynecks, were probably also in the vicinity of the cave. The ornithological record from Cueva del Llano suggests that in the early stages of the Holocene, the dominant climate in the Canary Islands was much wetter than it is today. In Fuerteventura, there were bodies of water with riparian vegetation and more or less dense forest areas with shrubby undergrowth. Higher global temperatures than the present ones [16] may have led to changes in the annual displacements of the Azores High and promoted a more intense rainfall regime, which fostered the maintenance of more diverse habitats and, consequently, a significantly more diverse avian fauna than today. The birds linked to these habitats likely disappeared with climate changes, which led to notably more xeric conditions.

The results of this work contribute to increasing the heritage value of the Cueva del Llano paleontological site as a world reference for a stratified volcanic tube fill from the Quaternary. Therefore, its protection is necessary through the application of a Paleontological Heritage Protection Framework.