Geographic situation of the Canary Islands showing the islands (Fuerteventura and Lanzarote) and islets (Lobos and La Graciosa) where bones of Malpaisomys had been found, and the sites where the materials for this work were collected (Cueva del Llano, Jameo de La Puerta Falsa and infilled fissured at malpaís de Montaña de La Arena) 

Contexts in source publication

Context 1

... Malpaisomys insularis is a mouse-like rodent endemic to the eastern Canary Islands. It became extinct during the fourteenth century. It was a remarkable species living under hyperarid conditions. A dental microwear analysis was performed in order to determine its former diet. The elevated number of fine scratches found in Malpaisomys molars suggests that it consumed a significant part of Poaceae, grass consumption leaving the most distinctive features on dental wear facets. A graminivorous diet with a high amount of abrasive items is in agreement with the broad teeth of Malpaisomys , considered as adapted to grass consumption. However, in the absence of potential competitors over its native range, it is likely that Malpaisomys also foraged on dicots to meet higher nutrient and energetic requirements. The ecology of Malpaisomys is discussed from these results in the context of the desertic climatic conditions of the eastern Canary Islands and with a special concern on its small body size in contrast to other large-sized island murine species such as the giant rats of the central Canary Islands. Keywords Canary Islands . Extinct endemic mammal . Island evolution . Muridae . Palaeoecology . Trophic ecology Malpaisomys insularis , the Lava mouse, is one of the three extinct endemic species of rodents from the Canary Islands. It inhabited the eastern islands of Lanzarote and Fuerteventura, and at least two close islets (Lobos and La Graciosa) (Fig. 1). This mouse-like rodent belonging to the Murinae subfamily was recovered from upper Pleistocene and Holocene deposits (Hutterer et al. 1988; Michaux et al. 1991). Its remains are very abundant in some localities, mostly fillings of lava tubes, and are also found in fissures and small caves within lava fields called “ malpaíses ” (meaning badlands) and even at the foot of some small hills where barn owls ( Tyto alba gracilirostris ) roosted in the past. The estimated weight of Malpaisomys is c40 g (Boye et al. 1992). Until the arrival of humans and the introduction of Mus musculus some time between 756 cal BC – 313 cal AD (Alcover et al. 2009), Malpaisomys was the main food resource of the barn owl. It constituted more than 85% of its consumed biomass. Later, its importance decreased quickly (Boye et al. 1992; Castillo et al. 2001). Malpaisomys became extinct during the fourteenth century, its extinction being a coeval event with the European colonization (Rando et al. 2008). The study of the palaeoecology of an island-endemic rodent like Malpaisomys requires special attention to the location, climate, soil, and vegetation of the eastern Canary Islands. The dominant bioclimate in these islands is desertic hyperarid infra-thermomediterranean, and the main potential vegetation is the tabaiba scrub dominated by species of Euphorbia spp. (Rodríguez Rodríguez et al. 2005). These conditions were prevailing throughout the Holocene in spite of the punctual, more humid phases documented in the North Atlantic region (deMenocal et al. 2000), but were likely slightly less harsh than today, the present day local conditions being affected by human activities (for a survey on Fuerteventura, see Rodríguez Rodríguez et al. 2005). Malpaisomys exhibits a remarkable combination of morphological features such as 1) rather high limb-to-body length ratios as found in living species adapted to climbing in a rocky environment (Boye et al. 1992), 2) a narrow interorbital region associated with a wide zygomatic breadth typical for rodents living in open environments (López-Martínez et al. 1998), and 3) wide molars presenting a longitudinal crest directed toward the poste- rior group of cusps on the lingual side of upper molars or in central position on lower ones (i.e., the stephanodont crown pattern) (Hutterer et al. 1988; Renaud and Michaux 2004). Moreover, while many other island-endemic murines express an island gigantism (body mass reaching c1 kg), Malpaisomys presents a relatively moderate size contrasting with the endemic giant rats Canariomys from the central Canary Islands (López-Martínez and López-Jurado 1987; Firmat et al. 2010) Morphological traits such as the enlarged stephanodont molars with crowns getting flatter with increasing wear, seems to be an adaptation to a diet dominated by green vegetation and abrasive plant materials (Denys 1994; Renaud and Michaux 2004). The absence of potential mammalian competitors could have promoted an enlarge- ment of its trophic niche, a common phenomenon in island animals (Van Valen 1965) and already observed in an extinct Balearic insular rodent (Hautier et al. 2009). Thus, the trophic ecology of this species appears to be an unanswered question. The diet of Malpaisomys was approached by Renaud and Michaux (2004) on the basis of molar outline comparative analysis. Their results suggest a herbivorous diet, but this hypothesis remains untested. In order to fill this gap, we used dental microwear patterns to characterize the dietary ecology of the Lava mouse or at least recognize the ingestion of items leaving characteristic traits. Microwear analysis considers the combination of pits and scratches left by food items on the enamel surfaces. It is a powerful tool for the study of the dietary habits of ungulates and primates (e.g., Solounias and Semprebon 2002; Merceron et al. 2005). Recently, this approach has been applied to murid rodents (Gomes Rodrigues et al. 2009). Comparing the microwear patterns of Malpaisomys to a set of living species offers the opportunity to provide insight into its diet. Results are discussed taking into account the pre-human environmental conditions of the eastern Canary Islands and the peculiar morphological features of Malpaisomys . From a group of more than 30 samples, 13 individuals, exhibiting a wear facet with no marks of post-mortem weathering, were selected for microwear analyses. These teeth came from three localities (ancient owl pellet accumu- lations): Cueva de Llano ( n =7) and Montaña de La Arena, (infilled fissures at malpaís n =2) from Fuerteventura and Jameo de la Puerta Falsa, ( n =4) from Lanzarote (Fig. 1). All the materials are of Holocene age (see Rando et al. 2008; Alcover et al. 2009) We used the protocol developed by Merceron et al. (2005), adapted to the dental morphologies of mice and rats (Gomes Rodrigues et al. 2009). Translucent casts of the dental wear facets were photographed in light stereo- microscopy at ×100 ...

Context 2

... Islands . Extinct endemic mammal . Island evolution . Muridae . Palaeoecology . Trophic ecology Malpaisomys insularis , the Lava mouse, is one of the three extinct endemic species of rodents from the Canary Islands. It inhabited the eastern islands of Lanzarote and Fuerteventura, and at least two close islets (Lobos and La Graciosa) (Fig. 1). This mouse-like rodent belonging to the Murinae subfamily was recovered from upper Pleistocene and Holocene deposits (Hutterer et al. 1988; Michaux et al. 1991). Its remains are very abundant in some localities, mostly fillings of lava tubes, and are also found in fissures and small caves within lava fields called “ malpaíses ” (meaning badlands) and even at the foot of some small hills where barn owls ( Tyto alba gracilirostris ) roosted in the past. The estimated weight of Malpaisomys is c40 g (Boye et al. 1992). Until the arrival of humans and the introduction of Mus musculus some time between 756 cal BC – 313 cal AD (Alcover et al. 2009), Malpaisomys was the main food resource of the barn owl. It constituted more than 85% of its consumed biomass. Later, its importance decreased quickly (Boye et al. 1992; Castillo et al. 2001). Malpaisomys became extinct during the fourteenth century, its extinction being a coeval event with the European colonization (Rando et al. 2008). The study of the palaeoecology of an island-endemic rodent like Malpaisomys requires special attention to the location, climate, soil, and vegetation of the eastern Canary Islands. The dominant bioclimate in these islands is desertic hyperarid infra-thermomediterranean, and the main potential vegetation is the tabaiba scrub dominated by species of Euphorbia spp. (Rodríguez Rodríguez et al. 2005). These conditions were prevailing throughout the Holocene in spite of the punctual, more humid phases documented in the North Atlantic region (deMenocal et al. 2000), but were likely slightly less harsh than today, the present day local conditions being affected by human activities (for a survey on Fuerteventura, see Rodríguez Rodríguez et al. 2005). Malpaisomys exhibits a remarkable combination of morphological features such as 1) rather high limb-to-body length ratios as found in living species adapted to climbing in a rocky environment (Boye et al. 1992), 2) a narrow interorbital region associated with a wide zygomatic breadth typical for rodents living in open environments (López-Martínez et al. 1998), and 3) wide molars presenting a longitudinal crest directed toward the poste- rior group of cusps on the lingual side of upper molars or in central position on lower ones (i.e., the stephanodont crown pattern) (Hutterer et al. 1988; Renaud and Michaux 2004). Moreover, while many other island-endemic murines express an island gigantism (body mass reaching c1 kg), Malpaisomys presents a relatively moderate size contrasting with the endemic giant rats Canariomys from the central Canary Islands (López-Martínez and López-Jurado 1987; Firmat et al. 2010) Morphological traits such as the enlarged stephanodont molars with crowns getting flatter with increasing wear, seems to be an adaptation to a diet dominated by green vegetation and abrasive plant materials (Denys 1994; Renaud and Michaux 2004). The absence of potential mammalian competitors could have promoted an enlarge- ment of its trophic niche, a common phenomenon in island animals (Van Valen 1965) and already observed in an extinct Balearic insular rodent (Hautier et al. 2009). Thus, the trophic ecology of this species appears to be an unanswered question. The diet of Malpaisomys was approached by Renaud and Michaux (2004) on the basis of molar outline comparative analysis. Their results suggest a herbivorous diet, but this hypothesis remains untested. In order to fill this gap, we used dental microwear patterns to characterize the dietary ecology of the Lava mouse or at least recognize the ingestion of items leaving characteristic traits. Microwear analysis considers the combination of pits and scratches left by food items on the enamel surfaces. It is a powerful tool for the study of the dietary habits of ungulates and primates (e.g., Solounias and Semprebon 2002; Merceron et al. 2005). Recently, this approach has been applied to murid rodents (Gomes Rodrigues et al. 2009). Comparing the microwear patterns of Malpaisomys to a set of living species offers the opportunity to provide insight into its diet. Results are discussed taking into account the pre-human environmental conditions of the eastern Canary Islands and the peculiar morphological features of Malpaisomys . From a group of more than 30 samples, 13 individuals, exhibiting a wear facet with no marks of post-mortem weathering, were selected for microwear analyses. These teeth came from three localities (ancient owl pellet accumu- lations): Cueva de Llano ( n =7) and Montaña de La Arena, (infilled fissures at malpaís n =2) from Fuerteventura and Jameo de la Puerta Falsa, ( n =4) from Lanzarote (Fig. 1). All the materials are of Holocene age (see Rando et al. 2008; Alcover et al. 2009) We used the protocol developed by Merceron et al. (2005), adapted to the dental morphologies of mice and rats (Gomes Rodrigues et al. 2009). Translucent casts of the dental wear facets were photographed in light stereo- microscopy at ×100 magnification. Several distinct features of the dental surface were counted on the lingual facet of the hypocone of the first upper molar within 0.01 mm 2 . We considered three types of microwear features: wide and fine scratches and large pits (see details in Gomes Rodrigues et al. 2009). Small pits (<5 μ m) were not considered because of the weak correlation with diet in murids (Gomes Rodrigues et al. 2009). We tested for differences in microwear variables between localities of Malpaisomys using a non-parametric Kruskal – Wallis rank sum test. For paleodietary inferences, we used a dataset that evaluates the average microwear patterns of 17 murid species representing three dietary classes: grass- eaters, fruit-plant-insect eaters, and insect-plant eaters (Online Resource 1). Microwear variables were computed in a linear discriminant analysis considering these three groups. The diet of Malpaisomys was inferred by projecting averaged microwear data onto the discriminant axes and by comparing its position to living species. The averages for each locality were also independently projected onto the discriminant axes. To summarize the proximity of Malpaisomys with the living species, a cluster diagram was created from the Euclidian distances using the average linkage method (UPGMA). Statistical analyses were performed with R 2.8.1. Microscopic observation of the dental wear facets of Malpaisomys revealed an elevated number of fine scratches (Fig. 2). No differences between localities were detected for the three microwear variables (for all tests, p >0.5), showing the homogeneity of our sample. The discriminant plane, considering extant murids as reference, showed a clear distinction between fruit-plant-insect and grass-eaters (Fig. 3). This difference was mainly related to an elevated number of fine and a low number of wide scratches (details in Firmat et al. 2010). The overall mean for Malpaisomys clearly fell within the range of the grass- eaters with a probability ( P ) reaching 82%. Similar classifications were obtained for the samples of Cueva de Llano ( P >81%) and Jameo de la Puerta Falsa ( P >90%). However, the sample from Montaña de La Arena was less clearly classified as grass-eater ( P >60%). It fell within the grass-eaters but on the limit of the 90% confidence interval for the plant-insect class. The proximity of Malpaisomys with the insectivorous Uranomys ruddi likely results from an unidentified bias in the U. ruddi microwear patterns (low number of large pits) instead of actual dietary similarities. This is supported by a UPGMA analysis showing that Malpaisomys is grouped with most of the grass-eaters (Online Resource 2). The distribution of Malpaisomys remains suggests that it inhabited all available island habitats and not only lava fields. A recent ecologically analogous rodent is by some aspect Nesoryzomys (Cricetidae) from the Galápagos Islands, which occupies all available habitats including recent lava fields (malpaíses) (Hutterer et al 1988). Up to now, data directly connected to the ingestion of food were lacking for reconstructing the diet of the extinct Malpaisomys . Our microwear analysis provides the first dietary inference for Malpaisomys independent of ecomorpholog- ical assumptions. Despite the fact that the frequent ingestion of dust particles could alter microwear patterns by artificially increasing the number of fine scratches like those produced by the ingestion of silica phytoliths of Poaceae (Sanson et al. 2007), the diet should be considered as the major factor controlling the genesis of microwear (Gomes Rodrigues et al. 2009). The overall similarities in the microwear patterns between the specimens collected on Lanzarote and Fuerteventura suggest homogeneous dietary habits for Malpaisomys over its geographic range. This is not surprising, given the similar environmental conditions on both islands. The abundance of fine scratches on the enamel surfaces of Malpaisomys indicates that abrasive items such as Poaceae probably account for a significant part of its diet. Since a diet, mainly based on abrasive items, requires an increased occlusal surface, our results are in agreement with the broad molar outline (Renaud and Michaux 2004) and the stephanodont pattern (Hutterer et al. 1988; López-Martínez et al. 1998) recorded for Malpaisomys . However, our results do not exclude the possibility that the diet of Malpaisomys included a significant part of softer food items (e.g., dicot leaves or soft fruits) which would not leave microwear traces as distinctive as Poaceae. Accordingly, dental specialization in mammals can be driven not only by trophic ...