Shrews (Eulipotyphla, Soricidae) are aggressive predators with a voracious appetite and a constant need to forage (^{Churchfield, 1990}). However, little ecological and behavioral information is available for the species inhabiting the American tropics. With over 40 species, Cryptotis is one of the most diverse and widespread genera of shrews in the New World (^{Guevara, Sánchez-Cordero, León-Paniagua, & Woodman, 2014}). Most Cryptotis species occur in cold and humid environments and are mainly distributed throughout tropical montane forests (^{Guevara et al., 2015}, ^{Woodman et al., 2012}), which might be related to a high abundance and variety of prey for shrews, such as worms and insects. It has been hypothesized that some modifications in postcranial morphology permit some species of Cryptotis to dig and forage in tunnels (^{Choate, 1970}, ^{Woodman and Morgan, 2005}). Recent morphological studies of limb skeletons among Cryptotis have revealed graded transitions in these adaptations from semifossorial to more fossorial behavior (^{Woodman and Gaffney, 2014}, ^{Woodman and Morgan, 2005}). Unfortunately, because most of the species of the genus inhabit areas with poor accessibility and capture success is relatively low compared with that of other small mammals, little is known about the digging behavior of Cryptotis species based on direct observations (^{Díaz-de Pascual and De Ascencao, 2000}, ^{Woodman and Gaffney, 2014}).

One species for which an intermediate mode of semifossoriality has been inferred is the Mexican small-eared shrew (Cryptotis mexicanus): a small-sized shrew with an average body mass of 7.5 g and head-body length of 70 mm, a mole-like form, and enlarged forefeet and elongated foreclaws (^{Woodman and Gaffney, 2014}, ^{Woodman and Timm, 2000}). During a short-term survey using pitfall traps, 7 shrews were collected in July 2015 from San Antonio Matlahuacales, in the municipality of Chignahuapan, Puebla, Mexico (19°41.871′ N, 98°05.592′ W; 2900 m). Permits for fieldwork were issued by the local administration from Chignahuapan and the Secretaría de Medio Ambiente y Recursos Naturales (Semarnat; FAUT 006), following the American Society of Mammalogists guidelines on sample care and management of specimens (^{Sikes, Gannon, & The Animal Care and Use Committee of the American Society of Mammalogists, 2011}). Specimens exhibited the measurements and diagnostic characteristics of C. mexicanus according to ^{Woodman and Timm (2000)} and ^{Guevara et al. (2014)}. The microhabitat where shrews were found consisted of pine and oak trees and a dense fog that covered the forest area, with a slope ranging from 30 to 40 degrees. The ground had a litter layer about 5 cm thick, formed mainly by litter from pine and oak. The only 2 individuals collected alive were used for behavioral observations, and were immediately transported to Mexico City and introduced to a terrarium (90 × 60 × 30 cm) with a layer of 7 cm of soil and 2 cm of leaf litter and moss, simulating the soil conditions of the collection site. Ad libitum feeding in captivity consisted of live crickets, mealworms, earthworms, and fresh beef. Water was supplied through a container and ice cubes. Observations, videos, and photographs were recorded for short periods along the day for 15 days in a row.

The shrews showed high burrowing activity, and spent nearly all of their time underground. They ate large quantities of food every day, from 80 to 100% of their body weight, and were extremely active from the first hour after sunset to until about 2 h after dawn. The 2 individuals commonly made and used tunnels at a depth of 2–3 cm from the soil surface through vigorous use of the manus and claws (Fig. 1a, 1b). To catch prey, the shrews alternated between subterranean tunnels and surface runways. When the shrews used tunnels, they dug open holes to the surface to breathe and find prey (Fig. 1c). Once they found their prey, they pursued them immediately above surface, which could be motivated by sense of smell or taste (Fig. 1d; ^{Saarikko, 1989}). Shrews also stored raw beef underground when they had enough food to consume immediately. Food hoarding behavior might be beneficial if it reduces the frequency of interactions with competitors or predators, or when food supply is unpredictable (^{Rychlik & Jancewicz, 2002}).

Figure 1 Digging and foraging behavior of the Mexican small-eared shrew, Cryptotis mexicanus, under semi-natural conditions. 

The observations presented here support the hypothesis that the mole-like form and additional characters of the forelimbs in C. mexicanus are adaptations for burrowing ability and a semifossorial life-style in the wild. C. mexicanus occurs in cloud and pine-oak forests and has been found to be not only sympatric but also syntopic with other soricids (^{Guevara et al., 2015}). A recent study indicates that C. mexicanus can live under similar environmental conditions with C. magnus, C. parvus, C. goldmani, S. ixtlanensis, S. salvini, S. veraepacis, or S. macrodon (^{Guevara et al., 2015}); however, at a fine scale, evidence of niche differentiation could be found in terms of foraging mode and prey selection (^{Churchfield, 1990}, ^{Churchfield and Rychlik, 2006}). Therefore, observations of substrate use and locomotor behavior can provide information on coexistence and niche partitioning among species of shrews, as well as the ecological requirements of this little-known semifossorial mammal.

I wish to express my thanks to Dra. R. Mata-López for her leadership in the project on parasites of wild mammals. Financial support was provided by the Programa de Apoyo a Proyectos de Investigación e Innovación Tecnológica (PAPIIT-UNAM IN220113). I also thank Jamie Kass for the friendly English corrections. V. Sánchez-Cordero for providing the permit for collecting mammals. I dedicate this contribution to Fernando A. Cervantes, teacher, colleague and friend, whose enthusiasm has promoted the knowledge of Mexican shrews. And I thank my wife Steph for sharing this wonderful journey with me.