Small Mammals, Rodents, and Bats in Andean Temperate Forests

What research at Kodkod has revealed about rodents, bats, and an ancient marsupial

Introduction

The small mammal community of the Andean temperate forests of southern Chile represents one of the most ecologically distinctive assemblages in the Southern Hemisphere, combining ancient marsupial lineages, endemic rodent species, and an increasingly prevalent invasive black rat, all sharing a structurally complex forest environment shaped by bamboo understorey, tree cavities, and steep elevational gradients.

A sustained programme of research centred around the Pucón region of the Araucanía — involving researchers from Pontificia Universidad Católica de Chile, Universidad de Chile, Universidad Austral de Chile, and international collaborators — has generated a cohesive body of work addressing the behavioural ecology, population dynamics, interspecific interactions, physiological adaptations, and conservation genetics of this community. This review synthesises findings across four interrelated themes: the role of bamboo understorey as critical small mammal habitat; coexistence dynamics and pathogen transmission risks within the rodent assemblage; behavioural responses of rodents to predation risk; and the physiology, social behaviour, and population genomics of the marsupial monito del monte (Dromiciops spp.).

1. Bamboo Understorey as Keystone Habitat for Small Mammals

A synthesis of eleven years of systematic surveys (2006–2017) in the Andean temperate forests of the Araucanía documented the breadth of vertebrate fauna dependent on the bamboo (Chusquea spp.) understorey, recording 20 mammal species using this habitat for feeding, shelter, and reproduction (Ibarra et al. 2018, La Chiricoca no. 23). Among small mammals, the monito del monte, the arboreal rat (Irenomys tarsalis), the long-tailed mouse (Oligoryzomys longicaudatus), and the long-haired mouse (Abrothrix hirta) were all recorded using bamboo for nesting and foraging. Four bat species — including Myotis chiloensis, Histiotus magellanicus, Lasiurus varius, and Tadarida brasiliensis — were detected in bamboo-rich areas using bioacoustic monitoring systems.

The ecological significance of bamboo for the monito del monte is particularly well documented. Bamboo leaves constitute 63% of nest construction material for this species, providing thermal insulation during winter torpor and anti-parasitic properties associated with bamboo chemistry. Population densities of D. gliroides are positively associated with bamboo cover, and the species' limited dispersal capacity across open habitat means that the degradation or clearance of bamboo-rich forest fragments can cause effective local extinction through isolation. The authors recommend maintaining heterogeneous patches and corridors of bamboo across the elevational gradient as a minimum management requirement, noting that bamboo removal confers no timber production benefit and disproportionately reduces the habitat available for dependent native fauna.

2. Rodent Activity Patterns, Coexistence, and Pathogen Transmission Risk

Salgado et al. (2022, BMC Zoology) examined nocturnal activity patterns and interspecific interactions within a four-species rodent assemblage — O. longicaudatus, A. hirta, A. olivaceus, and the introduced black rat Rattus rattus — in secondary temperate forest near Pucón, using camera traps across 4,474 recorded events in autumn and summer. Although all species showed significantly different activity patterns from one another, assemblage-wide temporal overlap was substantially greater than random expectation (Czekanowski index = 0.87 in autumn, p = 0.002), indicating that temporal partitioning does not appear to be a primary mechanism facilitating coexistence in this community.

Most notably, R. rattus exhibited predominantly aggressive behaviour toward all three native rodent taxa, with aggression being the most frequent category of interspecific interaction recorded. This contrasts sharply with the primarily agonistic (avoidance-based) interactions observed among native species themselves. The authors discuss the implications of this aggressive dominance for pathogen transmission: R. rattus, which arrived in Chile over 300 years ago and has colonised many natural areas, shares ectoparasites and endoparasites with native rodents, and serological evidence from related studies implicates it as a potential reservoir for Andes orthohantavirus. The temporal overlap between R. rattus and the primary hantavirus reservoir, O. longicaudatus, combined with observed aggressive physical encounters, creates conditions that may facilitate direct pathogen spillover between species. The study underscores the need for further research on the role of R. rattus as a long-term disruptor of native rodent community dynamics in continental temperate forests.

A companion study on cavity use examined the coexistence of R. rattus and D. gliroides in 90 artificial nest boxes monitored by camera trap across three sites of varying disturbance and elevation (Barz Cabezas 2024, Universidad de Chile). Black rats were detected only at the two lower-elevation, more disturbed sites, with camera trap rates up to 8.5 times higher than those of D. gliroides at the most disturbed site. The highest marsupial detection rates occurred at Llancalil, the undisturbed high-elevation site where rats were absent. In sites where both species co-occurred, D. gliroidesexhibited a measurable shift in the onset of peak activity, suggesting behavioural adjustment in response to R. rattuspresence, though no direct predation or aggressive encounters were recorded. Slope emerged as the strongest environmental predictor of marsupial occurrence, with steeper terrain associated with higher detection rates — likely reflecting both the species' preference for bamboo-rich habitat and its avoidance of areas disturbed by human settlement.

3. Behavioural Responses of Rodents to Owl Predation Risk

Hernández et al. (2021, Animals) conducted the first field-based study of rodent antipredator responses to native owl calls in Chilean temperate forest, using playback experiments in which three sympatric rodent species — Abrothrix spp., O. longicaudatus, and R. rattus — were exposed to calls of two native raptors: the Austral pygmy owl (Glaucidium nana) and the Rufous-legged owl (Strix rufipes). Camera traps at foraging stations recorded feeding time, locomotor activity, and vigilance across 981 independent events.

All three species showed measurable antipredator responses to owl calls, but the nature and magnitude of those responses were markedly species-dependent, consistent with predictions from the landscape of fear framework. Abrothrix spp. reduced feeding time under both owl call treatments, while O. longicaudatus increased vigilance and fast locomotor activity without altering feeding time, reflecting the species' agility and partial bipedalism as escape traits. R. rattusresponded most strongly to S. rufipes calls — reducing feeding time and slow locomotion — consistent with field evidence that this larger owl preys on R. rattus while the smaller G. nana does not. Contrary to expectations, R. rattus showed responses comparable in sophistication to those of native species, suggesting adaptation to native predators over the more than 300 years since its introduction to Chile.

Moonlight and vegetation cover significantly modified perceived predation risk for all species, with brighter conditions generally increasing vigilance and locomotion. The authors propose that playback of native owl calls represents a promising tool for ecologically based rodent management strategies, with the caveat that the choice of predator signal must be matched carefully to the target rodent species.

4. Communal Nesting, Physiology, and Conservation Genomics of the Monito del Monte

Nespolo et al. (2022, Journal of Experimental Biology) used biophysical models — taxidermic mannequins of D. gliroides filled with agar to simulate living tissue — to quantify the energetic benefits of communal nesting under realistic winter ambient temperatures (5°C). Comparing cooling curves across six treatment configurations (solitary, grouped, and with or without nest), they found that clustering combined with nest use minimised euthermic maintenance costs by approximately 10% relative to grouping alone, and the savings relative to solitary individuals outside a nest amounted to roughly half the species' daily basal metabolic rate. The finding confirms communal nesting as an adaptive thermoregulatory strategy and not merely a social by-product, resolving a long-standing ambiguity in the literature. The nests themselves — spherical structures built primarily from quila bamboo leaves (Chusquea quila) — proved critical to the energy savings, consistent with evidence that nest mass and volume increase with elevation and thus with thermal challenge.

Vázquez et al. (2020, Austral Ecology) documented phenotypic flexibility in D. gliroides nesting behaviour, showing that the proportion of cavity nests relative to non-cavity nests declines from 50% in old-growth forests to 25% in secondary forests, where cavity supply is more limited. This flexibility — switching between natural tree cavities, artificial nest boxes, and open non-cavity nests on bamboo or shrub substrates — may be critical to population persistence in degrading landscapes, but raises questions about whether non-cavity nests provide equivalent thermal and predation refuge.

The most recent genomic investigation of the genus (Quintero-Galvis et al. 2024, Ecology and Evolution) used restriction site-associated DNA sequencing of 242 individuals across 24 localities in Chile and Argentina to characterise population structure and identify putative loci under environmental selection in both recognised species of Dromiciops: D. gliroides(monito del monte) and D. bozinovici (Pancho's monito del monte). Both species showed strong genetic structure consistent with isolation by distance, with four genetic clusters identified in D. gliroides and three in D. bozinovici. Notably, one sampling locality for D. bozinovici was the Kodkod Field Studies Center at Pucón, contributing to the southernmost known cluster of that species. Environmental variables — particularly summer precipitation and temperature seasonality — explained approximately 42–44% of genomic variance in both species, with candidate loci mapped to genes involved in carbohydrate metabolism (ALG8), muscle and neural regulation (MEF2D), and stress response (PTGES3). Genomic vulnerability modelling projected that allele frequencies at climate-associated loci will shift substantially by 2100 under high-emissions scenarios, particularly in populations at the southern and Andean limits of each species' range. The authors recommend prioritising conservation of populations with unique genetic variation — particularly the DG-A cluster of D. gliroides near the Los Ríos coast, and high-Andean populations of both species — and caution that the loss of these genetically distinct lineages would irreversibly reduce the adaptive potential of both species in the face of ongoing climate change.

5. Conservation Implications

Across this body of work, several conservation conclusions emerge with clarity. The bamboo understorey is not simply a structural feature of Andean temperate forest but a critical functional habitat for a diverse community of small mammals, and its removal — whether for timber operations, livestock access, or land clearance — has cascading consequences that are poorly represented in conventional forest management plans. The invasion of R. rattus into secondary and disturbed forest stands represents a novel and underappreciated threat to native rodent communities through aggressive competitive displacement and facilitated pathogen transmission, warranting ongoing monitoring and research at the landscape scale. The monito del monte, as an ancient marsupial lineage and keystone seed disperser of the Valdivian rainforest, faces cumulative threats from habitat degradation, cavity loss, bamboo clearance, and now projected genomic vulnerability to climate change — making it a compelling focus for integrated conservation strategies that span habitat protection, forest structure management, and population connectivity.

References

Barz Cabezas, F.P. (2024). Coexistencia de la rata negra (Rattus rattus) con el monito del monte (Dromiciops gliroides) en la ocupación de cavidades artificiales en bosques templados andinos de la región de La Araucanía, Chile. Memoria de Título, Universidad de Chile.

Hernández, M.C., Jara-Stapfer, D.M., Muñoz, A., Bonacic, C., Barja, I. & Rubio, A.V. (2021). Behavioral responses of wild rodents to owl calls in an austral temperate forest. Animals, 11(2), 428.

Ibarra, J.T., Altamirano, T.A., Rojas, I.M., Honorato, M.T., Vermehren, A., Ossa, G., Gálvez, N., Martin, K. & Bonacic, C. (2018). Sotobosque de bambú: hábitat esencial para la biodiversidad del bosque templado andino de Chile. La Chiricoca, 23, 4–14.

Nespolo, R.F., Peña, I., Mejías, C., Ñunque, A., Altamirano, T. & Bozinovic, F.F. (2022). Communal nesting is the optimal strategy for heat conservation in a social marsupial: lessons from biophysical models. Journal of Experimental Biology, 225, jeb244606.

Quintero-Galvis, J.F., Saenz-Agudelo, P., D'Elía, G. & Nespolo, R.F. (2024). Local adaptation of Dromiciops marsupials (Microbiotheriidae) from southern South America: implications for species management facing climate change. Ecology and Evolution, 14, e70355.

Salgado, R., Barja, I., Hernández, M.C., Lucero, B., Castro-Arellano, I., Bonacic, C. & Rubio, A.V. (2022). Activity patterns and interactions of rodents in an assemblage composed by native species and the introduced black rat: implications for pathogen transmission. BMC Zoology, 7, 48.

Vázquez, M.S., Ibarra, J.T. & Altamirano, T.A. (2020). Austral Opossum adjusts to life in second-growth forests by nesting outside cavities. Austral Ecology, 45, 1179–1182.