The timid invasion: Behavioural adjustments and range expansion in a non-native rodent

Animal behaviour can moderate or favour biological invasion processes, and the native fauna’s ability to adapt. The importance of behavioural traits for colonization success remains debated, as well as the nature of behavioural traits that facilitate colonization. We investigated behavioural responses associated with risk-taking and exploration, in both non-native bank voles (Myodes glareolus, N = 225) accidentally introduced to Ireland a century ago, and in native wood mice (Apodemus sylvaticus, N = 189), that decline in numbers with vole expansion.

We repeatedly sampled behavioural responses in three colonization zones: source (2 sites), expansion edge (4) and pre-arrival(2).  Only source and edge zones are occupied by bank voles, for more than 80 years and 1-4 years, respectively.

Individuals of both species varied consistently in risk-taking and exploration, and mice had not adjusted their behaviour to the presence of non-native voles, as their behaviour did not differ between the zones. Voles at the expansion edge were initially more risk-averse but habituated to the repeated testing situation, compared to voles that were consistently less risk-averse in the source population, indicating spatial sorting for risk-taking propensity along the expansion front We suggest that in non-native species under high predation pressure, the ability to develop timid phenotypes may thus represent a fundamental component for range expansions.

Methods
Trapping
Rodents were trapped in single-capture Longworth traps (Penlon Ltd., Oxford, UK) which were pre-baited with seeds from commercial bird feed for 2 nights prior to trapping. At each site, 24 traps were set in lines with 10 m spacing between them. Traps were baited and set overnight and checked and deactivated in the morning. Animals were kept inside their traps in the shade until testing, with replenished food and pieces of carrots and cucumber as water sources. Trapping was not conducted during rainy days. Sites were visited for two trapping sessions within 4 days and re-visited after 2–4 weeks to allow repeated testing on recaptured individuals. 

Behavioural tests

For behavioural testing, we adopted two standardized behavioural tests that are commonly used in animal personality studies of small mammals to assess boldness and spatial exploration in a novel environment. The set-up combined the dark-light test and the open-field test and was adjusted to be executed under field conditions without prior handling of animals. Test arenas were always set in shady locations or under canvas roofs to avoid direct sunlight, shade patterns and open sky above the arena. At the end of each test, the arenas were cleaned with ethanol 70%. All tests were conducted between 1000 and 1800, under natural light conditions. 

Dark-light test
We assessed boldness by measuring individuals’ willingness to emerge from a dark-covered shelter into an open, illuminated, and empty arena, which is perceived as dangerous for a rodent species with a plethora of predators. An opaque plastic tunnel (15 cm inner diameter, 30 cm long) was connected to an opening in the circular arena at ground level, blocked by a wooden door. Shortly before the tests, the animal was carefully transferred from the trap into a shuttle tube (15 cm outer diameter, 30 cm long, open at one side and blocked by a cylindrical sponge of the same diameter on the other) avoiding touching or shooing. The shuttle tube with the animal inside was then inserted with its full length into the plastic tunnel. After 1 minute of acclimatization to the dark tunnel, the door leading into the arena was opened and we measured the rodent’s latency to emerge (emergence) from the dark tunnel into the arena with its full body (without the tail). If animals did not emerge from the tunnel within 5 min (66% of tests in voles and 24% in mice), we slowly reduced the space in the tube by pushing in the soft, cylindrical sponge in order to gently displace the animal without touching it. 

Open field test
We quantified exploration by measuring the proportion of time each individual spent moving in the novel space of an experimental arena (activity), and also how often they crossed into its central part (crossings). We used round, plastic, foldable arenas (1.20 m diameter, 40 cm high) as portable test set-ups. Round arenas prevent animals sitting in corners. The central part of the arena (90cm diameter) was more exposed (open field) and could be perceived as potentially dangerous by the animal, compared to the area (15 cm wide) nearby the wall. We quantified exploration activity by measuring the proportion of time each individual spent moving in the novel space of an experimental arena, and also how often they crossed into its central part, which should be perceived as being even more exposed and potentially dangerous. Arenas were covered with mesh to prevent wood mice from jumping out. The open field test started as soon as the animals entered the arena from the dark tunnel of the dark-light test with the full body (without the tail). The door leading to the dark tube was closed quietly by the observer, or, if the animal was pushed out, the tube remained blocked by the sponge. For 5 minutes after emergence, we counted the number of 10 sec intervals spent moving in the arena (activity), and the number of crossings of the line dividing the wall area and the central part of the arena with the whole body (crossings). After testing, animals were removed from the arena by offering a hiding place, either by pulling back the sponge in the tube or by placing their familiar trap in front of them.

After the behavioural tests, animals were sexed and weighed, marked by an individual fur cut if not already marked, and photographed for later identification.

The dataset represents the raw data on individual traits and behavioural measures taken in the field and was not in any way processed.