Background: Anthropogenic changes are often studied in isolation but may interact to affect biodiversity. For example, climate change could exacerbate the impacts of biological invasions if climate change differentially affects invasive and native species. Behavioural plasticity may mitigate some of the impacts of climate change, but species vary in their degree of behavioural plasticity. In particular, invasive species may have greater behavioural plasticity than native species since plasticity helps invasive species establish and spread in new environments. This plasticity could make invasives better able to cope with climate change.

Goal: Here our goal was to examine whether reproductive behaviours and behavioural plasticity vary between an introduced and a native Onthophagus dung beetle species in response to warming temperatures and how differences in behaviour influence offspring survival.

Methods: Using a repeated measures design, we exposed small colonies of introduced O. taurus and native O. hecate to three temperature treatments, including a control, low warming, and high warming treatment, and then measured reproductive behaviours, including the number, size, and burial depth of brood balls. We reared offspring in their brood balls in developmental temperatures that matched those of the brood ball burial depth to quantify survival.

Results: We found that the introduced O. taurus produced more brood balls and larger brood balls, and buried brood balls deeper than the native O. hecate in all treatments. However, the two species did not vary in the degree of behavioural plasticity in response to warming. Differences in reproductive behaviours did affect survival, such that warming temperatures had a greater effect on survival of offspring of native O. hecate compared to introduced O. taurus.

Broader implications: Overall, our results suggest that differences in behaviour between native and introduced species is one mechanism through which climate change may exacerbate negative impacts of biological invasions. 

To investigate behavioural plasticity in these species, we used a repeated measures design to quantify reproductive behaviours of single species colonies (n=18 colonies per species) in response to changes in their thermal environment. Each colony had five beetles (two males and three females). Before the start of the trials, we weighed each beetle to control for the effects of body size on brood ball size. We placed all experimental beetle colonies in plastic 2 L rectangular containers (13.5 x 10.2 x 28.2 cm) filled to a depth of 24 cm with a 4:1 mixture of topsoil:sand. We mixed the soil with water to create a standardized moisture level across colonies, and we covered the container with aluminum mesh to prevent escape of the beetles. 

 We used 43W halogen light bulbs to heat experimental colonies because the bulbs produce a gradient of warming in the soil, mimicking soil gradients produced by the sun. The distance of the bulb to the soil surface determined the degree of warming at the surface of the soil and the steepness of the thermal gradient. While this set-up allowed us to produce soil gradients similar to field settings, we were unable to simultaneously measure the effect of increased temperature variation on reproductive behaviour; the temperature gradients produced by our warming treatments were thus consistent throughout the trial and did not fluctuate. The control treatment temperature is slightly lower than field averages over the breeding season but has led to high reproductive output in laboratory conditions for these species. The low warming treatment mimics average ambient high temperatures in the field throughout the summer breeding season (29.5°C). The high warming treatment reflects temperatures commonly reached during heat waves at our collection sites, and such heat waves are predicted to become more common due to climate change (IPCC, 2014). To record soil temperatures experienced by our colonies in all three treatments, we buried data loggers (Onset Hobo Pendant Temperature/Light Logger) at the surface, middle, and bottom of containers that were filled with soil but did not have beetles three times during the experimental period. To maintain warming conditions throughout the trial length, we kept bulbs on during the entire trial, such that all colonies experienced consistent light (no dark periods).

We held each experimental colony at each of the three temperature treatments in random order for ten days (30 days total/colony). We fed colonies 130 ± 5g of autoclaved cow dung on days one, three, and six of each ten-day trial. On day ten, we searched through the soil in three cm sections (0-3cm, 3-6 cm, 6-9cm, 9-12 cm, 12-15cm, 15-18cm, 18-21cm, 21-24cm) and removed any brood balls produced by the experimental colony. For each brood ball we recorded mass and soil section where it was buried (i.e. burial depth).

To determine the effect of behavioural plasticity on offspring size and survival, we reared all brood balls at the average temperature of the soil layer in which they were buried. To approximate this burial temperature, we binned the container into thirds based on depth, including the top (0-9cm deep), middle (9-15 cm), and bottom (15-24 cm) of the container. We used data from the data loggers to quantify temperatures for each treatment and depth. We then used incubators to rear offspring in temperatures that reflected the brood ball location and, thus, the soil temperature in the containers. For the control treatment, we reared offspring in brood balls found in all three sections of the container at 25°C, which reflects the lack of thermal gradient in these containers. For the low warming treatment, we reared offspring in brood balls found in the top third at 29°C, those in the middle third at 26°C, and those in the bottom third at 25°C. For the high warming treatment, we reared offspring in brood balls found in the top third at 33°C, those in the middle third at 27°C, and those in the bottom third at 25°C. For rearing offspring, we placed each brood ball in an individual, sealed 75ml plastic cup with holes punched in the lid. We placed each brood ball at the bottom of the plastic cup and packed soil around the brood ball up to the lip of the cup. Throughout development, we added water to the cups using a spray bottle to maintain soil moisture.

We checked brood balls for beetle emergence starting four weeks after the end of the experimental trial. If beetles had not emerged after six weeks, we determined if the brood ball had an egg chamber (hollow portion of the brood ball). If the brood ball did not have an egg chamber, we considered the brood ball empty and we removed it from data analysis since it could be a food cache that does not reflect parental investment or reproductive behaviour. If the brood ball had an egg chamber, we categorized it as a mortality event during development.  

The dataset contains raw data for every brood ball produced by experimental colonies.