Skip to main content

Source data for: Antagonistic effects of intraspecific cooperation and interspecific competition on thermal performance

Cite this dataset

Shen, Sheng-Feng et al. (2020). Source data for: Antagonistic effects of intraspecific cooperation and interspecific competition on thermal performance [Dataset]. Dryad.


Understanding how climate-mediated biotic interactions shape thermal niche width is critical in an era of global change. Yet, most previous work on thermal niches has ignored detailed mechanistic information about the relationship between temperature and organismal performance, which can be described by a thermal performance curve. Here, we develop a model that predicts the width of thermal performance curves will be narrower in the presence of interspecific competitors, causing a species’ optimal breeding temperature to diverge from that of its competitor. We test this prediction in the Asian burying beetle Nicrophorus nepalensis, confirming that the divergence in actual and optimal breeding temperatures is the result of competition with their primary competitor, blowflies. However, we further show that intraspecific cooperation enables beetles to outcompete blowflies by recovering their optimal breeding temperature. Ultimately, linking abiotic factors and biotic interactions on niche width will be critical for understanding species-specific responses to climate change.


Beetle collection and maintenance


Lab experiments were conducted using N. nepalensis individuals from a laboratory-reared population. Our stable lab population was established in 2014 from 24 male and 24 female beetles caught near Meifeng, which is 2100m above sea level on Mt. Hehuan, Taiwan (24°5' N, 121°10'E). Since establishment, we have supplemented the lab strain with new individuals from the same location every one or two years to avoid inbreeding. We used hanging pitfall traps baited with rotting pork (mean ± SE: 100 ± 10 g) to collect adult beetles in the field. We checked traps and collected beetles on the fourth day after traps were set. In every generation, we established at least 20 families, approximately 600 individuals in total, to maintain the population within the lab. To ensure that beetles in the lab population were unrelated to each other, we always paired beetles collected from different traps. We then put one female and one male in a 20 × 13 × 13 cm box with 10 cm of soil and a rat carcass (75 ± 7.5 g). Approximately two weeks after introducing adult beetles, all of the dispersing larvae that were ready to pupate from each breeding box were collected and allocated to a small, individual pupation box. After roughly 45 days, beetles that emerged from pupae were housed individually in 320 ml transparent plastic cups and fed once a week with superworms (Zophobas morio). All breeding experiments were conducted in walk-in growth chambers that imitated natural conditions where the lab population was collected at 2100 m on Mt. Hehuan. Temperature was set to daily cycles between 19 at noon and 13 at midnight, and relative humidity was set to 83-100%. We completed all of the laboratory experiments within three generations.


Breeding performance in the common garden experiment

To investigate breeding TPCs, we conducted solitary pairing experiments in six temperature conditions—8, 10, 12, 16, 20 and 22ºC—in a common garden with no temperature variation in the lab. For each replicate, one male and one female were arbitrarily chosen from different nests to avoid inbreeding. We chose adult beetles that were sexually mature, roughly 2 to 3 weeks after their emergence. Each individual was weighed to the nearest 0.1 mg. We then placed the pair with a mouse carcass (75 ± 7.5 g) under each temperature condition in a transparent plastic container (21 × 13 × 13 cm with 10 cm of soil depth) for two weeks. Cases in which pairs fully buried the carcass and produced offspring were regarded as successful breeding attempts. Cases in which pairs failed to bury a carcass, or they buried it but did not produce offspring, were regarded as failed breeding attempts. The only instance (of 118 replicates) when beetles died during the experiment was excluded from analysis.


Thermal regulation of locomotor performance

To determine the TPC for locomotor behaviors, we conducted a series of treadmill experiments under three temperature conditions—12, 16 and 20ºC—in a common garden with no temperature variation in the lab. We chose these temperatures because 16ºC is the optimal performance temperature for reproduction, we wanted to further test whether this optimal temperature coincided with optimal physiological function.

We set 72 replicates in total (12ºC: 25 replicates; 16ºC: 25 replicates; 20ºC: 22 replicates). We arbitrarily selected individuals from different nests for replicates at each temperature. The beetles were brought to the experimental chamber one day before data collection began. Monofilament glued to the pronotum by UV glue attached each beetle to the treadmill, where it was allowed to walk at a stable speed of 1.5 m per min (see Figure 3). We turned off the treadmill if a beetle’s abdomen began to drag or if the beetle started to fly, both behaviors that indicated that the beetle could no longer walk. An individual was tested only once per day. After each experiment, beetles were returned to the transparent container with 3 cm of soil for recovery.

We measured each beetle’s pronotum and the ambient temperature during running with a thermal imaging infrared camera (FLIR Systems, Inc., SC305; thermal sensitivity of < 0.05ºC) at a resolution of 320*240 pixels. Pronotum temperature was measured at the center of the thorax and calculated as the average pronotum body temperature each minute until an individual dragged its abdomen or started flying. The ambient temperature was the average temperature of a 6 x 6 cm surface of the treadmill located near where the beetle was tested. The temperature difference was depicted by the difference between the beetle’s body and ambient temperatures.


Breeding performance in the field

Since our previous study showed that blowflies are the beetle’s main interspecific competitor (Sun et al. 2014), we conducted series of experiments in the field to investigate the breeding TPC with and without interspecific competition. In 2013 to 2016 (May-October), we investigated the natural pattern of N. nepalensis reproduction and its breeding success along an elevational gradient from 673m to 3422m on Mt. Hehuan in central Taiwan (24°11’ N, 121°17’ E) that encompasses broadleaf forests at lower elevations and mixed conifer-broadleaf forests at higher elevations. We chose 37 study sites, primarily in natural forests to avoid cultivated or open areas where temperatures are more variable (De Frenne et al. 2019) and replicated each treatment at least 3 times at each site. In each trial, a 75 g (± 7.5 g) rat carcass was placed on the soil to attract beetles and covered with a 21 × 21 × 21 cm (length x width x height) iron cage with 2 × 2 cm mesh to prevent vertebrate scavengers from accessing the carcass. We checked each carcass daily until it began to decay due to microbial activity (Payne 1965), was consumed by maggots or other insects, or was buried under the soil by beetles. If burying beetles completely buried the carcass, we checked the experiment after 14 days to determine if third-instar larvae appeared. Cases in which pairs produced third-instar larvae were regarded as successful breeding attempts. Cases in which pairs failed to produce larvae were regarded as failed breeding attempts.

Breeding experiments without blowflies were conducted in the same experimental sites from 2014 to 2017, and 2019 (May-October). The experimental design was the same as that described above, but we used screen mesh above the pots to also keep blowflies out. To record air temperature at every site, we placed iButton® devices approximately 120 cm above the ground within a T-shaped PVC pipe to prevent direct exposure to the sun but allow for air to circulate. One male and one female beetle that were reared in the lab were released into the pot to record fundamental breeding performance. After 14 days, we checked the pots to determine whether the burying beetles’ third-instar larvae appear. Cases in which pairs fully buried the carcass and produced larvae after 14 days were regarded as successful breeding attempts. Cases in which pairs failed to bury a carcass, or they buried it but did not produce larvae, were regarded as failed breeding attempts. Criteria of data exclusion was the same as the common garden experiments. The only 3 instances (of 178 replicates) when beetles died during the experiment were excluded from analysis.

The influence of cooperation on thermal performance in the field

To investigate how cooperative behavior influences TPCs, we manipulated the group size of beetles in the field at 38 sites along the elevational gradient. Our experimental device comprised a small plastic container (21 × 13 × 13 cm with 10 cm of soil) placed inside a large container (41 × 31 × 21.5 cm with 11 cm of soil). There were several holes on the small container’s side wall that allowed beetles to move freely between the two containers. A 2 x 2 cm iron mesh was placed around the top of the large container’s wall to let flies access the carcass but to keep out larger animals that might scavenge the carcass. Small, non-cooperative groups contained one male and one female, whereas large, cooperative groups contained three males and three females (Chen et al. 2020; Liu et al. 2020). We captured the local beetles using the same hanging pitfall traps described above, and then conducted two group size treatments at each site. Based on our previous work exploring the natural pattern of arrival times, we released the marked beetles into the experimental device 1, 2 and 3 days after the trials began at elevations of 1700–2000 m (low), 2000–2400 m (intermediate) and 2400–2800 m (high), respectively (for details, see (Sun et al. 2014).

Each experiment was recorded by a digital video recorder (DVR) to determine whether N. nepalensis successfully buried the carcass. We placed the same temperature measurement device as described above at every site. Cases in which beetles buried the carcass completely and produced larvae after 14 days after were regarded as successful breeding attempts. Cases in which beetles failed to produce larvae were regarded as failed breeding attempts.


Data analysis

We used generalized linear mixed models (GLMMs) with binomial error structure to compare thermal performance curves among treatments (with/without interspecific competitors; with/without intraspecific cooperation) in the field. The outcome of breeding success (1 = success, 0 = failure) was fitted as a binomial response term to test for differences in the probability of breeding successfully. The variables of interest (i.e. mean daily temperature, type of experimental treatment) were fitted as fixed factors. Other environmental factors (elevation, daily minimum air temperature) were fitted to test the generality of the results. However, since elevation, daily minimum air temperature, and mean air temperature were highly correlated, we only included mean daily temperature in the final model. We also modeled the potential nonlinear effects of the environmental factors by fitting a quadratic regression model and compared the model fit with the linear model. Thus, the thermal performance curves (TPC) were determined statistically by the GLMM. To account for repeated sampling in the same plot, we set the field plot ID as a random factor (coded as 1|plot ID) in the R package lme4 (Bates et al. 2014). We also included year as a random factor to account for sampling at different time points. (See Source Code File 1 for further details)

We used a general linear models (GLMs) to determine the rates of N. nepalensis breeding and locomotor performance in the lab. The outcome (1 = success, 0 = failure) was fitted as a binomial response term to test the difference in the probabilities of interest (burial or non-burial) under different temperatures. For locomotor performance, the outcome (1 = flying, 0 = not flying) was fitted as a binomial response term to test for a difference in the probability of interest (flying or not) under different temperatures conditions. The relationship of temperature difference (body temperature minus ambient temperature) was fitted as a Gaussian response to different temperature conditions. (See Source Code File 1 for further details)

Finally, the optimal temperature (Toptimal) of the TPC was calculated by taking the derivative of the regression line that described the relationship between temperature and the likelihood of breeding successfully. In other words, Toptimal was estimated from the unimodal statistical model of the TPC. To estimate TPC breadth, we calculated the 95% confidence interval of the regression line. The boundaries of the TPC were the points that there was not significant difference between regression lines and zero. All statistical analyses were performed in the R v3.0.2 statistical software package (Team. 2018). (See Source Code File 1 for further details)

Usage notes

Beetles died during the experiment were excluded from analysis. We have marked that in the dataset.


National Science and Technology Council, Award: 103-2621-B-001 -003 -MY3

National Science Foundation, Award: IOS-1656098

National Science and Technology Council, Award: 101-2313-B-001 -008 -MY3

Academia Sinica, Award: AS-SS-106-05

Academia Sinica, Award: AS-IA-106-L01