The beetles used in this study is Nicrophorus nepalensis Hope, they were collected in Alishan, Taiwan (23o47'N, 120o70'E), at altitudes ranging from 1100 to 1650 m above sea level. The collection took place from June 2010 to May 2011, utilizing 30 hanging pitfall traps, each containing 5 g of chicken as bait. To avoid the effect of parasites on the experiment, all the beetles collected in field were anesthetized with carbon dioxide and placed under a microscope to remove mites with forceps. After that they were reared in 10.4 × 10.4 × 6 cm transparent plastic containers. The whole experiment would be using the same size containers. Each container was filled with a layer of moist peat that is approximately 4 cm thick. The temperature and the light-dark cycle in them were set to 18°C and 12 h/12 h respectively. Females could obtain and store sperm by mating in the field44,43, in order to clear the fertile sperm within their spermatheca, all the adults were placed in containers separated by genders for 30 days before the experiment, with each container accommodating up to six beetles. And they were fed with freshly decapitated Tenebrio molitor twice a week.
Measure of morphological characteristics
The pronotum width and the secondary sexual characteristic ratio (SSC) are regarded as the morphological characteristics for both genders in this research. The pronotum width is used as an indicator of body size, measured from the widest lateral point of pronotum with an electronic vernier caliper. The orange semitransparent and black pentagonal plaque on clypeus is the SSC of burying beetles, captured through a microscope (Fig. 3a-c). And the SSC ratio is defined as the ratio of the length of plaque to the length of clypeus.
Congenital parental genetic test
To examine morphological influence of parental body size and paternal SSC on offspring, field-collected males were divided into three groups: large and small males with complete SSC and small ones with degenerative SSC, while females were only divided into two groups by size (large and small). Here a large beetle is defined as the one with pronotum width greater than 5.5 mm, and the one with pronotum width less than 4.5 mm is called small. The threshold value for the ratio between complete and degenerative SSC is defined as 0.8. Whereby 6 different parental combinations can be obtained, which are LL, LS, SL, SS, DL and DS (L: large, S: small, D: small males with degenerative SSC; the first and second letters indicate the paternal size and the maternal size respectively). To facilitate breeding, each parental combination was placed in a container maintained under 12h/12h light-dark cycle at a constant temperature of 18°C. After feeding each pair with 25g of chicken breast for about 3 days, all containers were transferred to a darkroom at the same temperature. There was no interference in the development of larvae during the breeding until they left to pupate. To provide sufficient space for pupation, up to eight larvae preparing to pupate were accommodated in one container with 4 cm-thick moist peat. And all the containers were kept in an 18°C incubator with no light exposure. A month later, after the adults emerged, the gender determination, the measurement and calculation of their pronotum width and SSC ratio would be implemented.
Acquired food intake test
The parasite-reduced F1 offspring of field-collected beetles were used as parents in this test. Newly emerged laboratory-reared adults were transferred to new containers with 3 cm thick moist peat and kept in an incubator under the same temperature and light-dark cycle as before. Each container accommodates up to six individuals with same gender. And they were fed with freshly decapitated mealworms twice a week. Each unrelated breeding pair was placed in another new container with 5 cm thick moist peat about 30 days after they emerged. After a week passed, 25 g of chicken breast was put into each container. All the containers were transferred to the dark following oviposition. Four days after the chicken was put in, the parents finishing burying carcass were transferred to a new container with 3 cm thick moist peat. In order to send the newly hatched larvae in the old containers back to their parents, a piece of chicken was used as a bait to gather them. As larvae reached third-instar stage, the measurement of larval mass was conducted every 12 hours for the control of larvae’s food intake. Larvae were divided into seven groups based on their size, which weighed 0-50mg, 50-100mg, 100-150mg, 150-200mg, 200-250mg, 250-300m and 300mg-350mg respectively. For each group, several larvae were randomly selected to be blocked feeding and placed into a new container with 5 cm thick moist peat to pupate. Their gender, pronotum width and the SSC ratio were determined after their eclosion. To verify whether parental genetic influence is present in this test or not, 4 different parental morphological combinations, LL, LS, SS and DS, were additionally used to detect the influence of paternal or maternal morphology on the feeding blocked offspring.
Numerical simulation method
A segmented linear regression model was used to simulate the data in this study, and SegRegA program48 was applied to accomplish the simulation. According to the data analysis, there is a threshold value of male pronotum width in each group, and the growth rates of SSC before the threshold values were calculated. The fitted segmented linear regression models are as follow:
Y = KiX + Bi, when X < Bp,
Y = Ci, when X ≥ Bp,
where i={1,…,7}, Y is the SSC ratio; X is the male pronotum width; Bp is the breakpoint of the segmented linear equation (the threshold value of male pronotum width); Ki is the slope of ith regression function (the increase rates of SSC ratio) when X < Bp; Bi is the regression constant of ith regression function when X < Bp; Ci is the intercept of horizontal line of ith regression function (saturation value of ratio of SSC) when X ≥ Bp.
The determination of break-point is through assuming a range of break-points and then selecting the one with largest coefficient of determination, which is denoted R2, where R2 = 1-\(\frac{{SS}_{res}}{{SS}_{tot}}\), \({SS}_{res}\) is the residual sum of squares, \({SS}_{tot}\) is the total sum of squares49. By conducting the analysis of variation and significance, the best segmented linear regression models yield under the significance of regression coefficient is at least 90% when X < Bp, regression coefficient is insignificant when X ≥ Bp and standard error is relatively small.
Vocal organ and sound spectrum
The vocal organ of burying beetles is composed of scraper and file. Scraper is the two short transverse ridges on inner surface of trailing edge of two elytra and next to suture (Fig. 3d). File is the two longitudinal stridulatories on the fifth dorsal tergum of the abdomen with small ridges arranged finely (Fig. 3e). Beetles vocalize by swinging file up and down with abdomen and rubbing scraper which makes the air in elytra resonates. Observed by the anatomy, beetles’ auditory organ might be the trachea inside pronotum (Fig. 3f). Ten individuals were each selected from wild-caught males with complete SSC, relatively large males, males with degenerative SSC and large females. Photography and measurement of them were conducted under a dissecting microscope, which was used to compare differences of their small ridges spacings, the file lengths and the densities of small ridges.
In the interest of analyzing the impact of gender and SSC of beetles on their sound spectrum, this research simulated five scenarios: males with complete SSC running into females, males of the same size and males with degenerative SSC, as well as females encountering females of the same size and larvae. To record the sound of subjects, they were placed in transparent plastic containers (10.4 x 10.4 x 6 cm) to prevent them from drilling into the soil. In each experimental scenario, every two subjects were sequentially put into a container with chicken ten minutes apart so that the first subject can fit into the environment. The recording would be started after the second subject was placed in the container. The monitoring of subjects’ sound by earphones and observation of their behavior were conducted at the same time. Their sound was tracked by Audio-Technica AT9944 Microphone and recorded by Marantz PMD 620 Recorder. The recall level of recorder was set to -24dB, and the Low-cut function of microphone was turned on during recording. To avoid disturbance to beetles and simulate their state in the dark, this experiment was carried out in a homemade Styrofoam box and red lights on only. Three audio recordings would be made for each encounter between subjects. And for each recording, only the audio between tenth to twentieth second after first calling was selected for analysis. Avisoft Bioacoustics was applied to analyze the dominant frequency, fundamental frequency, pulse train and pulse rate (number of pulses / sec). To guarantee the sound was clear and from beetles, the analysis standard was set to 0.05V.
Calling behaviour test
For the sake of studying the influence of gender and body size on the calling behavior of beetles, we conducted five calling behaviour experiments during intrasexual and intersexual encounters and parental care. Beetles used for this calling experiment were the laboratory-reared F1 offspring, and they were all sexually mature and unmated adults. To avoid the interference of noise, these experiments were all carried in the laboratory with walls made of sound-absorbing cotton during midnight. In the first three experiments, the reaction displayed by males of different types in intrasexual competition was examined. In Experiment 1, multiple pairs of large males with complete SSC were sequentially put in a container with 20 g chicken. They have similar body size which means the difference between their pronotum widths does not exceed 0.1 mm. To distinguish them, one of them was marked with liquid whiteout. Once they met, the number of individuals calling within one minute would be counted. In Experiment 2, the calling behaviour of males with distinct body size competing for carcass was compared. For each male pair, two subjects were sequentially placed into a container with chicken 30 minutes apart, and their pronotum widths was greater than 1 mm. The numbers of individuals calling when they encountered other males with distinct body size would be counted correspondingly. The operation of the third experiment is basically same as the second one, except for using much more smaller males with degenerative SSC instead of small males. And the operation of the fourth and the fifth experiment are same as the first one. However, Experiment 4 focused on the calling behaviour of large males and small males with degenerative SSC when they met females while Experiment 5 analyzed the call behavior of average-size females when they ran into the individuals of same size in both genders and larvae. In Experiment 5, eggs were picked out and placed in a petri dish first, males would be removed when larvae were hatching. After that, larvae were placed onto carcass balls.
Statistical analysis
We used one-way ANOVA test and Scheffe’s test as post-hoc test to examine the differences in numbers, pronotum width and the SSC ratio of offspring among various parental combinations, different mass of offspring in the third-instar larvae, as well as beetles’ sound spectrum (Fundamental frequency, Pulse number and Pulse rate) in all experimental scenarios. And the same statistical methods were performed to analyze the differences in small ridges spacings, the numbers of small ridges and the file lengths among males with varying SSC, and beetles of different genders as well. Chi-square test was applied to find the correlations between the vocal behaviours of beetles and the SSC, body size and gender of the individuals they met. All statistical analyses were carried out in SPSS Statistics 17.0.