Evolution of Expending Extra Effort in Making a Dung Mass before Making a Brood Ball in the Nesting Behavior of the Female Dung Beetle Copris acutidens (Coleoptera; Scarabaeoidea)

Abstract

Nutrient limitations have often caused the evolution of mechanisms for efficient nutrient acquisition. The mouthparts of adult dung beetles efficiently acquire nutrients from a fiber-rich diet. Conversely, primitive mouthparts force larvae to survive on a low-quality diet despite experiencing the most demanding growth stages. In this study, we investigated the nutritional conditions and microbial community of the larval diet through the nesting behavior of the dung beetle Copris acutidens. We revealed that diet quality (C/N ratio) increased during the process of making the brood ball, irrespective of dung type. The sequencing of the bacterial community based on a partial 16S rRNA gene and the fungal community that targeted ITS2 region revealed that the fungal community in the female gut was the closest to the larval diet, whereas the bacterial community was not. The proportion of fungal Trichosporonaceae tended to increase with a decreasing C/N ratio irrespective of dung type and was alive in the larval gut. We suggest that Trichosporonaceae is a gut symbiont of both the adult female and larvae of C. acutidens, which is transmitted to the dung mass and then to larval gut through the brood ball, and that females have evolved the extra effort processes in their nesting behavior to compensate for larval diet quality, which is likely associated with symbiont fungi within the family Trichosporonaceae.

1. Introduction

In the ecology of growth, reproduction, and survival, nutrient limitations are a central concern for many organisms [1]. Herbivores can often be limited by nitrogen deficiency; thus, they have evolved mechanisms to efficiently acquire and conserve nutrients [2]. Lignocellulosic biomass is one of the most abundant carbon resources on Earth; however, the consumption of the resource is accompanied by the difficulties of degradation [3,4] and nitrogen deficiency [1]. Therefore, many herbivores rely on symbiotic microorganisms to digest lignocellulose and have capacious guts to maximize the function of gut symbionts and enzymes because the degradation of the stable polymer is a slow process. Even though many insects have been known to have intrinsic cellulases [5], a combination of intrinsic enzymes and microbes is still important for many insects that utilize fiber-rich diets [3,4]. In dung beetles, both adults and larvae utilize mammalian excrement as their food source. Dung has been suggested to be a nutritionally complete diet for dung beetles [6,7]; however, it is still a fiber-rich diet, and the average dung composition of several herbivores consists of approximately 30% cellulose, 20% hemicellulose, and 20% lignin [6]. In response to this fiber-rich diet, adult dung beetles have mouthparts that filter out coarse dietary fibers and can enrich the proportion of nitrogen [8,9,10,11]. In contrast, larvae consume a diet that is richer in cellulose and poorer in nitrogen compared to the adult diet because larvae are primitive and unselective bulk feeders with chewing mouthparts that are unable to filter out dietary fiber [12,13]. Holter (2016) revealed that the nitrogen content of dung pats moderately decreased between one week and 30 days [6]. This means that the aged dung that larvae usually consume has less nutritional value than the fresh dung that the adults eat. In short, larvae are stationary, and they depend on the provisioning they receive from their parents, whereas adults can supplement their nutritional needs by moving the sporadically distributed dung pats. In some dung beetle species that make brood balls, which function as both a food source and a chamber for larvae, the quantity of available resources through their developmental period is limited. These limitations are even more severe when larvae are in the most demanding growth stages. We, therefore, hypothesize that dung beetles must have a mechanism to efficiently acquire nutrition for the growth and survival of the larvae from the brood balls provided by parental care.

Preparing a dung mass before forming it into brood balls and then making brood balls for larvae are activities that are assumed to function as a “fermentation device” or “external rumen” [14,15,16]. However, Rougon et al. (1990) investigated the microbial symbiont activities in three dung beetle species (Onitis alexis, Euoniticellus intermedius, and Onthophagus gazella) and suggested that microbial activity occurred in the digestive tract of the larva rather than in brood balls [17]. In addition, by testing the location of egg pedestals inside brood balls, considered to be a maternal gift in E. intermedius, Byrne et al. (2013) concluded that feeding larvae did not depend on microbial symbionts within the brood balls [18]. In contrast, studies on dung beetles (O. taurus and E. intermedius) have suggested that the bacterial community of the female gut is transmitted to the offspring through the brood ball to compensate for the nutritionally incomplete diet [19] and to assist with cellulose degradation [20]. The oldest assumption for such a fermentation device was described for a Copris species in Souvenirs Entomologiques, edited by Jean-Henri Fabre [16]. This assumption was adopted for other taxa with similar nesting behaviors [14,15]. In these species, however, the assumption of a dung mass as a fermentation device remains to be clarified.

Copris species do not prepare brood balls directly from dung that is carried into the nest. Rather, females make a dung mass from the dung carried into the nest, which is kept for approximately a week as is. The females then divide the dung mass so as to make several brood balls [14] (refer to [21,22] for detailed photos on Copris acutidens). These nesting behaviors are highly conserved in this genus. In this study, we investigated the nutritional value and microbial composition of the larval diet through the different stages of the nesting behavior in C. acutidens.

2. Materials and Methods

2.1. Beetle Breeding Experiments

2.1.1. Using Cow Dung

Adult C. acutidens beetles (Figure 1A) were collected from central Japan between May and June (early reproductive season). Each female and male pair was bred in a container half filled with soil (Figure S1A). Fresh cow dung was mixed well to a uniform quality, and 250 mL of fresh cow dung was provided to each pair. A sample was stored in the freezer for baseline values for nutrition and microbiota. The containers were kept at room temperature, and a dung mass made artificially with approximately half the volume of 250 mL was placed into a hole dug into the soil as a control. Seven days later, some nests were checked, and the dung masses that the female beetles had made in the nest were sampled (defined as “Early dung mass”). A sample of the artificially made dung mass (control) was also collected at this time, defined as “Early artificial dung mass”. Similarly, 14 days after the beginning of the breeding experiment, the remaining containers were checked, and the brood balls with an egg and the remaining dung mass were collected from the beetle nest and defined as “Dung ball” and “Late dung mass”, respectively. A sample of the artificially made dung mass was also collected as “Late artificial dung mass”. A small amount of the inner part of each sample was used for DNA extraction and the remaining for C/N ratio measurements, detailed below. All samples were stored in the freezer until used.

2.1.2. Using Horse Dung

Each female and male pair was bred in a container half filled with soil by providing 250 mL of fresh horse dung mixed well to a uniform quality (Figure S1A,C). After 17 days, the containers were checked, and the brood balls made by the females were sampled. The remaining dung on the soil that the pair did not bury in the nest (Figure S1D) was defined as “residual dung”, and the residual dung in each container was also sampled for the baseline values for nutrition and microbiota. After a small amount of the inner part of the brood balls was used for the culture-dependent method detailed below, all remaining samples were stored in the freezer until used for DNA extraction and the measurement of the C/N ratio. The larvae were used for dissection, as described below.

2.2. Dissection

To obtain the gut samples, adult females and larvae were dissected (Figure 1). Adult abdominal part under the elytra was dissected (Figure 1B,C), and the midgut and hindgut were observed, whereas larvae were observed from the base of the head to the tip of the abdomen (Figure 1D,E). The adult gut was used to determine microbial communities based on DNA extraction and subsequent sequencing, and adult and larval guts were used for fungal symbiont isolation on artificial media, as described below.

2.3. Microbial Community

2.3.1. Bacteria

To examine the changes in the bacterial communities through the nesting processes, a cloning method was conducted. A Power Soil DNA isolation KIT (Mo Bio Laboratories, Carlsbad, CA, USA) was used to extract the total DNA of the microbes from each sample in the breeding experiments using cow dung and the female gut. To clarify the bacterial community, a fragment of the 16S rRNA gene was amplified via PCR with Ex Taq (Takara, Kusatsu, Shiga, Japan) and the primers 16SA2 (5′-GTGCCAGCAGCCGCGGTAATAC-3′) and 16SB2 (5′-CGAGCTGACGACARCCATGCA-3′), using a PCR protocol of initial denaturation at 95 °C for 5 min, followed by 25 cycles of 95 °C for 20 s, 55 °C for 20 s, and 72 °C for 50 s, and a final extension step at 72 °C for 5 min. The PCR product was cloned with the TA-cloning vector pT7Blue and Escherichia coli DH5α competent cells (Takara, Shaga, Japan) using ampicillin and X-gal blue–white as a color selection system. A second PCR was performed by using the white colonies as template DNA and AmpliTaq Gold DNA polymerase (Applied Biosystems, Waltham, MA, USA). The universal primers U-19 (5′-GTTTTCCCAGTCACGACGT-3′) and BT7 (5′-TAATACGACTCACTATAGGG-3′) were used via a PCR protocol with an initial denaturation at 95 °C for 10 min, followed by 30 cycles of 95 °C for 45 s, 54 °C for 30 s, and 72 °C for 1 min. The PCR products were purified with the ExoSAP-IT Express PCR Cleanup Reagents (Applied Biosystems, USA). Sequence analysis was performed on an ABI 3100 DNA Sequencer. The BLAST searches of the National Center for Biotechnology Information (NCBI) were performed to identify the sequence reads. All reads were deposited in the DDBJ database with accession numbers LC770449–LC770838.

2.3.2. Fungi

Culture-Independent Method

To clarify the fungal community, sequencing on an Illumina MiSeq high-throughput sequencing platform was performed by targeting the internal transcribed spacer 2 (ITS2) region. The same DNA samples that were used for the bacterial community and newly added DNA samples from the breeding experiment using horse dung were used in a two-step tailed PCR method. The ITS2 region was first amplified via PCR using Ex Taq (Takara, Shiga, Japan); the primers gITS7 and ITS4, with an overhang adapter; and a PCR protocol described as follows: initial denaturation at 94 °C for 2 min, followed by 30–35 cycles consisting of 94 °C for 30 s, 50 °C for 30 s, and 72 °C for 60 s and a final extension step at 72 °C for 5 min. The second PCR was performed using the first PCR products as the template and Ex Taq (Takara, Shiga, Japan). The primers consisted of overhang adapters and indices for the sample-specific adapters to generate a product that was compatible with the flow cell of MiSeq and a PCR protocol of an initial denaturation at 94 °C for 2 min, followed by 10–12 cycles consisting of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 60 s and a final extension step at 72 °C for 5 min. The library quality was assessed using a DNF-915 dsDNA Reagent Kit and Fragment Analyzer™ (Advanced Analytical Technologies, Orangeburg, NY, USA). Libraries were sequenced on an Illumina MiSeq platform, and 300 bp paired-end reads were generated. All the raw reads were deposited in the DDBJdatabase with the BioProject accession number PRJDB15647.

Sequence reads (total 577,565 reads) were processed using the QIIME2 genome analysis computing platform. Paired-end raw data with primer, adapter, and conserved regions were trimmed using the Q2-ITSxpress plugin. Trimmed reads with less than Q20 were denoised by setting both truncation end lengths as 0 bp using the DADA2 plugin because the quality of the data was good (average 96% > Q20). To assign taxonomy, a classifier was trained by naïve Bayes in QIIME2 using the sequence and taxonomy data of the dynamic thresholds in the UNITE database (ver. 9.0). The taxonomy of the obtained sequence reads was assigned by the trained classifier. Sequence depth was set to 1500 from the rarefaction curve.

Culture-Dependent Method

We investigated whether the symbiont fungi could be isolated from the adult and larval gut and brood balls from the breeding experiments using horse dung. Each sample from the brood balls was directly streaked on yeast and malt (YM) medium and potato dextrose agar (PDA) medium. Each gut sample (adult and larva) was crushed in a tube with a 0.7% sterile saline solution. This 50 μL solution was plated on YM and PDA media. All media were incubated at 25 °C. Picking and streaking were repeated until a single colony was obtained. DNA extraction from a single colony was conducted using a Power Soil DNA isolation KIT (Mo Bio Laboratories, USA). The ITS2 region was amplified via PCR with Ex Taq (Takara, Shiga, Japan), primers (ITS5, ITS4), and the PCR products were purified using MagExtrator (TOYPBO, Osaka, Japan). Sequence analyses and BLAST searches were performed, as detailed above.

2.4. Measuring the Carbon: Nitrogen Ratio

To clarify the change in the nutritional conditions through nesting progress, the C/N ratio was measured. Each sample that was used to clarify the microbial communities and an additional early dung mass was dried at 60 °C in an incubator and ground with a food blender. Each sample of 200 g dry weight was placed on a tray with quartz filter paper and was analyzed using SUMIGRAPH NC-220F (Sumika Chemical Analysis Service, Tokyo, Japan).

2.5. Statistical Analysis

2.5.1. Linear-Mixed Effect Models

To clarify the change in C/N ratio of the cow dung in female nesting, we developed a linear-mixed effect model (LME) using the lmer function in the lme4 package for R. The C/N ratio values were included as a response variable, the number of days (the period from the start of breeding to when containers were checked) as a fixed effect parameter, and the different container as the random effect. To determine the contribution of females to the C/N ratio of the horse dung, an LME model was developed (as described above), in which the C/N ratio values of the brood ball and the residual dung were included as a response variable, with the presence or absence of the females as a fixed effect parameter, and the different container as the random effect. To determine the contribution of the females to the observed read number of a selected taxon from the fungal community in the samples from the horse dung experiment, an LME model was developed (as described above), in which the read numbers of the selected taxon for the brood ball and residual dung were included as a response variable, with the presence or absence of the female as a fixed effect parameter, and the different container as the random effect, as well as the total read number of each sample as an offset function.

2.5.2. Principal Coordinate Analysis

To analyze the relationship among the operational taxonomic units (OTUs) at the species level in each community, principal coordinate analysis (PCoA) was performed on the Bray–Curtis distance and Jaccard distance for bacteria using PC-ORD 6.08 software and QIIME2 for fungi.

3. Results

3.1. C/N Ratio

Regarding the C/N ratio in cow dung processed by the females, in the LME model, there was a significant difference in deviance between the fit and null models (type III analysis of variance: F value: 67.0, p < 0.05). Similarly, in the LME model for the C/N ratio in horse dung, there was a significant difference in deviance between the fit and null models (F value: 121.1, p < 0.01). Thus, the C/N ratio in the fresh cow dung that was fed to the beetle pair was significantly decreased after being altered into the larval diet through the nesting behavior of the females (Figure 2A). The C/N ratio of the horse dung was also significantly decreased after being processed by females (Figure 2B).

3.2. Bacterial Community

A total of 390 reads of 16S rRNA were obtained and assigned to 167 OTUs at the species level bacteria (Table S1). The PCoA analyses of based on both Bray–Curtis and Jaccard distances of these OTUs showed the female gut was clustered at a distance from the brood ball and dung masses, which was supported by randomization tests (axis 1, p = 0.001). The fresh dung was also clustered at a distance from them, irrespective of distance, but the randomization tests did not support this result (axis 2, p > 0.05) (Figure 3A,B). In fresh dung (total 59 reads), most of the OTUs belong to members of the phylum of Bacteroidota (Bacteroidetes) (44%, 13OTUs) or Bacillota (Firmicutes) (44%, 17OTUs) (Figure 4). In the early dung mass (total 57 reads), the members of Pseudomoadota (Proteobacteria) increased to 70% (16 OTUs). Similarly, in the artificial early dung mass (total 37 reads), Pseudomoadota were dominant, accounting for 70% (10 OTUs). In the late dung mass, Bacteroidota increased to 39% (10 OTUs), in which Bacteroidia accounted for 73%. In contrast, Bacillota remained at a low proportion of approximately 11%. The remaining 38% (15 OTUs) was accounted for by Pseudomonadota. Similarly, in the brood ball (total 76 reads), Bacteroidota and Bacillota accounted for 37% (10 OTUs) and 8% (6 OTUs), respectively. Pseudomonadota accounted for 46% (27 OTUs).

In the artificial late dung mass, unlike the late dung mass and brood ball, the Bacteroidota were not recovered (20%, 3 OTUs). Instead, Pseudomonadota were kept in the proportion of 63% (17 OTUs). In both female guts, members of Bacillota accounted for 87% (7 OTUs) and 82% (11 OTUs), whereas Bacteroidota accounted for approximately 10%.

3.3. Fungal Community

In the breeding experiment using cow dung, after quality control by DADA2, 125,297 nonchimeric reads were retained. After taxonomic assignment, a total of 114 OTUs at the species level were identified as fungi (Table S2). The PCoA analysis based on the Bray–Curtis distance showed that the female gut was closely clustered to the late dung mass and the brood ball (cluster 1), and fresh dung were closely clustered to the early dung mass (cluster 2). These clusters were distantly clustered from the artificial dung masses (cluster 3), and each artificial dung mass was distantly clustered (Figure 5A). PERMANOVA showed a significant difference among these clusters (F = 1.87, p = 0.009) but pairwise tests did not. Similarly, the PCoA analysis based on Jaccard distances showed that the fresh dung was closely clustered to the early dung mass and the artificial early dung mass, and the late dung mass was closely clustered to the brood ball. The former cluster was distantly clustered from the artificial late dung mass and female gut, and the latter cluster was positioned in the middle between the former cluster and the female gut (Figure 5B). However, PERMANOVA could not be conducted between one distant plot (artificial late dung mas, female gut) and another.

In the fresh dung (total 31,061 reads), approximately 86% of the reads (11 OTUs) belonged to members of the family Neocallimastigaceae (Figure 6A). This family still accounted for approximately 60% (10 OTUs) of those in the early dung mass (total 15,186 reads) and approximately 14% (9 OTUs) of those in the brood ball (total 17,338 reads), and 5% (7 OTUs) of those in the late dung mass (total 16,532 reads). However, it was almost absent in the artificial dung masses (early: 4%; late: 0.1%) and adult gut (0%). The family Trichosporonaceae accounted for approximately 67% of the female gut (3 OTUs). The number of Trichosporonaceae increased from the early dung mass (approximately 15%; 2 OTUs) to the late stage (85% 2 OTUs) and accounted for approximately 68% (3 OTUs) in the brood ball, whereas this family was almost absent in the fresh dung (0%) and artificial dung masses (Early: 0.3%, Late: 0%). In addition, of the Trichosporonaceae reads, Apiotrichum scarabaeorum and the other Apiotrichum accounted for 35% and 58% in the female gut, respectively (Figure 7). A. scarabaeorum accounted for more than 87% of the reads in the early and late dung masses and brood ball. In contrast, no Apiotrichum was found in the artificial dung masses.

In the horse dung experiment, after quality control via DADA2, 147,576 nonchimeric reads were retained. After taxonomic assignment, a total of 126 OTUs at the species level were identified as fungi (Table S2). The number of OTUs was not significantly different between the presence and absence of processing by the females (Kruskal–Wallis test, H = 0.07, p > 0.05), but the Shannon diversity value was significantly lower in the presence of processing by the females than in their absence (Kruskal–Wallis test, H = 4.27, p < 0.05). The PCoA analysis based on Bray–Curtis distance showed that each cluster in the presence and absence of processing by the females overlapped (permutational multivariate ANOVA (PERMANOVA), F = 1.18, p =  0.31), The analysis based on the Jaccard distance showed each cluster was divided by PC2 axis, although the difference did not reach statistical significance (PERMANOVA, F = 1.27, p =  0.056) (Figure 8), and a similar result was found through the dendrogram analysis of Trichosporonaceae, which included experiments using cow dung (Figure 7). Members of the family Neocallimastigaceae were limited regardless of the presence and absence of processing by the females (Figure 6B). The families of Lasiosphaeriaceae, Podosporaceae, Trichosporonaceae, and Psathyrellaceae accounted for a maximum of 30–40% of the total reads in each sample. In all containers, a similar tendency of higher read proportions found in those samples in which the females are present was found in the family Trichosporonaceae (Figure S2 and Figure 7), but the LME model for the read number of this family did not fit between the presence and absence of processing by the females (type III analysis of variance: F value, 2.91, p > 0.05).

The fungi isolated from the plate cultures that were assigned to Trichosporonaceae are presented in

Table 1. Fungi isolated from plate culture and assigned to Trichosporonaceae. Only Fungi with BLAST identity over 99% were shown. Used females are different from beetles of the fungal community. Sequence reads were deposited with the following accession number in DDBJ. Pictures from some plate cultures were taken under a microscope.

Samples	Assigned Fungal Name	BLAST Identity (%)	Accession Number
Brood balls
1–2	Apiotrichum scarabaeorum	99	LC763495
2–3	A. siamense	99	LC763496
2–3	A. siamense	99	LC763497
Larval gut
2–1 midgut	A. scarabaeorum	99	LC763498
2–1 hindgut	A. scarabaeorum	99	LC763499
2–2 midgut1	A. scarabaeorum	99	LC763500
2–2 midgut2	Trichosporon sp.	100	LC763501
2–2 hindgut	Trichosporon sp.	99	LC763502
Female gut
Female1 midgut	A.siamense	99	LC763503
Female2 hindgut	A.siamense	99	LC763504

. The genus Apiotrichum (Trichosporon) was isolated from brood balls, adult gut, and larval gut.

4. Discussion

4.1. Diet Quality for Larva

Many organisms have evolved mechanisms to efficiently acquire nutrients in nutrient-limited conditions. Adult dung beetles have mouthparts that efficiently acquire nutrients from a cellulose-rich and nitrogen-poor diet. In contrast, the primitive mouthparts of larvae only permit them to access low-quality nutrients, despite experiencing the most demanding growth stages [8,9,10,11,12,13].

In this study, we revealed that the quality of the diet of the larvae increased through the nesting process, irrespective of dung type. The C/N ratio is a widely used quality index for organic matter, with quality rising with a decreasing C/N ratio [6,10]. The C/N ratio in the breeding experiment using cow dung was observed to decrease in phases. In the first 7 days, the decrease in the C/N ratio in the early dung mass was at a similar level to the early artificial dung mass. In the following 7 days, the C/N ratio in the late dung mass and brood ball decreased further, which caused a significant increase in nutritional quality. In addition, in a different type of dung (horse dung), the presence of processing by the females caused a similar increase. This study is the first to demonstrate that (in species that exert extra effort in this regard) the extra effort of making a dung mass before the brood ball contributes to diet quality to compensate for larval nutrition. This is exciting and indicates that parental care modifies the dung microbiota to achieve this feat, as discussed in next subchapter.

4.2. Diet Quality Changes with Microbial Communities

Studies in dung beetles have suggested that the bacterial community of the female gut that is transmitted to the offspring through the brood balls compensates for the nutritionally incomplete diet of larvae and helps with cellulose degradation [19,20]. This study revealed that the change in bacterial and fungal communities occurs through the nesting processes in C. acutidens, suggesting that, rather than the bacterial community, the fungal community of the female gut is transmitted to the larval diet. In addition, from the perspective of diet quality from fungal communities, the proportion of Trichosporonaceae in the fungal community tended to increase with decreasing C/N ratio in both breeding experiments, irrespective of dung type. Trichosporonaceae species have been reported to assimilate or degrade lignocelluloses [23,24,25,26,27,28,29,30,31]. These facts suggest that Trichosporonaceae, which accounted for a high proportion in the female gut, is likely to be transmitted to the larval diet and increase the diet quality of the larvae via the ability of symbiotic fungi to assimilate or degrade lignocelluloses. Furthermore, the most identified genus from the female gut and larval diet was Apiotrichum (which has recently been used as a new genus name for members of the Brassicae, Gracil, and Porosum clades of the phylogenetic tree in old Trichosporon [32]), in which A. scarabaeorum accounted for 36% in the female gut and the majority in the larval diet. The cultivation from the larval gut also detected this species. These results suggest that at least this particular species is a gut symbiont of both the adult and larval C. acutidens, which is likely to be transmitted to the larval gut through brood balls. In contrast, the cultivation from the female gut detected A. siamense, which is from the same Brassicae clade and has the same ability to assimilate cellobiose and D-Xylose as A. scarabaeorum [33], yet it was found to not play a major part in the larval diet. We could not clarify whether these are facts because the registration of species level is not enough in the database for NGS amplicons or due to this species not being major species in the female gut.

The remaining question about the changes in the C/N ratio in this study is reflected in the fact that the artificial early dung mass also tended to decrease the C/N ratio to a similar level as the early dung mass, despite there being no female processing. Furthermore, the bacterial community in the artificial early dung mass revealed the closest relationship to the brood ball. Considering that there is no evidence in this study that the bacterial community of the female gut was transmitted to the larval diet, this seems to indicate that only the physical impact of dung processing, despite the presence and absence of female processing, was important for altering the bacterial community in this early phase, so the decrease in the C/N ratio in both the early dung masses could be caused by bacteria before the increase in potential symbiont fungi [34].

Fabre described the making of dung mass before dividing brood balls in Copris as similar to “bread yeast fermentation [16]. Halffter and Edmonds (1982) assumed that the accelerated anaerobic fermentation of dung mass was due to the compaction by and coating of thin layers of soil in two Copris species [14]. In contrast, the detailed observations of most Copris species, including C. acutidens, suggest that females coat brood balls with soil but not the dung mass because the thin layer of soil became thick as the larva grew [15,21]. In this study, the late dung mass and brood ball were more anaerobic than the artificial ones but aerobic activity was maintained by Pseudomoadota (Proteobacteria), which accounted for half the proportion of the bacterial community. Our findings support the assumption of the role of dung mass as a “fermentation device”, meaning “fermented food”. This is better food for the species, as suggested by Fabre rather than Halffter and Edmonds. More surprisingly, we also support the anaerobic conditions during the making processes suggested by Halffter and Edmonds; however, we suggest that it is present in more moderate conditions than previously considered.

As a corresponding study, the study of greenhouse gas emissions from dung pats revealed that cumulative methane fluxes were lower in the nests of C. lunaris than in the nest of other dung beetles up until 15 days, after which it changed to being higher after 20 days [35]. This change was caused by altering a moderate emission velocity from 7–14 days to a rapid increase in emission velocity after 15 days. Thus, the report by Piccini et al. (2017) supports our result of moderate anaerobic conditions during the making of the dung mass [35]. Furthermore, according to Piccini et al. (2017), more accelerated anaerobic conditions could occur during the later period of brood ball care by females [35]. This seems to be supported by the detailed observation that females coated brood balls with soil. Unfortunately, this study did not cover the later period of brood ball care, and the class Methanobacteria (Archaea) was not identified. The overall elucidation of lignocellulose degradation in the larval diet will advance the comprehensive understanding of the nesting behavior of Copris species.

4.3. Parental Care

In this study, we demonstrated that extra effort in making a dung mass before the brood ball contributes to diet quality and compensates for the lack of larval nutrition in C. acutidens. This is likely to be associated with the ability of possible symbiont Trichosporonaceae fungi to assimilate/degrade lignocelluloses. This extra effort is common in the tribe Coprini [14]. In addition, A. hepliocopridis (Trichosporonaceae) is isolated from the gut of the beetle Heliocopris bucephalus [31]. A. hepliocopridis (of the same Barassicae clade as A. scarabaeidae) is a unique species that does not assimilate cellobiose and D-Xylose, as well as urease, unlike the other major species of the Barassicae clade. Dung beetles generally utilize mammalian dung. Large Heliocopris species, in particular, prefer the larger dung pats of large mammals. Such large pats will also contain a large amount of mammalian urine. Although we do not know whether A. hepliocopridis is major species in the larval diets of H. bucephalus, instead of reducing the carbon source, such as A. scarabaeorum, preserving and not decomposing urea as a nitrogen source might be two sides of the same coin when increasing the proportion of nitrogen content, and both strategies might have been favored by natural selection to improve the quality of food for larvae. Further study is expected to test whether this association with Trichosporonaceae fungi is common within the genus Copris.

Finally, we would like to add that our findings that the extra effort of nesting behavior by females is likely to be associated with symbiont fungi might not be completely unrelated to the subsocial nesting behavior of C. acutidens. All females in the Copris species stay in the nest until the offspring emerge as adults [14]. When there is no female parental care, the larval survival rate decreases remarkably [36,37]. This has been considered to be because females add soil coats onto brood balls, repairing fractures in the brood balls and protecting them from desiccation, fungal infection, predators, and parasites [14,37]. Here, we suggest two hypotheses from our findings. First, the increases in the cost of female effort toward their offspring can simply lead to further generous protection over the survival of offspring through other means. Most dung beetles directly make brood balls from dung, but Copris females use time and energy in the extra effort preparing a dung mass, not only preparing the dung mass but also keeping it as is for approximately a week. While keeping it in this state, the females work hard around the dung mass [21]. In contrast, the males often leave the nests between the time the females take to finish making a dung mass and finishing the brood balls. Second, parental care in insects can coevolve with certain microbiota in both competitive and mutualistic relationships as an ecological factor of evolution [38,39]. As already mentioned in this study, we revealed that the extra effort of nesting behavior by females is likely to be associated with symbiont fungi. Although it is yet determined how this association with potential symbiont Trichosporonaceae fungi continues through nesting behavior, it seems to be worth testing the relationship between the subsocial behavior of C. acutidens and the Trichosporonaceae fungi as one hypothesis.

5. Conclusions

This study is the first to demonstrate that extra effort expended in making a dung mass before making the brood balls improves the nutritional value of the larval diet. From the microbial community analyses, we suggest that rather than the bacterial community, the fungal community of the female gut is transmitted to the larval diet. Trichosporonaceae fungi, which accounted for a high proportion in the female gut, are likely to be transmitted to the larval diet and increase the quality of the diet of the larvae via the ability to assimilate or degrade lignocelluloses. At least, A. scarabaeorum of Trichosporonaceae is a gut symbiont of both the adult and larvae C. acutidens. Our findings support the assumption of “fermentation device” or “external rumen” roles for making the better food. And it is noted that the anaerobic conditions during the making process of a dung mass occur more moderately than has been previously considered.