The Rolling of Food by Dung Beetles Affects the Oviposition of Competing Flies

Flies are the main competitors of dung beetles for oviposition sites and rolling dung beetles relocate their food to reduce interspecific competition. Furthermore, dung beetles deposit chemical substances on the food ball that may repel fly larvae and certain predators. In the present study, using Deltochilum furcatum, a dung beetle that does not exhibit parental care and the blow-fly, Lucilia cuprina, we tested the hypothesis that pygidial secretions deposited on the food ball could also make it less attractive as an oviposition site for flies. Food balls rolled by either D. furcatum males or females received significantly fewer eggs that balls that had not been rolled by beetles. Also, flies laid significantly fewer eggs on food balls treated with secretions collected from male pygidial glands. Reduced fly oviposition may be a direct effect of compounds the beetles deposited, acting as an allomone, and/or an indirect negative effect on the microbial community that stimulates fly oviposition. A model of the reproductive biology of this species is proposed.


Introduction
Coprophagous beetles use dung and/or carrion as a food and nesting resource [1,2]. However, these nutrient-rich resources are ephemeral both in time and space, resulting in intense intra and interspecific competition for food [3]. Dipterans are the main interspecific competitors [2][3][4][5][6], so coprophagous beetles have evolved relocation strategies whereby resources are cut up and rolled to less accessible locations [1].
Favila [7] demonstrated that fresh male secretions, probably from the fifth abdominal sternite, were attractive to conspecific females, and secretions from different glands in Canthon species serve to deter competitors and predators [5,8], leading Cortez et al. [9] to suggest that these secretions may serve multiple purposes. This idea is supported by studies on other dung beetles: for example, male secretions from glands situated on the first abdominal sternite of Kheper species are involved in mating [10,11], while adults of Oniticellus egregius (Klug, 1855) produce secretions from the lateral edges the anterior abdominal segments that play a role in defence [12,13].
In the present study we carried out experiments to test the hypothesis that chemicals deposited during the rolling process by Deltochilum furcatum (Castelnau, 1840), a species strongly attracted to carrion [14], would reduce the ball's attractiveness as an oviposition site for Lucilia cuprina (Wiedemann, 1830) (Diptera: Calliphoridae), as it also uses carrion as an oviposition site [15]. The secretions deposited by either sex during rolling reduced oviposition by L. cuprina, as did the mechanical application of male pygidial glands.

Biological Samples
The D. furcatum adults used in these experiments were collected using sardine baited pitfall traps set out in a fragment of a seasonal semideciduous Atlantic forest (20 • 45 22 S/42 • 51 44 W), near Viçosa, Minas Gerais State, Brazil. The beetles were separated by sex, held in 1000-mL plastic pots containing humid soil and fed fresh meat or sardine once a week.
A colony of L. cuprina was established from maggots collected on meat left out in the field for several days on the campus of the Federal University of Viçosa. Adult flies were fed a mixture of honey and water, as well as fresh beef blood as the protein source for egg formation [15]. Pieces of meat placed in the cage served as oviposition sites and the eggs then transferred to 250-mL plastic containers (about 100 eggs/container) with milk powder diluted in water as a larval food source. Both species were reared at 27 ± 1 • C, 55% RH, 12L:12D.

Behavioural Analysis
To determine the peak period of oviposition by L. cuprina, 10 pairs of 13 day-old flies were provided a 3-g ball (Ø = 19 mm) of fresh hamburger meat, similar in size and shape as those formed by beetles. The number of eggs were counted every hour throughout the photophase, as pretests showed that no eggs were laid during the scotophase. The experiment, performed at room temperature, had 16 replicates per hour (see Supplementary Materials File 1) (Figure 1). secretions deposited by either sex during rolling reduced oviposition by L. cuprina, as did the mechanical application of male pygidial glands.

Biological Samples
The D. furcatum adults used in these experiments were collected using sardine baited pitfall traps set out in a fragment of a seasonal semideciduous Atlantic forest (20°45′22′′ S/42°51′44′′ W), near Viçosa, Minas Gerais State, Brazil. The beetles were separated by sex, held in 1000-mL plastic pots containing humid soil and fed fresh meat or sardine once a week.
A colony of L. cuprina was established from maggots collected on meat left out in the field for several days on the campus of the Federal University of Viçosa. Adult flies were fed a mixture of honey and water, as well as fresh beef blood as the protein source for egg formation [15]. Pieces of meat placed in the cage served as oviposition sites and the eggs then transferred to 250-mL plastic containers (about 100 eggs/container) with milk powder diluted in water as a larval food source. Both species were reared at 27 ± 1 °C, 55% RH, 12L:12D.

Behavioural Analysis
To determine the peak period of oviposition by L. cuprina, 10 pairs of 13 day-old flies were provided a 3-g ball (Ø = 19 mm) of fresh hamburger meat, similar in size and shape as those formed by beetles. The number of eggs were counted every hour throughout the photophase, as pretests showed that no eggs were laid during the scotophase. The experiment, performed at room temperature, had 16 replicates per hour (see Supplementary Materials File 1) (Figure 1). To determine the effect of D. furcatum rolling on subsequent fly oviposition, either a male or female beetle was randomly taken from the colony and left in a plastic arena with a meatball for 24 h. Control balls were held for 24 h under the same conditions but in the absence of a beetle. Subsequently, we placed a rolled and a control ball in a 0.5 m 3 organza-covered cage during the last 4 h of photophase (peak oviposition period) with forty 13 ± 1 day old L. cuprina males and females (Figure 2A). At the end of each assay (six per sex) the total number of eggs on each ball was counted. To determine the effect of D. furcatum rolling on subsequent fly oviposition, either a male or female beetle was randomly taken from the colony and left in a plastic arena with a meatball for 24 h. Control balls were held for 24 h under the same conditions but in the absence of a beetle. Subsequently, we placed a rolled and a control ball in a 0.5 m 3 organza-covered cage during the last 4 h of photophase (peak oviposition period) with forty 13 ± 1 day old L. cuprina males and females (Figure 2A). At the end of each assay (six per sex) the total number of eggs on each ball was counted. As observed with other Scarabaeinae species [8,12], D. furcatum releases pygidial secretions when disturbed. Therefore, male and female beetles were gently tapped and the secretions extruded collected under a stereoscope using sterilized glass capillary tubes (Micron capillaries glass-tube for micro-hematocrit without heparin). Samples of male and female secretions were held in 200 μL glass tubes at −6 °C until used in bioassays. The surface of two balls was treated with 20 μL of either male or female pygidial secretions and placed in an arena with two untreated balls. The attractiveness of treated and untreated balls as oviposition sites was determined using the same protocol described for beetle rolled vs. control balls above. There were three replicates for male and female secretions ( Figure 2B).

Statistical Analyses
The effect of D. furcatum food rolling and pygidial secretions on the oviposition of L. cuprina were analyzed with a paired t-tests, using R software [16].

Results
The mean number of L. cuprina eggs laid during the 4 h experiment comparing unrolled balls and those rolled by beetles was 389.7 ± 146.3 for male and 665.8 ± 82.2 for female D. furcatum treatments (See Supplementary Materials File 2). Significantly fewer eggs were laid on rolled balls than control balls, regardless of which sex did the rolling ( Figure 3A). As observed with other Scarabaeinae species [8,12], D. furcatum releases pygidial secretions when disturbed. Therefore, male and female beetles were gently tapped and the secretions extruded collected under a stereoscope using sterilized glass capillary tubes (Micron capillaries glass-tube for micro-hematocrit without heparin). Samples of male and female secretions were held in 200 µL glass tubes at −6 • C until used in bioassays. The surface of two balls was treated with 20 µL of either male or female pygidial secretions and placed in an arena with two untreated balls. The attractiveness of treated and untreated balls as oviposition sites was determined using the same protocol described for beetle rolled vs. control balls above. There were three replicates for male and female secretions ( Figure 2B).

Statistical Analyses
The effect of D. furcatum food rolling and pygidial secretions on the oviposition of L. cuprina were analyzed with a paired t-tests, using R software [16].

Results
The mean number of L. cuprina eggs laid during the 4 h experiment comparing unrolled balls and those rolled by beetles was 389.7 ± 146.3 for male and 665.8 ± 82.2 for female D. furcatum treatments (See Supplementary Materials File 2). Significantly fewer eggs were laid on rolled balls than control balls, regardless of which sex did the rolling ( Figure 3A). In the experiment comparing unrolled balls and those treated with either male and female extracts, the mean number of eggs laid was 1145.3 ± 566.5 and 1767.7 ± 720.2 for male and female treatments, respectively (see Supplementary Materials File 3). Significantly fewer eggs were laid in balls treated with pygidial secretions from males but not of females ( Figure 3B).

Discussion
Both male and female D. furcatum adults deposit chemicals on the ball that limit oviposition by L. cuprina. This is very different from the situation reported for Canthon cyanellus cyanellus (LeConte, 1859), where only balls marked by males are less attractive to Calliphora larvae in a choice assay [5]. Obviously the two studies were testing the choice behaviors of different life history stages of the same species, but it seems rather unlikely that this would be the cause of the difference observed in the two studies. We believe that the marked differences in the exploitation of resources between the two beetle species would explain the difference in marking behavior.
In the case of C. cyanellus cyanellus only males mark the ball, but both sexes collaborate in ball rolling and parental care of the brood, which is located in a chamber containing several food balls [5,17]. In contrast, D. furcatum buries one ball at a time and exhibit no parental care. Thus, in the absence of either pair formation or brood care, the most effective means of reducing interspecific competition would be when both sexes marked.
Based our findings, together with observations of adults in a large arena, we believe there are several reproductive scenarios for D. furcatum, that would explain why both sexes produce secretions during the rolling of food that reduce interspecific competition (Figure 4). Adults of both sexes are attracted to the carcass, where immature beetles independently prepare and feed on food balls, while sexually mature individuals could also mate (Figure 4 scenario a). A female mating on the carcass may then independently roll a ball for some distance, oviposit an egg and bury the ball before In the experiment comparing unrolled balls and those treated with either male and female extracts, the mean number of eggs laid was 1145.3 ± 566.5 and 1767.7 ± 720.2 for male and female treatments, respectively (see Supplementary Materials File 3). Significantly fewer eggs were laid in balls treated with pygidial secretions from males but not of females ( Figure 3B).

Discussion
Both male and female D. furcatum adults deposit chemicals on the ball that limit oviposition by L. cuprina. This is very different from the situation reported for Canthon cyanellus cyanellus (LeConte, 1859), where only balls marked by males are less attractive to Calliphora larvae in a choice assay [5]. Obviously the two studies were testing the choice behaviors of different life history stages of the same species, but it seems rather unlikely that this would be the cause of the difference observed in the two studies. We believe that the marked differences in the exploitation of resources between the two beetle species would explain the difference in marking behavior.
In the case of C. cyanellus cyanellus only males mark the ball, but both sexes collaborate in ball rolling and parental care of the brood, which is located in a chamber containing several food balls [5,17]. In contrast, D. furcatum buries one ball at a time and exhibit no parental care. Thus, in the absence of either pair formation or brood care, the most effective means of reducing interspecific competition would be when both sexes marked.
Based our findings, together with observations of adults in a large arena, we believe there are several reproductive scenarios for D. furcatum, that would explain why both sexes produce secretions during the rolling of food that reduce interspecific competition (Figure 4). Adults of both sexes are attracted to the carcass, where immature beetles independently prepare and feed on food balls, while sexually mature individuals could also mate (Figure 4 scenario a). A female mating on the carcass may then independently roll a ball for some distance, oviposit an egg and bury the ball before returning to repeat the process; this she could do repeatedly without a male as she would have an adequate supply of sperm (Figure 4 scenario b). A mature male could prepare a ball, roll it some distance from the carcass and then release sex pheromones to attract a virgin female (Figure 4 scenario c). Following mating, the male returns to the food source and repeat the process to locate additional mates, while the mated female oviposits, buries the ball and then returns to the food source to continue as in scenario b.
Insects 2018, 9, x FOR PEER REVIEW 5 of 8 returning to repeat the process; this she could do repeatedly without a male as she would have an adequate supply of sperm (Figure 4 scenario b). A mature male could prepare a ball, roll it some distance from the carcass and then release sex pheromones to attract a virgin female (Figure 4 scenario c). Following mating, the male returns to the food source and repeat the process to locate additional mates, while the mated female oviposits, buries the ball and then returns to the food source to continue as in scenario b. There are two other possible variations: where a female sits on a ball being rolled by a male but does not mark or help in any way, or when a female is rolling a ball and a male joins her (Figure 4 scenario d). The sex rolling the ball usually buries it, mating may occur, the female oviposits, the sexes separate, and continue independently to repeat any one of the three scenarios described. In all cases, the balls would be susceptible to attack by dipterans, so in the absence of parental care or true collaboration between adults, as with D. furcatum, marking by both sexes would certainly play an important role in reducing interspecific competition.
While balls rolled by either sex reduced oviposition, only male pygidial secretions of showed activity when applied mechanically. There are a number of possible alternate explanations for the difference between the experiments using balls rolled by adult females beetles and those just treated with secretions only. First, the pygidial secretions were collected from adults of unknown age/sexual maturity and these factors are known to affect gland content in dung beetles [2,5,7,18,19]. Second, the There are two other possible variations: where a female sits on a ball being rolled by a male but does not mark or help in any way, or when a female is rolling a ball and a male joins her (Figure 4 scenario d). The sex rolling the ball usually buries it, mating may occur, the female oviposits, the sexes separate, and continue independently to repeat any one of the three scenarios described. In all cases, the balls would be susceptible to attack by dipterans, so in the absence of parental care or true collaboration between adults, as with D. furcatum, marking by both sexes would certainly play an important role in reducing interspecific competition.
While balls rolled by either sex reduced oviposition, only male pygidial secretions of showed activity when applied mechanically. There are a number of possible alternate explanations for the difference between the experiments using balls rolled by adult females beetles and those just treated with secretions only. First, the pygidial secretions were collected from adults of unknown age/sexual maturity and these factors are known to affect gland content in dung beetles [2,5,7,18,19]. Second, the samples were extracted by disturbing the beetles rather than through normal production in the presence of a food ball. Third, the time and mode of application involved in deposition of the secretions by beetles is obviously more complex and longer that for manual application using a syringe. These factors, individually or in combination, could affect the outcome of the different assays and merit further attention. Also, there are a number of exocrine glands located on the abdomens of Deltochilum species [18], which in Canthon sp. are known to serve several functions [8,9], so it is possible that secretions from several different glands may be applied during rolling.
The secretions deposited may act as allomones, directly affecting female fly oviposition behavior. However, during rolling the food may be contaminated with fungi and bacteria that negatively affect the beetle larvae [20] and chemical substances from abdominal exocrine glands help inhibit the growth of these microorganisms [21], including the pygidial glands [9]. Several fly species have been shown to preferentially lay eggs on substrates with bacteria rather than on sterile ones [22][23][24][25][26] as the bacteria may be essential dietary components for developing maggots [25,27] and/or may cause the release essential nutrients from the substrate though enzymatic action [25,28]. Furthermore, female L. cuprina uses volatile compounds released by bacteria developing on the food source when selecting oviposition sites [22]. Thus, it is possible that the beetle secretions modify the microbial community in a way that reduces the suitability of the resource and L. cuprina females avoid oviposition sites that would be suboptimal for the development of their progeny.

Conclusions
The data presented certainly support the hypothesis that glandular secretions placed on the ball by both sexes of D. furcatum during rolling reduces oviposition by L. cuprina and could reduce interspecific competition. However, considerably more research is required to (i) determine the origin; (ii) the composition of the beetle secretions (preliminary analyses of glands suggest lipids and carbohydrates) that deter oviposition; (iii) elucidate persistence of the compounds and the changes that render the resource less attractive as an oviposition site for flies; and (iv) conduct experiments to quantify the benefits accrued from reduced interspecific competition.