Heat production in carrion biofilm formed by communally

Insects regulate their body temperature mostly behaviorally, by changing posture or microhabitat. These strategies may be ineffective in some habitats, for example on carrion. Carrion beetles create a biofilm-like matrix by applying to cadaver surface anal or oral exudates. We tested the hypothesis that biofilm formed by communally breeding Necrodes littoralis L. beetles (Silphidae) produces heat, that enhances beetle fitness. We demonstrated that heat produced in the biofilm is larger than in meat decomposing without insects. Beetles regularly warmed up in the biofilm. Moreover, we provide evidence that biofilm formation by adult beetles has deferred thermal effects for larval microhabitat. We found an increase in heat production of a biofilm and a decrease in development time and mortality of larvae, after adult beetles applied their exudates on meat. Behavioral strategy revealed here for N. littoralis is basically a new form of thermoregulation and parental care in insects.

Temperature is a key component of animal environment. Insects usually use external 32 heat to regulate their body temperature (1, 2). By changing body orientation or selecting 33 microhabitat with specific thermal characteristics, insects may maintain their body 34 temperature within thermal optima (3,4). These mechanisms may be ineffective at certain 35 developmental stages (e.g. larvae) and in some microhabitats (e.g. carrion or dung), indicating 36 that other thermal options may be important for some insect ectotherms. 37 Carrion is an example of a "bonanza" resource, i.e. very rich but at the same time 38 scattered and ephemeral (5). There is severe competition between microbes, insects, and 39 vertebrates over carrion resources (6)(7)(8)(9)(10)(11)(12). Insects, e.g. blow flies (Calliphoridae) or carrion 40 beetles (Silphidae) use carrion for breeding and their larvae are main carrion reducers in 41 terrestrial environments (11,13,14). Necrophagous larvae usually feed in aggregations (15).

61
Necrodes beetles colonize large vertebrate cadavers, where its larvae feed on carrion 62 tissues and under favorable conditions may reduce them into dry remains (44,45). Beetles 63 start visiting carrion after it becomes bloated (under summer temperatures usually 4-8 days 64 after death) and after some time many of them (even hundreds) may be present on a cadaver 65 (44,(46)(47)(48). Females lay eggs in a nearby soil and larvae abundantly colonize carrion usually 66 during late decomposition (44,46,47). Necrodes beetles are members of Silphinae and, in 67 contrast to Nicrophorinae, they colonize large carrion, breed communally and reveal no 68 parental care (46,49,50). 69 We demonstrate that a biofilm-like matrix formed on meat by adult or larval Necrodes 70 littoralis L. (Silphidae) produces heat, which is larger than heat emitted by meat decomposing 71 without insects. Beetles regularly warm up in the matrix. Consequently, smearing carrion with 72 exudates may be categorized as a novel mechanism for insect thermoregulation. Moreover, 73 when adult beetles applied exudates, heat produced in a biofilm formed by feeding larvae was 74 larger, resulting in a decrease of larval development time and mortality. Carrion manipulation 75 by adult beetles had deferred thermal effects, that benefited beetle offspring, and this is 76 basically a new form of parental care among carrion insects.

81
Heat production in carrion biofilm formed by Necrodes littoralis. 82 To test if biofilm produced by carrion beetles (Fig. 1) generates heat, we monitored 83 with thermal imaging camera conditions in colonies of adult and larval N. littoralis 84 subsequently feeding on a meat resource (M+A+L) and in colonies of larval beetles only 85 (M+L), and compared them against thermal conditions in equivalent colonies but without 86 beetles (M). We found that adult and larval beetles smeared meat with anal exudates, forming 87 greasy, biofilm-like coating on meat and surrounding soil, the surface of which increased with 88 colony age (Fig. 1). The biofilm had a higher temperature than the background (Fig. 2) and 89 beetles, particularly larvae, were regularly warming up on its surface (Fig. 3).

90
By quantifying the average temperature of biofilm-covered surface (meat or soil) and 91 meat without beetle-derived biofilm (M), we found that heat emission in the biofilm enlarged 92 with colony age and was significantly larger than heat produced in meat alone (rmANOVA 93 for the average heat production in the pre-larval phase, F 1,27 =122, N=30, P<0.001, Fig. 4a, b).

94
Heat of the biofilm during larval feeding phase significantly increased after adult beetles 95 prepared the meat (rmANOVA for the average heat production in the larval feeding phase, 96 F 1,27 =154.8, N=30, P<0.001, Fig. 4c). Moreover, heat production in larval feeding phase 97 became larger, when we extended the pre-larval phase (rmANOVA for the average heat 98 production in the larval feeding phase, F 2,27 =24, N=30, P<0.001, Fig. 4a, c).

99
To test the effect of environmental temperature on heat emission in a biofilm, we   To test if biofilm formed on meat by adult beetles affects fitness of larvae, we  Fig. 6b). Larval mortality was significantly lower in colonies 116 with meat prepared by adult beetles (rmANOVA for larval mortality, F 1,27 =8.3, N=30, 117 P<0.01), differences between M+A+L and M+L trials were the largest in case of the shortest 118 pre-larval phase (Fig. 7a). Increase in the duration of pre-larval phase enlarged larval 119 mortality (rmANOVA for larval mortality, F 1,27 =3.5, N=30, P=0.045, Fig. 7a). Postfeeding 120 larval mass was significantly larger when adult beetles prepared the meat, but only for the 121 shortest pre-larval phase (Fisher LSD post-hoc test, P<0.001, Fig. 7b). We demonstrated that Necrodes beetles form a biofilm-like matrix on carrion in which 125 heat is produced and beetle larvae warm up while feeding. This thermoregulation strategy 126 probably involves microbial activity. Carrion biofilm is a mixture of microbes, their 127 metabolites, and compounds released into the matrix directly by beetles or through carrion 128 digestion (24, 26). Microbes, e.g. bacteria involved in composting (51), are known to produce

137
Heat produced in a biofilm formed by adult and larval carrion beetles may benefit 138 them in several, non-mutually exclusive ways. By increasing larval development rate, heat 139 may reduce the time that larvae spend on carrion. Current results support this interpretation.

140
When we compared treatments with and without application of exudates by adult beetles, we 141 found that heat production during larval feeding phase was larger and larval development 142 times were shorter after application of exudates by adult beetles. Shorter larval development 143 times were found also for Nicrophorus mexicanus (53) and Nicrophorus vespilloides (28) in a 144 full care or prehatch care conditions (i.e. with exudates application by adult beetles) as 145 compared to the no care conditions, and for Nicrophorus orbicollis (54) in the biparental care 146 conditions (i.e. more exudates) as compared to the uniparental care conditions. Although 147 temperature of a biofilm was not measured in these studies, and we do not even know if there 148 is any heat production in the biofilm formed by Nicrophorus beetles, it is tempting to 149 hypothesize that decrease in larval development time resulted from larger heat production of 150 the biofilm after application of exudates by adult burying beetles.

151
Benefits of beetles from heating carrion may be less obvious. Heat produced in the 152 biofilm may be an external and social mechanism of immune response to entomopathogenic 153 microbes (55-58). Recent studies of Nicrophorus beetles highlighted the importance of beetle-154 microbe interactions, by indicating that smearing carrion with beetle exudates may simply 155 suppress their microbial competitors (24,26,28,29,31). Similarly, anal exudates of Necrodes 156 surinamensis were found to have a negative effect on bacterial survival (37). Although previous studies linked the suppression with the production of antimicrobial compounds (24, 158 26, 29, 38, 39, 41), the heat of the biofilm may act synergistically with these compounds.

159
There is evidence, from other insect groups, that heat may act in this way. A wax moth larvae 160 Galleria mellonella (Lepidoptera: Pyralidae), colonizing dead or weak honeybee colonies, 161 were found to elevate temperature when aggregated inside beehives (4, 59). A recent study 162 showed substantial mortality of Galleria larvae after infection with Metarhizium fungi at 163 24°C, whereas at 34°C a 10-fold higher dose of the fungus was necessary to reach similar 164 mortality rate, indicating that temperature elevation by communally feeding Galleria larvae longer this phase, the more time microbes have to multiply and produce harmful metabolites, 171 and the larger is the probability that beetle survival will be affected.

172
Heat in a carrion biofilm may also support the growth and activity of mutualistic All experiments were performed in rearing containers with a volume of 1.5 l (initial 215 experiments and experiment 1) or 3.5 l (experiment 2) with about 5 cm layer of humid soil.

216
On one side of a container meat was placed, and on the opposite side, we put wet cotton wool 217 to maintain high humidity and water for the beetles (Fig. 2). To prevent meat from drying out 218 and to mimic skin cover below which N. littoralis usually feed and breed on carrion, we 219 covered the setup with aluminum foil.

221
Thermal imaging and temperature quantification in a biofilm-covered surface 222 We monitored temperature conditions inside beetle colonies using thermal imaging 223 camera Testo 885-2 with 30° x 23° infrared lens (Testo, Germany), mounted to a tripod while 224 making images. Images were made in a room temperature and humidity (20-23°C, 50-60%), 225 with containers taken outside of the temperature chamber and images made within no more 226 than 2 minutes.

227
Based on initial tests with pork meat, all temperature measurements were made with 228 the emissivity set on 0.8 and reflected temperature set on 17°C. Heat production of the 229 biofilm-covered surface was defined as a difference between the average temperature of the 230 surface and the background surface temperature. To obtain background temperature, we 231 measured the average temperature of a rectangular soil area located in the upper, meat-232 opposite side of a container, for the first three days starting from the establishment of a 233 colony. During these initial days, biofilm temperature was usually only a little higher than 234 background temperature, so heat production in the biofilm had negligible effects on the 235 background temperature. Therefore, by choosing the most distant part of a container and by 236 quantifying average soil surface temperature there, we obtained the best measure of the 237 background surface temperature. As a reference against which heat production in the biofilm 238 was quantified, we used grand mean calculated from average background surface 239 temperatures measured during the first three days in each container. Because biofilm-covered 240 surface enlarged with colony age, and at some time the biofilm covered also soil surrounding 241 meat, it was impossible to quantify heat production always in the same place and within the 242 same surface area. Instead, we quantified the average temperature in a largest possible, 243 circular or ellipsoidal area, depending on the shape of the surface covered with biofilm. We 244 used for this purpose the average temperature calculation tools from IRSoft 4.5 software 245 (Testo, Germany). To determine the optimal number of beetle larvae and the optimal quantity of meat, we 249 performed several initial trials. By comparing heat production in larval colonies of different 250 abundance (10, 20, 30, 40, 50 and 100 larvae), we found that normal growth and clear heat 251 production were already present in colonies of 40 larvae (Fig. 8). Accordingly, we used this 252 number in main experiments. During initial trials, we also tested setups with different quantity 253 of meat and decided to use 2 g per larva in Experiment 1 and 3.5 g per larva in Experiment 2.

254
In Experiment 2 we studied effects of heat production on larval fitness, and to maintain 255 optimal growth of larvae, we decided to provide them with more meat than we used to study 256 heat production only (Experiment 1).

258
Environmental temperature and heat production in carrion biofilm (Experiment 1) 259 To test the effect of environmental rearing temperature on biofilm heat production, we pork. In M+A+L containers 10 adult beetles (sampled at random from our main colony, sex 298 ratio 1:1) were kept for the duration of 3, 4 or 5 days, depending on the treatment. Afterward, 299 meat prepared by adult beetles was transferred to a new container and 40, sampled at random, 300 freshly hatched first instar larvae were added. We had to transfer meat to new containers, 301 because adult beetles usually oviposited during the pre-larval phase and it was difficult to 302 control number of larvae in the containers. In M+L containers meat was decomposing without 303 insects for the duration of 3, 4 or 5 days, depending on the treatment, and then was transferred (10 replicates). Once a day colonies were taken out of a chamber to make thermal images.

308
After larvae stopped feeding and began to bury themselves, we counted and weighed 309 them (laboratory scale AS 82/220.R2, Radwag, Poland). Then, they were transferred to soil-   Longitudinal thermal profiles (a), differences in the average heat production of a biofilm 498 during pre-larval phase (b) and larval feeding phase (c) between Necrodes littoralis colonies 499 differing in the presence of adult beetles and in the length of pre-larval phase. All colonies 500 were reared at constant 23°C. As a no-beetle reference in (a), we used results for 23°C from 501 environmental temperature experiment. Thermal profiles were fitted to the data using the 502 distance-weighted least-squares smoothing procedure. Symbolsmeans, whiskers -95% 503 confidence intervals, different letters denote significant differences in pairwise comparisons.