The effects of social environment and the metapleural gland on disease resistance in acorn ants

Eusocial species differ in living conditions when compared to solitary species primarily due to the dense living conditions of genetically related individuals. Consequently, these crowded conditions can induce a high rate of pathogen transmission and pathogen susceptibility. To resist an epidemic, individuals rely on sets of behaviors, known as social immunity, to decrease pathogen transmission among nestmates. Alongside social immunity, ants have a pair of secretory metapleural glands (MG), thought to secrete antimicrobial compounds important for antisepsis, that are believed to be transferred among nestmates by social immune behaviors such as grooming. To investigate the effects of social immunity on pathogen resistance, we performed a series of experiments by inoculating acorn ants Temnothorax curvispinosus with a lethal spore concentration of the entomopathogenic fungus Metarhizium brunneum. After inoculation ant survival was monitored in two environments: solitary and in groups. Additionally, the MG role in pathogen resistance was evaluated for both solitary and grouped living ants, by sealing the MG prior to inoculations. Individuals within a group environment had a higher survival compared to those in a solitary environment, and individuals with sealed glands had significantly decreased survival than ants with non-sealed-MG in both solitary and social environments. We observed the lowest survival for solitary-sealed-MG individuals. Although sealing the MG reduced survival probability, sealing the MG did not remove the benefits of grouped living. We show here that social living plays a crucial role in pathogen resistance and that the MG has an important role in pathogen resistance of individual T. curvispinosus ants. Although important for an individual’s pathogen resistance, our data show that the MG does not play a strong role in social immunity as previously believed. Overall, this study provides insights into mechanisms of social immunity and the role of MG in disease resistance.


Introduction 52
There are many benefits of social living, such as shared resource and protection from predators, 53 but the inherent colony structure could also be thought to increase pathogen spread among individuals [1].

54
Historically, the infection rate across a group of individuals is assumed to be linearly dependent on the 55 density of hosts and pathogens [2]. Pathogen transmission is increased with contact, and this would 56 certainly be true for dense colonies, especially observed in the insect order Hymenoptera, where 57 individuals are constantly in contact by exchanging food and information [3,4]. Furthermore, a high 58 interaction rate among genetically related individuals has the potential to increase both the spread and 59 virulence of the pathogen [1,5]. However, eusocial insects such as ants, bees, termites and wasps have an 60 ecologically high success rate by inhabiting many types of ecosystems, and often dominate their 61 surrounding environments [6,7].

62
The large prominence of social insects must indicate evolutionary adaptations to the increased , or within their nests, such as plant sap in a form of propolis in honey bees [11,22]. Along with 72 pathogen transmission prevention, social insects can provide nestmates with contact immunization, in 73 which a non-infected individual receives the pathogen to increase resilience in potential future pathogen 74 contact [23,24,25]. Similarly, pharaoh ants prefer a pathogen-exposed nest over noninfected nests,

75
increasing the immune priming of the whole colony [26]. The presence of social immunity is thought to have arisen to maximize inclusive fitness of each individual, opening many altruistic traits and actions 77 that are vital for the colony as a whole and are comparable to an organism level immune system [27].

II. Survival in groups vs. solitary environment 143
Six single queen colonies containing over 110 workers were cooled at 4°C until movement 144 slowed; colonies were then placed on an ice pack while 110 workers were randomly selected from each 145 colony into colony specific Fluon-coated petri dishes with water and diluted honey.

146
Ants from each colony were divided into two inoculation treatments: 1. M. brunneum 147 conidiospore inoculated (1 x 10 8 conidiospores/mL) (n=48 per colony), and 2. sham inoculated (0.05% Triton-X solution) (n=62 per colony). For each ant in each treatment, the ant was submerged within the 149 corresponding solution for 5 seconds and set upon a paper towel to dry. Immediately after drying, the ant 150 was placed into a microcentrifuge tube that was then capped with moist cotton. Ants remained in 151 individual microcentrifuge tubes for 24 hours.

152
After the 24-hour isolation period, ants from each inoculation treatment were divided into social Three single queen colonies containing over 120 workers were sedated using carbon dioxide until 163 movement slowed; 120 workers were randomly selected from each colony into colony specific Fluon-

183
Ants from each sealing treatment were then divided into two inoculation treatments: 1. M. conidiospore inoculated and non-sealed ants with eight sham inoculated and gland-sealed ants. All ants were moved to the respective tubes and moist cotton was used to seal the tubes and to provide water to the 198 ants. Survival was checked and recorded daily for a total duration of 12 days post-inoculation.

212
Dosage response and immune priming 213 The M. brunneum conidiospore concentration treatment had an effect on the survival of exposed 214 ants (See supplemental). The lowest pathogen concentration tested was 10 5 conidiospores/mL that had no 215 significant difference on ant survival compared to sham inoculated ants (χ2 = 1.05, d.f. = 1, p = 0.305).

223
Effects of social grouping on survival 224 Social environment vs. host isolation (solitary) had a significant effect on the survival of ants that 225 were inoculated with M. brunneum (Fig 1). Furthermore, high survival rate in a control treatment with a

Effects of metapleural gland on survival 246
To validate our experimental protocol, ants were subjected to CO 2, a glue treatment on their back, 247 or sham inoculation, and all ants were kept in individual tubes by themselves. These two treatments did 248 not differ from ant's survival that were only sham inoculated (CO 2 : χ 2 = 0, d.f. = 1, p = 1.000; Glue: χ 2 = 249 3.140, d.f. = 1, p = 0.076) (Fig 2). The survival probabilities at the end of twelve-day experiment, of sham 250 inoculated ants with no further treatment with CO 2 and glue was as follows: sham 1.00, CO 2 1.00, and

281
Groups in which sham and pathogen exposed ants with differing sealing treatments were mixed did not functional MG is necessary for individual survival but not so much for the nestmate survival (Fig 3). probability than species with functional MG glands when exposed to a pathogen, and in addition appear

350
to have an increase in grooming behavior [36].

351
Overall, our data suggest that group living is central for pathogen resistance in T. curvispinosus