Available via license: CC BY 4.0
Content may be subject to copyright.
Injury-dependent wound care behavior in the desert
ant Cataglyphis nodus
Narmin I. Beydizada
Masaryk University: Masarykova Univerzita https://orcid.org/0000-0002-2882-0292
Antonia Abel
University of Wurzburg: Julius-Maximilians-Universitat Wurzburg
Patrick Schultheiss
University of Wurzburg: Julius-Maximilians-Universitat Wurzburg
Erik T. Frank
University of Wurzburg: Julius-Maximilians-Universitat Wurzburg
Research Article
Keywords: Formicidae, helping behavior, survival, social immunity, infection, pro-social behavior
Posted Date: March 7th, 2024
DOI: https://doi.org/10.21203/rs.3.rs-3964303/v1
License: This work is licensed under a Creative Commons Attribution 4.0 International License.
Read Full License
1
Injury-dependent wound care behavior in the desert ant Cataglyphis nodus 1
Narmin I. Beydizada1,2*, Antonia Abels1, Patrick Schultheiss3 & Erik T. Frank1*
2
1Department of Animal Ecology and Tropical Biology, Biocenter, Am Hubland, University of Würzburg, 97074 3
Würzburg, Germany 4
2Department of Botany and Zoology, Faculty of Science, Masaryk University, Kotlářská 2, 611 37 Brno, Czech 5
Republic 6
3Department of Behavioral Physiology and Sociobiology, Biocenter, Am Hubland, University of Würzburg, 7
97074 Würzburg, Germany 8
Narmin I. Beydizada, 0000-0002-2882-0292 9
Patrick Schultheiss, 0000-0002-6906-3507 10
Erik T. Frank, 0000-0002-2066-3202 11
*Corresponding authors: erik.frank@uni-wuerzburg.de, and beydizade.n@gmail.com. 12
2
Abstract 13
Ants often face injuries during foraging, or interspecific competition, elevating infection risk and mortality among 14
the wounded. To avoid this, ants engage in wound care on injured nestmates as a form of social immunity. In this 15
study, we show that Cataglyphis nodus desert ants perform differentiated wound care behavior, depending on 16
wound location and state. Leg-injured ants received significantly more wound care than antenna-injured ants. 17
However, leg wounds with induced infections received barely any wound care from nestmates, leading to similar 18
levels of mortality in isolation and inside the nest. Instead, such leg-infected ants are expelled from the nest, 19
emphasizing the care-kill dichotomy in social immunity. Infections of antennal wounds showed no change in the 20
level of wound care, nor increased mortality. Our results suggest that the level of wound care in ants can be flexibly 21
adjusted to the perceived mortality risk of injuries. Leg injuries pose a greater risk of infection and mortality 22
compared to antennal injuries, likely because of the larger wound area and increased vascular circulation, 23
necessitating intensive prophylactic care to prevent infection. This study is the first to show wound care in 24
Cataglyphis ants despite their short lifespan and offers significant insights into social immunity mechanisms. 25
Significance statement 26
Woun d car e in ant s i nc re as es su rv iv al o f n es tm at es wh ile p re ve nt in g t he r is k o f inf ec tio n tra ns mi ssi on a mo ng 27
colony members. Our study on wound care in thermophilic Cataglyphis nodus ant foragers explores whether this 28
species invests in care for injured foragers and how wound care efficiency varies between injuries at different 29
locations (leg or antenna) or infection status. Here we show that leg-injured ants received significantly more care 30
than antenna-injured ones. However, mortality risks varied significantly based on whether injuries were infected 31
or received nestmate assistance. Sterile leg injuries have a low mortality risk and are given priority for treatment 32
(prophylactic care). Ants with infected leg injuries suffer from increased mortality risk and are expelled from the 33
nest to avoid further spread of infection in the colony. 34
35
Keywords: Formicidae, helping behavior, survival, social immunity, infection, pro-social behavior 36
Introduction 37
Predators are frequently injured while hunting, when competing for mates, or defending their territory, and the risk 38
of injury can be an important foraging or overall fitness cost (Mukherjee and Heithaus 20131). Therefore, animals 39
3
have developed a series of behavioral strategies to recognize, avoid, control, or eliminate pathogens and parasites 40
from their injuries (Hart 1992, 2011). Such strategies include grooming, licking, or applying plant-based 41
antimicrobial materials to the injured area (We st er ga ar d and Fragaszy 1987; Mascaro et al. 2022; Frank et al. 42
2023). While the disinfection or medical treatment of infected wounds has typically been associated with human 43
behavior, some mammals have been observed to engage in wound licking and saliva application (Hart 2011; 44
Kessler 2020). Although many antiseptic compounds and wound-healing proteins have been identified in 45
mammalian saliva, this behavior remains poorly understood and seems to occur regardless of the actual condition, 46
state, or location of the wound (Hart 2011). 47
Social insects also frequently encounter injury risks in many situations, such as during territorial defense, 48
interspecific competition (Porter and Jorgensen 1981), or when hunting dangerous prey (Frank et al. 2017). These 49
events may pose a significant threat of injury, like cut-off extremities (Frank et al. 2017). Open wounds resulting 50
from such events can be one of the major mortality causes in ants (Frank et al. 2017, 2018) because such wounds 51
are likely to get infected (Cremer et al. 2007; Nunn and Altizer 2006; Schmid-Hempel 2017) by opportunistic 52
pathogens on the dirty ground or soil (Frank et al. 2023). In addition, the intensive interactions among group 53
members as well as their low genetic diversity (Alexander 1974; Anderson and May 1979; Côté and Paulin 1995; 54
Schmid-Hempel 1998) make social colonies easy targets for pathogenic transmission and spread (Schmid-Hempel 55
and Koella 1994; Cremer et al. 2018). To prevent this process in a colony, ants engage in social immunity - a 56
collective effort that can be performed through a range of behaviors such as the removal of waste or dead nestmates 57
from the colony (Hart and Ratnieks 2002; Cremer et al. 2007; Wilson-Rich et al. 2009; Diez et al. 2014), social 58
grooming to remove potential pathogenic agents from the cuticle (Büchler et al. 1992; Schmid-Hempel 1998; 59
Tragust 2013), the use of antimicrobials to disinfect substrates and the cuticle (Oi and Pereira 1993) and wound 60
care of injured nestmates (Frank et al. 2017, 2018, 2023). Wound care behavior involves the “licking” of wounds 61
and the direct application of metapleural gland secretions (Frank et al. 2023). In the termite hunting ant 62
Megaponera analis, injured individuals can reach a mortality rate of 80% within 24 hours due to a systemic 63
infection if wound care is inhibited. However, even if only one hour of wound care was provided mortality can be 64
reduced to 10% (Frank et al. 2018). Wou nd c ar e be ha vi or in a nt s is r el at iv el y ra re a nd – apart from Megaponera – 65
has so far only been observed in two other genera, Eciton and Camponotus (preprint, Frank et al. 2023). 66
In this study, we address the question of whether the solitary foraging ant species Cataglyphis nodus performs 67
wound care behavior towards their injured nestmates. Although this species does not hunt dangerous prey, foragers 68
are sometimes injured during aggressive altercations with members of neighboring colonies (personal 69
4
observation). While this species has never been observed to pick up injured nestmates from outside the nest, we 70
can assume that caring behavior may occur inside the nest. The average life expectancy of C. nodus foragers has 71
been calculated to be a mere 6 days, probably due to the extreme harshness of their environment, where ants 72
continue to forage at soil surface temperatures of up to 70°C (Harkness and Weh ne r 197 7; We hn er 19 83 ; Weh ne r 73
et al. 1987). It is therefore vital to maximize the efficiency of foragers to ensure the survival of the colonies. We 74
thus wanted to see if, in a species in which foragers only live for a very short time frame, the colony would still 75
invest in their care. We sp ec if ic all y ai me d t o fo cus o n how the location of the wound affects the behaviors related 76
to wound care. By injuring two different types of extremities (antenna and leg) with both physiological and 77
functional differences we expected to observe a difference in wound care investment and severity of the wound 78
type on mortality with and without care. This is the first report providing evidence of the presence and efficacy of 79
wound care behavior in this genus. 80
Material and Methods 81
The research conducted in this study complies with all relevant ethical regulations and was approved by the Natural 82
Environment and Climate Change Agency of the Greek Government under research permit number 11697. 83
(i) Study animal – laboratory colonies 84
Cataglyphis nodus (BRULLÉ 1833) is a common species in the Eastern Mediterranean shrubland. Like other species 85
of Cataglyphis, it is highly thermophilic and diurnally active, with solitary foragers venturing out in the heat of 86
the day to scavenge opportunistically for dead arthropods. Worker ants show continuous size polymorphism and 87
have long legs, which enable very high running speeds (Sommer and We hn er 2 01 2). In Greece, mature colonies 88
of C. nodus are estimated to harbor 1500 – 2000 individuals, of which around 200 are active foragers at any given 89
day (Weh ne r et al . 1 98 3). Colony density can be high, and the nearest neighbor distance between colonies has been 90
estimated at only 13 m (Wehn er 19 87 ). 91
Natural rates of injury were evaluated at a population of C. nodus in Schinias National Park, Greece, in June 2023. 92
At each of 10 different colonies, 100 or more active foragers were briefly captured to quantify visible injuries to 93
their appendages (legs and antenna). 94
Three colonies of C. nodus were collected in May 2023 from Strofylia National Park, Greece, where this species 95
inhabits coastal open forests of Pinus halepensis and Pinus pinea. All three colonies were located in very sandy 96
5
soil at the edge of the forest. Each colony consisted of many small chambers, distributed over an area of about 97
1 x 1 m and up to 0.5 m in depth, and contained a single queen as well as brood (eggs, larvae, and cocoons). 98
Subsequently, the colonies were kept in a climate-controlled room at the University of Würzburg at 27°C and 40 99
% RH, in artificially created nests (37 x 17 cm, height: 11 cm) with access (connected with a plastic tube, length: 100
10 cm, diameter: 3 cm) to a small foraging arena (19 x 9.5 cm, height: 9 cm) at a 12/12 h light/dark cycle. The 101
walls of the nest container and the foraging area were thinly covered with fluon or paraffin oil to prevent the ants 102
from escaping and covered with a lid that included net-covered openings to ensure air exchange. A humidity 103
gradient was created through a water reservoir on one side of the nest box, which enabled moisture to diffuse into 104
the nest. The colonies were fed every two days with fresh water, diluted honey, and frozen beetle larvae 105
(mealworms). 106
(ii) Wou nd m an ipu la ti on tr ia l 107
To i nve st iga te w he the r an ts are involved in wound care of nestmates and if so, whether they adjust the level of 108
wound care provision depending on the type of injury, we created three treatment groups, two of them were 109
physically injured, either at the leg or antenna and the third was left without injury (healthy control). We fi rs t 110
removed individuals one by one from the foraging area of the colony using soft tweezers and placed them in a 111
glass tube (height: 5 cm, diameter: 2.5 cm). Ants were marked on the thorax with acrylic paint (Wildenhof & 112
Scholz GmbH & Co. KG) for individual recognition. The marked ant was first placed in a sterile Petri dish for 40 113
min. After the paint dried, the ant was prepared for wound manipulation: using sterile scissors one hind leg was 114
severed at the center of the femur segment, ensuring that injuries to the left or right leg were balanced across 115
individuals. Antennal injuries were done in the same way, with the cut placed at the central part of the funiculus. 116
The control ants were handled in the exact same way but did not receive any injury. Three individuals from each 117
of the treatment groups were used per colony (n = 9 per treatment). 118
After manipulation, the ant was re-introduced into its colony immediately, to ensure that the wound remained fresh. 119
The wound care behaviors performed by nestmates toward the focal individual were recorded for 3 hours (with 120
observation periods lasting 10 minutes followed by a 5-minute break) through visual observation. The sequence 121
of treatments (leg injury, antenna injury, healthy control) was randomized. To quantify how injured ants were 122
treated inside the nest by nestmates, we analyzed the following three types of behavior: (a) aggression: when ants 123
show aggression toward the wounded individual, i.e., biting on legs, antenna or abdomen of the injured ant; (b) 124
allogrooming: when ants engage in grooming behaviors with their mouthparts to clean the different parts of the 125
6
body or to remove debris (color-grooming – where the interaction was restricted to the thorax area where acrylic 126
paint had been applied, was not considered as caring behavior); (c) wound care: when nestmates inspect the injured 127
part of the body, placing their mouthparts on the wound or close to the wound (i.e., licking, cleaning) and (d) acid 128
spraying: when nestmates spray acid on the injured individual close to the wounded area. 129
(iv) Survival (isolation) experiment 130
To test if wound care by nestmates reduces the mortality rate of injured ants, we conducted two different survival 131
trials in parallel. In the first assay, 24 individuals from each of the three experimental colonies were removed (in 132
total 3 x 24 individuals) and were marked in the same way as described before (see, (ii) Wound manipulation trial). 133
Following this procedure, 24 were leg-injured; 24 antenna-injured, and 24 left without injury (healthy control). 134
Half of the injured individuals (leg, n = 12 and antenna, n = 12) were then exposed to 10 μL of distilled water with 135
0.005 optical density (OD) of Pseudomonas aeruginosa (hereafter referred to as infected leg and antenna injury), 136
for two seconds on their wounds, following the protocol in Frank et al. (2023). 137
P. a e r u g i n o s a cultures were created from frozen samples kept in Tryptic Soy Broth (TSB) medium with 25 % 138
glycerol (stored and transported at -23°C). After replating the bacterial culture once on a fresh Trypticase soy agar 139
(TSA) plate, we waited 24 hours before applying the pathogen on fresh wounds. 140
The other half of the leg (n = 12) and antenna (n = 12) injured ants were exposed to 10 μL of sterile distilled water 141
(hereafter referred to as sterile leg and antenna injury) for two seconds on their wounds. To c ontr ol f or a ny e ffec ts 142
of this procedure, the leg of the healthy control group was also exposed to sterile distilled water (n = 12). 143
Afterward, each ant was kept in isolation inside a sterile Petri dish (diameter: 5.5 cm, height: 1.5 cm) and fed with 144
diluted honey and water. Experiments were conducted at 25°C. 145
The same experimental procedure was conducted in parallel with the same proportion of individuals (n = 24 per 146
treatment group), but instead of keeping them in isolation they were then released into the nest. Ants were 147
monitored once per hour (in isolation) or every two hours (in the nest) for the next 36 hours. In isolation, if the ant 148
did not react after shaking the dish, the ant was classified as dead. Inside the nest, we defined the death of an ant 149
if the body was completely motionless even after nestmates interacted with it. At the end of the experiment, all 150
ants were reintroduced to their respective colonies. 151
7
(v) Infection assay 152
To o bs erv e if wound care behavior, as well as the survival of ants with a sterile or infected injury, differs 153
quantitatively inside the nest, we maintained four experimental groups (for each group n = 9; 3 per colony): (1) 154
sterile leg-injured ants, (2) sterile antenna-injured ants, (3) infected leg-injured ants, and (4) infected antenna-155
injured ants. All ants were marked for individual recognition, as outlined previously. 156
All focal individuals were then released into the nest to monitor whether the behavioral approach of nestmates to 157
infected and sterile nestmates with different wound types differed over time. Each ant was observed every hour 158
for 1 minute for a total of 4 days. We quantified 6 measures for each observation period: (i) the presence of wound 159
care, (ii) allogrooming, (iii) aggression, (iv) acid spraying, (v) the location at the final stage of observation and (vi) 160
time of death. 161
All three ant colonies used in this study were kept in the experimental laboratory for three months (08-11.2023), 162
during which they underwent the aforementioned behavioral experiments. Only foragers were used for the 163
experiments (selected from the foraging chamber) and no individual was used twice during the experiments. 164
Data analysis 165
For statistical analyses and graphical illustration, we used the statistical software R v4.1.057 with the user interface 166
RStudio v1.4.1717 and the R package ggplot2 v3.3.558. To select the appropriate statistical tests, we tested for 167
deviations from the normal distribution with the Shapiro-Wil ks test (P > 0.05). A Bartlett test was used to verify 168
homoscedasticity (P > 0.05). All our models included colony as a random factor. In case of multiple testing, a 169
Holm-Bonferroni correction was performed with the adjusted P-values given throughout the text. To test for 170
significant differences in the survival curves, we conducted mixed effect Cox proportional hazards regression 171
models (Fig. S2) using the R package survminer (v0.4.9) followed by post-hoc analyses using least square means 172
with the R package lsmeans (v.2.30). The survival curves were illustrated using Kaplan-Meier cumulative survival 173
curves (Fig. 2 and 4a). For the nest leaving time (Fig. 4b) we also used Kaplan-Meier cumulative survival curves 174
for illustration and statistical analysis. For behavioral differences in wound care across injury types (Fig. 1), we 175
conducted linear mixed effect models (LMER) with Individual ID nested within Colony as random factor followed 176
by a Type III Analysis of Variance with Satterthwaite’s method (ANOVA) using the package lme4 (v1.1-27.1) and 177
lmerTest (v3.1-3). For behavioural differences between infected and sterile wounds (Fig. 3), we first pooled all 178
positive interactions (allogrooming, wound care, and acid spraying) for each observation time. We then modelled 179
8
the positive interactions as a binary event using binomial generalized additive models with posthoc contrasts to 180
identify intervals of time during which the probability of receiving care differed between sterile and infected 181
individuals. Reported values refer to the variance, mean ± standard deviation. 182
Results 183
Natural injury rates 184
Under natural conditions, C. nodus foragers had an average injury rate of around 7% (6.6 ± 2.85, n = 10 colonies). 185
Injuries were located on the legs (4.9%), antennae (1.6%) or both legs and antennae (0.2%) (see, SI). Such injuries 186
were most likely received during aggression events between neighboring colonies, as C. nodus avoids most 187
predation and interspecific competition by foraging during the hottest part of the day. 188
Behavioral Ethograms 189
Cataglyphis nodus ants performed wound care behavior toward injured nestmates. Leg-injured nestmates received 190
significantly more care (264.8 ± 166.2 sec, n = 9) compared to antenna-injured ones (21.4 ± 26.1 sec, n = 9; 191
Wilcox test: W = 1, P < 0.001; Fig.1), while the duration of wound care decreased over time in both wound types 192
(linear model: R2 = 0.39, F(3, 212) = 47.4, t = -4.97, P < 0.001). 193
194
Fig. 1. Wound c are based on the injury type in C. nodus. (a) Percentage of time spent on wound care in 10 min 195
intervals over 3 hours with a local polynomial regression (loess) showing a 95% confidence interval fitted for n = 196
9 antenna-injured ants (red, Antenna) and n = 9 leg-injured ants (blue, Leg). LMER: formula= 197
woundcare~time*treatment; Random effect (ID: Colony): Variance=38.29; Std. Dev. = 6.2; Colony: 29.12; Std. 198
***
0
200
400
600
Antenna Leg
Wound type
wound care duration per individual [sec]
colony
C1
C2
C3
0
5
10
15
0 25 50 75 100 125 150
Time [min]
Percentage of Woundcare
Antenna
Leg
ab
9
Dev. = 5.40; Residual: Variance = 239.23; Std. Dev. = 15.47; Fixed effects: Time: t = -1.49, P = 0.14, Treatment: 199
t = 8.00, P < 0.001; Time:Treatment: t = -5.68, P < 0.001. (b) To tal a moun t of ti me spent on wound care for each 200
individual in the first 3 hours after wounding. n = 9 per group (Wilcox test: W = 1, P < 0.001). 201
202
When comparing wound care efficiency between colonies, the colony that consisted of considerably fewer 203
individuals (Colony 3: approximately 568) than the other two colonies (Colony 2 and Colony 1: approximately 204
1119 and 1059 respectively), spent more time on wound care for both leg (454.3 ± 131.9 sec, n = 3) and antenna-205
injured (43.3 ± 37.0 sec, n = 3) nestmates (Fig. 1b; emmeans: C1-C3: t = -2.93, P = 0.028; C2-C3: t = -2.87, P = 206
0.031). The other two colonies invested approximately the same amount of care for leg (170.7.2 ± 55.2 and 169.3 207
± 95.8 sec respectively, n = 3 each) and antenna-injured ants (6.3 ± 1.2 and 14.7 ± 15.0 sec respectively, n = 3 208
each, emmeans: C1-C2: t = -0.06, P = 0.99; Fig. 1b). 209
While overall attention (allogrooming) did decrease over time (ANOVA: F = 170.5, P < 0.001), it did not differ 210
significantly across treatments (ANOVA: F = 2.44, P = 0.094, Fig. S1a). Aggression did not differ significantly 211
across treatments (ANOVA: F = 0.1.38, P = 0.25; Fig. S1b), but there was a significant interaction with time, with 212
increased aggression towards focal individuals upon introduction to the nest, which quickly receded thereafter 213
(ANOVA: F = 14.22; P < 0.001). 214
The use of acid spraying was not a frequent behavior. We could only observe seven cases (for a total of 92 sec), 215
during which the caring ant applied acid either to a nestmate’s wound (more often to leg-injured ants) or close to 216
the wounded part or the full body. We ob se rv ed o ne co nt rol a nt (healthy) receive acid spray, all other acid spraying 217
was directed at injured ants. 218
Survival assay 219
Wound care did not significantly affect survival in antenna-injured ants, which always showed high levels of 220
survival when exposed to pathogens (isolation vs nest infected; least square means: Z = 1.98; P = 0.72) or not 221
(isolation vs nest sterile; Z = 1.13, P = 1; Fig. 2a). In leg-injured ants, infected wounds had significantly higher 222
mortality in isolation when compared to their sterile counterpart (least square means: Z = 3.01, P = 0.04, Fig. 2b). 223
Surprisingly, mortality remained high in leg infected ants even inside the nest, showing no significant difference 224
to the mortality in isolation (least square means: Z = 1.29, P = 1, Fig. 2b). 225
10
226
Fig. 2. Survival probability of sterile and infected antenna-injured (a) and leg-injured ants (b). Kaplan – 227
Meier cumulative survival rates of workers in isolation (dotted line) or inside the nest (solid line) whose wounds 228
were exposed to P. a e ru gi no s a diluted in PBS (Infected, OD = 0.005) or distilled water (n = 12 per group). For 229
detailed statistical results see Fig. S2a. 230
231
Infection assay 232
While antennal injuries showed no difference in care between infected and sterile injuries (Fig. 3a), infected leg 233
injuries received significantly less care between 3 and 22 hours when compared to sterile leg injuries (hierarchical 234
generalized additive model: P < 0.05; Fig. 3b). Instead, infected leg injured ants tended to move sooner to the 235
foraging arena when compared to the other groups (cox mixed effect model: Z = 2.43, P = 0.015, Fig. 4b) and 236
showed higher mortality (Fig. 4a, detailed statistical results in Fig. S2b), thus further corroborating our 237
observations from the previous survival assay (Fig. 2b). 238
+
+
+
+
0
010 20 30 40 50 60 70 80 90
Time [h]
25
75
100
Survival probability [%]
50
Leg injury
Nest
Isol ati on
Sterile
Infec ted
Healthy
+
+
+
0
010 20 30 40 50 60 70 80 90
Time [h]
25
75
100
Survival probability [%]
Nes t
Iso lat ion
Steri le
Inf ect ed
Hea lt hy
50
Ant e nn a i nj ury
a b
11
239
Fig. 3. Positive interactions received by ants with sterile and infected wounds across injury types. A) 240
Probability of receiving care over 24h fitted with a hierarchical generalized (binomial) additive model (HGAM) 241
for antennal injuries (a) or leg injuries (b); shaded bars indicate periods during which the probability of receiving 242
care was significantly (P < 0.05) higher for sterile (yellow, n = 9) than infected (green; n = 9) individuals. The 243
line represents the predicted probability by the HGAM, with the colored shaded area representing the 95% 244
confidence interval. 245
246
Fig. 4. Survival and nest leaving time of ants with sterile and infected wounds across injury types. (a) Kaplan 247
– Meier cumulative survival rates of workers inside the nest whose wounds were exposed to P. a e r u g i n o sa diluted 248
in water (Infected, OD = 0.005) or distilled water (n = 9 per group). For detailed statistical results see Fig. S2b. 249
(b) Percentage of focal individuals inside the nest over time for the same individuals as in Fig. 4a. 250
0.00
0.25
0.50
0.75
1.00
0 4 8 12 16 20 24
Time [h]
Probability of receiving care
0.00
0.25
0.50
0.75
1.00
0 4 8 12 16 20 24
Time [h]
Probability of receiving care
infected
sterile
a b
+
+
++
0
25
50
75
100
0 10 20 30 40 50 60 70 80 90
Time [h]
Survival probability [%]
Treatment
+
+
+
+
Antenna − Infected
Leg − Infected
Antenna − Sterile
Leg − Sterile
++
++
+
+ +
+
0
25
50
75
100
0 10 20 30 40 50 60 70 80 90
Time [h]
Percentage inside the nest [%]
Treatment
+
+
+
+
Antenna − Infected
Leg − Infected
Antenna − Sterile
Leg − Sterile
a b
12
251
Discussion 252
In this study, we show that Cataglyphis nodus ants can display wound care behavior toward injured nestmates, but 253
the amount of care given to the injured individual depends on the state and location of the wound. While sterile 254
leg injuries received more care than antennal injuries, infected leg injuries were barely treated. Instead, the injured 255
individuals were isolated in the foraging arena until death. Interestingly, this was not the case in infected antennal 256
injuries, which received similar levels of care to sterile injuries, suggesting that the risk of systemic infections is 257
not equal across wound types. 258
Woun d ca re be ha vi or an d sur vi va l 259
To as sess h ow w ou nd ca re be havi or va ries d epen ding o n th e type o f wou nd in i njur ed an ts, we cre at ed tw o type s 260
of wounds: one on the ant's leg and the other on its antennae. Since these extremities differ physiologically and 261
functionally from each other (Hillyer and Pass 2020), we expected wound care to differ depending on the injury 262
type. Different types of wounds can lead to differing responses in nestmates depending on the mortality risks 263
involved. A recent study in Camponotus floridanus demonstrated that infected tibia injuries carried a far greater 264
mortality risk than femur injuries and subsequently received more attention from nestmates (Frank et al. 2023 265
preprint). Furthermore, the type of injury could affect the usefulness of the injured individual after recovery, with 266
leg injuries potentially considered to be “light” due to the only slight reduction in locomotive ability, while antennal 267
injuries could more seriously inhibit foraging due to the potential communication or perception errors resulting 268
from the loss of sensory receptors (Gellert et al. 2022). Alternatively, the physiological state of the injury (i.e., the 269
size of the open wound) might also be linked to how much care is needed to increase the likelihood of survival of 270
the injured individual. 271
We observed that ants provided more wound care towards leg-injuries than antennal injuries. Previous studies 272
show that ants can differentiate wounds (Frank et al. 2023a; Frank et al. 2023b preprint), giving greater attention 273
towards wounds that carry a greater mortality risk. Our survival assays show that infected leg injuries carry greater 274
mortality risks than infected antennal injuries, yet our infection assay revealed the highest levels of wound care 275
towards sterile leg injuries. This may be a strategy of reducing the risk of wound infection in the first place, thus 276
increasing the survival chances of leg-injured ants. Antennal injuries on the other hand carry a relatively low 277
mortality risk, even when exposed to pathogens, likely because of the reduced wound size compared to leg injuries 278
13
and a lower hemolymph flow (Pass 1991). It is interesting to note here, that a previous study also reported no 279
differences in foraging activity or other tasks in antenna-injured Cataglyphis niger ants (Gilad et al. 2022). 280
Following this, antennal injuries likely do not represent a major handicap during nestmate interactions or foraging. 281
A point of interes t is that sterile leg injuries survive better in isolation than in the nest (see Fig. 2b), where they get 282
the highest level of wound care. Although this comparison lacks statistical significance, it hints at the possibility 283
that these injuries remain fairly clean in isolation, whereas they are exposed to environmental pathogens in the 284
nest and foraging arena despite wound care. 285
Interestingly, infected wounds were treated differently. The survival analysis revealed that infection resulted in a 286
substantial increase in mortality for ants with leg injuries, both when isolated and within the nest, but not in 287
antennal injuries. The high mortality of infected leg injured ants in the nest will be discussed below (see Care/Kill 288
dichotomy). While infected antenna-injured ants had a slightly higher chance of survival when nestmates' help was 289
present, their survival was still high even in isolation. This underscores the considerable difference in mortality 290
risks across wound types and the role of nestmate assistance in the survival of infected ants. 291
We c an c on cl ude t ha t leg-injuries carry a greater risk of causing lethal infections when exposed to pathogens than 292
antennal injuries. Both the wound size – which increases the surface for pathogen exposure – and a potentially 293
greater vascular (hemolymph) circulation in the leg may contribute to the lethality of the pathogen on the wound 294
(Hantschk 1991; Pass 2000; Oh et al. 2019; Frank et al. 2022; Frank et al. 2023 preprint). 295
It is interesting to note that wound care also quantitatively differed between colonies. The colony with fewer 296
members engaged more in wound care compared to the other two colonies which consisted of approx. twice as 297
many individuals. This difference could be associated with the higher relative value of an individual in smaller 298
colonies as theorized by Frank et al. (2017). 299
Allogrooming, as a collective hygiene measure (Cremer et al. 2007), was consistently provided, irrespective of 300
wound status, even for healthy ants. This allogrooming behaviour is part of the general hygienic habit of ants and 301
was sometimes directed towards the experimental paint mark on the thorax of nestmates, in an attempt to remove 302
it. However, we did not score color grooming in our study to make a conclusive statement. Allogrooming might 303
serve as a form of social care, particularly for nestmates undergoing stress, regardless of pathogen exposure. While 304
the occurrence of aggression was primarily observed during the initial phase when focal ants were reintroduced 305
into the nest, suggesting it could be an initial response independent of wound type. 306
14
We, fur th erm or e, i de nt if ie d a p ot en ti al ly new t yp e o f in ter ac ti on in wo un d car e beh av ior : aci d spr ay ing o n a n 307
injured nestmates wounded area or body. Previous studies showed that acid spraying improves disease resistance 308
in ant colonies (Tranter and Hughes 2015; Tragust et al. 2013) and is even used for destructive disinfection of 309
brood (Pull et al. 2018). However, we only observed this behavior seven times during the study. Nevertheless, this 310
can be seen as a strong indication of acid use to potentially disinfect wounds, constituting a novel care strategy in 311
social insects. 312
Care/Kill dichotomy 313
While both infected and sterile leg injuries received care for the first hours after wounding this drastically changed 314
afterward, with infected injuries receiving no care and the infected individual being socially isolated. These 315
infected ants soon behaved like moribund ants and left the nest of their own volition to die in isolation (Heinze 316
and Walter 2010, Bos et al. 2012; Miler 2016). In a few cases, the ant’s dead or immobile body, which was located 317
inside the nest, was then transported by nestmates to a garbage pile in the foraging area. The contrast between 318
sterile and infected injuries suggests the presence of a care-kill breakpoint in C. nodus. Leg injuries exposed to 319
even a low concentration of P. ae ru gi no s a, may be considered “beyond” saving, either by the injured ant itself 320
and/or its nestmates. The following social isolation of the infected ant reduces the risk of infection transmission 321
inside the colony and the energetic investment needed to care for a wound by the nestmates. 322
Since our observations in the infection assay were limited to only 60 sec for every individual per hour, further 323
detailed and more frequent observations would be needed to clarify the precise nature of this nest-leaving behavior. 324
It would be interesting to find out how exactly the nest-leaving behavior of moribund ants is happening, and to 325
what extent it is a collective or an individual decision. In this regard, further manipulation of different pathogen 326
concentrations, combined with analyses of cuticular chemical profiles of sterile and infected ants could provide 327
promising avenues to pursue. 328
Conclusion 329
In conclusion, our study confirms the presence of wound care behavior in Cataglyphis nodus ants. These ants assist 330
injured nestmates differently depending on the wound type. Leg-injured nestmates received more wound care than 331
antennal-injured ones. However, leg injuries carried a greater mortality risk when exposed to pathogens than 332
antennal injuries. Infected leg-injured ants therefore not only had lower survival in isolation but also received less 333
wound care inside the nest compared to sterile ones. Leg injuries have a wider wound opening compared to 334
15
antennal injuries, increasing the amount of pathogens that can enter the wound at exposure. Hence, ants with 335
infected wounds on their leg are expelled from the nest, either out of their own volition or by nestmates. This not 336
only exemplifies the care-kill dichotomy in social immunity when it comes to saving injured nestmates but also 337
shows that these ants can differentiate between injuries with high or low mortality risks, leading to more care for 338
sterile leg injuries to prevent pathogens from entering the wound. 339
Data Availability 340
All data generated during this study are provided as supplementary material (S1 and S2). 341
342
Competing Interest 343
Authors have no conflict of interest to declare. 344
345
Acknowledgments 346
N. I. Beydizada was supported by a grant (program in support of the doctoral students mobilities - A6, ID: 347
2126043) from the Department of Botany and Zoology, Masaryk University, Brno, Czech Republic for her 348
internship in the Animal Ecology and Tropical Biology Department, Ant Care Lab, at the University of Würzburg, 349
Germany. E. T. Frank was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research 350
Foundation) Emmy Noether Program - no. 511474012. P. Schultheiss was supported by the Deutsche 351
Forschungsgemeinschaft (DFG, German Research Foundation) - no. 499479766. 352
Author Contributions 353
E. T. Frank, N. I. Beydizada & A. Abels: Conceptualization; Designing the experiment; Methodology; Data 354
preparation; Data analysis; N. I. Beydizada: Writing–Original draft; N. I. Beydizada, E. T. Frank & P. Schultheiss: 355
Writing–Review & editing; P. Schultheiss: Material collection; N. I. Beydizada: Conducting experiments; E. T. 356
Frank: Supervision. All authors contributed critically to the drafts and gave final approval for publication. 357
16
References 358
Alexander RD (1974) The evolution of social behavior. Annual Review of Ecology and Systematics 5(1):325-383.
Anderson RM, May RM (1979) Population biology of infectious diseases: Part I. Nature, London 280:361-7.
Bos N, Lefèvre T, Jensen AB, d’Ettorre P (2012) Sick ants become unsociable. Journal of Evolutionary Biology
25(2):342-351.
Büchler R, Drescher W, Tornier I (1992) Grooming behaviour of Apis cerana, Apis mellifera and Apis dorsata
and its effect on the parasitic mites Varroa jacobsoni and Tropilaelaps clareae. Experimental & Applied
Acarology 16(4):313-319.
Cremer S, Armitage SA, Schmid-Hempel P (2007) Social immunity. Current biology, 17(16):693-702.
Cremer S, Pull CD, Fürst MA (2018) Social immunity: emergence and evolution of colony-level disease protection
Annual Review of Entomology 63:105-123. https://doi.org/10.1146/annurev-ento-020117-043110
Cremer S (2019) Social immunity in insects. Current Biology 29(11):458-463.
Côté IM, Poulin R (1995) Parasitism and group size in social mammals: a meta-analysis. Behavioral Ecology
6:159-165.
Diez L, Lejeune P, Detrain C (2014) Keep the nest clean: survival advantages of corpse removal in ants. Biology
letters 10(7):20140306. https://doi.org/10.1098/rsbl.2014.0306
Frank ET, Schmitt T, Hovestadt T, Mitesser O, Stiegler J, Linsenmair KE (2017) Saving the injured: rescue
behavior in the termite-hunting ant Megaponera analis. Science Advances 3(4):e1602187. https://doi:
10.1126/sciadv.1602187
Frank ET, Wehrhahn M, Linsenmair KE (2018) Wound treatment and selective help in a termite-hunting ant.
Proceedings of the Royal Society B: Biological Sciences 285(1872):20172457.
https://doi.org/10.1098/rspb.2017.2457
Frank ET, Kesner L, Liberti J, Helleu Q, LeBoeuf AC, Dascalu A, ... & Keller L (2022) Antimicrobial wound care
in an ant society. Available at BioRxiv 2022-04. https://doi.org/10.1101/2022.04.26.489514
Frank ET, Kesner L, Liberti J, Helleu Q, LeBoeuf AC, Dascalu A, ... & Keller L (2023a) Targeted treatment of
injured nestmates with antimicrobial compounds in an ant society. Nature Communications 14(1):8446.
https://doi.org/10.1038/s41467-023-43885-w
17
Frank ET, Buffat D, Liberti J, Keller L (2023b) Wound Dependent Leg Amputations to Combat Infections in an
Ant Society. Available at SSRN 4612970.
Gellert HR, Halley DC, Sieb ZJ. et al. (2022) Microstructures at the distal tip of ant chemosensory sensilla.
Scientific Report 12:19328. https://doi.org/10.1038/s41598-022-21507-7
Gilad T, Dorfman A, Subach A, Scharf I (2022) Leg or antenna injury in Cataglyphis ants impairs survival but
does not hinder searching for food. Current Zoology 68(4):441-450. https://doi.org/10.1093/cz/zoab027
Hantschk AM (1991) Functional morphology of accessory circulatory organs in the legs of Hemiptera.
International Journal of Insect Morphology and Embryology 20(6):259-273.
Harkness RD, Wehner R (1977) Cataglyphis. Endeavour 1(3-4):115-121.
Hart BL (1992) Behavioral adaptations to parasites: an ethological approach. The Journal of Parasitology 256-265.
Hart BL (2011) Behavioural defences in animals against pathogens and parasites: parallels with the pillars of
medicine in humans. Phil Trans R Soc B 366:3406-3417. https://doi.org/10.1098/rstb.2011.0092
Hart AG, Ratnieks FL (2002) Waste management in the leaf-cutting ant Atta colombica. Behavioral Ecology
13(2):224-231. https://doi.org/10.1093/beheco/13.2.224
Heinze J, Walter B (2010) Moribund ants leave their nests to die in social isolation. Current Biology 20(3):249-
252. https://doi.org/10.1016/j.cub.2009.12.031
Hillyer JF, Pass G (2020) The insect circulatory system: structure, function, and evolution. Annual Review of
Entomology 65:121-143.
Kessler SE (2020) Why Care: Complex Evolutionary History of Human Healthcare Networks. Frontiers in
Psychology 11:199. https://doi.org/10.3389/fpsyg.2020.00199
Mascaro A, Southern LM, Deschner T, Pika S (2022) Application of insects to wounds of self and others by
chimpanzees in the wild. Current Biology 32(3):112-113.
Miler K (2016) Moribund ants do not call for help. PLoS One 11(3):e0151925.
https://doi.org/10.1371/journal.pone.0151925
Mukherjee S, Heithaus MR (2013) Dangerous prey and daring predators: a review. Biological reviews 88(3):550-
563. https://doi.org/10.1111/brv.12014
Nunn C, Altizer SM (2006) Infectious diseases in primates: behavior, ecology and evolution. Oxford University
Press, USA.
18
Oh JK, Behmer ST, Marquess R, Scholar EA, Akbulut M (2019) Investigation of mechanical properties of tibia
and femur articulations of insect joints with different joint functions. MRS Communications 9(3):900-
903. https://doi.org/10.1557/mrc.2019.71
Oi DH, Pereira RM (1993) Ant behavior and microbial pathogens (Hymenoptera: Formicidae). Florida
Entomologist 63-74.
Pass G (1991) Antennal circulatory organs in Onychophora, Myriapoda and Hexapoda: functional morphology
and evolutionary implications. Zoomorphology 110:145-164.
Pass G (2000) Accessory pulsatile organs: evolutionary innovations in insects. Annual Review of Entomology
45(1):495-518.
Porter SD, Jorgensen CD (1981) Foragers of the harvester ant, Pogonomyrmex owyheei: a disposable caste?.
Behavioral Ecology and Sociobiology 9:247-256
Pull CD, Ugelvig LV, Wiesenhofer F, Grasse AV, Tragust S, Schmitt T, ... & Cremer S (2018). Destructive
disinfection of infected brood prevents systemic disease spread in ant colonies. Elife, 7, e32073.
https://doi.org/10.7554/eLife.32073
Schmid-Hempel P, Koella JC (1994) Variability and its implications for host-parasite interactions. Parasitology
Today 10(3):98-102.
Schmid-Hempel P (1998) Parasites in social insects (Vol. 60). Princeton University Press.
Schmid-Hempel P (2017) Parasites and their social hosts. Trends in Parasitology 33(6):453-462.
https://doi.org/10.1016/j.pt.2017.01.003
Sommer S, Wehner R (2012) Leg allometry in ants: extreme long-leggedness in thermophilic species. Arthropod
Structure and Development 41:71-77. https://doi.org/10.1016/j.asd.2011.08.002
Tragust S, Mitteregger B, Barone V, Konrad M, Ugelvig LV, Cremer S (2013) Ants disinfect fungus-exposed
brood by oral uptake and spread of their poison. Current Biology 23(1):76-82.
http://dx.doi.org/10.1016/j.cub.2012.11.034
Tranter C, Hughes WOH (2015) Acid, silk and grooming: alternative strategies in social immunity in ants?.
Behavioral Ecology and Sociobiology 69:1687-1699.
Wehner R (1987) Spatial organization of foraging behavior in individually searching desert ants, Cataglyphis
(Sahara Desert) and Ocymyrmex (Namib Desert). Experientia Supplementum 54:15-42.
19
Wehner R, Harkness RD, Schmid-Hempel P (1983) Foraging strategies in individually searching ants Cataglyphis
bicolor (Hymenoptera: Formicidae). Gustav Fischer, 79 p.
Westergaard G, Fragaszy D (1987) Self-treatment of wounds by a capuchin monkey (Cebus apella). Human
Evolution 2:557-562.
Wilson-Rich N, Spivak M, Fefferman NH, Starks PT (2009) Genetic, individual, and group facilitation of disease
resistance in insect societies. Annual Review of Entomology 54:405-423.
https://doi.org/10.1146/annurev.ento.53.103106.093301
