The indispensable soma of Cardiocondyla obscurior ants

The evolutionary mechanisms that shape aging in social insects are not well understood. It is commonly assumed that queens live long and prosperous, while workers are regarded as a short-lived disposable caste because of their low reproductive potential. Queens of the ant Cardiocondyla obscurior gain high fitness late in life by increasing investment into sexual offspring as they age. This results in strong selection against senescence until shortly before death. Here, we show that workers have the same lifespan and shape of aging as queens, even though workers lack reproductive organs and cannot gain direct fitness. Under consideration of the prevailing aging theories and the biology of the species, we hypothesize that programmed aging has possibly evolved under kin selection. Impact statement Morphologically distinct fertile queen and sterile worker castes in the model ant Cardiocondyla obscurior show the same pace and shape of aging, contradicting the paradigm of queen/worker lifespan divergence in social insects.


Introduction 43
Queens of some social Hymenoptera (ants and bees) live long while being highly 44 fertile, seemingly avoiding a trade-off between lifespan and reproduction (Hartmann Cardiocondyla obscurior, a model for social insect aging (reviewed in Oettler and 47 Schrempf, 2016), we recently identified molecular processes associated with queen 48 aging , and demonstrated that the strength of selection on 49 age-biased genes differs between social and solitary insects (Harrison et al., 2021). 50 Based on these insights and lifespan data from C. obscurior queens, including 51 transcriptomic data from queens shortly before death, a life history framework was 52 proposed (Jaimes-Nino et al., 2022). 53 54 The concept of "continuusparity" is grounded in a pattern observed in many ants, 55 namely that resources are first invested into workers, and only when colony size has 56 reached a certain threshold, resource investment is diverted to the production of 57 sexual offspring (males, queens) (Oster and Wilson, 1978). Thus the strength of 58 selection against senescence does not decline with age despite life-long 59 reproduction (Jaimes-Nino et al., 2022). Irreversible reproductive division of labor 60 between queens and workers is a pre-requisite for the evolution of continuusparity, 61 however, not all social insects must exhibit continuusparity, for example, species with 62 only one caste. In respect to senescence, continuusparous social insects should sit 63 between iteroparous and semelparous species, the former experiencing decreasing 64 selection strength after the first reproductive bout, while the latter undergo strong 65 selection against senescence until a single, and final, reproductive bout, followed by 66 reproductive death.

68
As progress is made in understanding aging in ant queens, the aging patterns of 69 workers remain largely unknown. Anecdotal references such as "mother queens live 70 much longer than workers in all groups of ants" (Hölldobler and Wilson, 1990) microscope slides and covered with a dark foil, placed in a square petri dish half-filled 105 with plaster. The ants always had access to water, and were fed with honey, 106 Drosophila and pieces of cockroaches twice per week. To test for an effect of 107 workload on worker lifespan, colonies were subjected to one of three treatments: 1) 108 no larvae (NL), 2) with two second instar larvae (low workload, LW), and 3) with ten 109 second instar larvae (high workload, HW) (n=20 each). To keep workload constant, in 110 the two treatments containing brood, colonies were checked every few days and 111 larvae that had pupated were removed and replaced by new second instar larvae.

113
After 18 weeks there was no difference in survival between workers with or without 114 larvae (see results). This led us to adjust treatments to provoke more variation in 115 lifespan. Colonies were standardized to five workers, and those containing fewer than 116 five workers were excluded (n after exclusion: NL=18, LW=20, HW=16). The 117 remaining replicates from each treatment were split into two groups; in one group, 118 workers were kept without larvae while in the other, two second instar larvae were 119 added to simulate low workload (n after split: NL=9, NL-LW=9, LW-NL=10, LW=10, 120 HW-NL=8, HW-LW=8). Larvae were replaced as described above. As after 36 weeks 121 still no effect of workload treatment was apparent (see results), we removed all larvae 122 and continued to monitor worker survival weekly until all workers had died. 123 124 Differences in survival across worker treatments and between workers and queens 125 (using queen data from Jaimes-Nino et al. After this bout of sexual production, mortality increases. Queens die soon after they 172 cease egg laying, suggesting that reproduction is optimized and intrinsic resources 173 are depleted (Jaimes-Nino et al., 2022). Finding a similar pace and shape of aging in 174 workers is surprising because queens and workers differ in morphology (Fig 3,  architecture, temperature, humidity, feeding rate). In the present study, workers also 206 lived considerably longer than in the previous experiment in which focal workers were 207 kept with 10 or 20 workers (each marked by tarsal clipping), and a fertile queen, 208 presumably not apply here, including e.g., antagonistic pleiotropy, which postulates that genes 256 with positive effects on the germline in early life can be selected for even if they cause 257 senescence in the soma later in life. As antagonistic pleiotropy requires a phase after 258 the point of strongest selection, it is unlikely to be effective in C. obscurior because 259 there is no such phase (Jaimes-Nino et al., 2022). Along the same lines, 260 metaphorically speaking, the results suggest that C. obscurior workers are not a 261 disposable soma.

263
In contrast to the first two approaches aimed at understanding senescence, a third 264 controversial perspective considers whether death is not just a consequence of iteroparous species exhibiting repeated reproduction and increasing risks and costs 272 of somatic maintenance, and a senescent phase that is rarely expressed in nature.

273
This is partly a problem of semantics because "iteroparity" comprises a wide range 274 of reproductive strategies, including humans with an exceptionally long post-275 reproductive and senescent phase. The concept of programmed aging seems easier 276 to imagine in semelparous species, where reproduction is a single event. However, 277 in semelparous species it is not lifespan that is determined. Instead, reproduction is 278 optimized towards one episode, which is often triggered by season, probability of 279 encountering prey or hosts, or other extrinsic factors.

281
Despite all valid counterarguments, including inconceivable proximate mechanisms, 282 programmed aging can explain not only the shape of aging of C. obscurior queens, 283 but also the similarity of queen and worker aging patterns. All queens show increasing 284 investment into sexuals with age, irrespective of overall fertility, lifespan or colony 285 size, followed by increasing mortality. This indicates some sort of Zeitgeber, informing 286 the queen to increase investment into sexual production. As potential environmental 287 signals are absent under controlled conditions in the lab, such a Zeitgeber must be 288 linked to the organism's physiological condition, and it must act as an honest signal, 289 immune to mutations and cheating. This Zeitgeber is likely a finite or limited resource.

290
A putative mechanism may be related to cellular aging, such as a  and a confidence interval of 95%. The dashed gray line at y = 1 indicates when relative 327 mortality is equivalent to mean mortality. 328 329 330 331 Figure 3. A virgin C. obscurior queen (left) and a worker tending to some brood. Ó 332 Lukas Schrader.

334
Glossary 335 Aging -Time passing by. Neutral in terms of fitness -often confused with senescence. 336 Life expectancy -Species (mean) lifespan estimate based on mortality rates. 337 Lifespan -The maximum age an organism can reach -often confused with life 338 expectancy.

339
Mortality rate (age-specific) -Expected probability to die at a given age in a particular 340 population.

341
Pace of aging -Related to the extension of the life of an organism (either mean or 342 maximum lifespan) See Baudisch, 2011. 343 Senescence -When an organism shows signs of increasing mortality and decreasing 344 fertility.