Mating and longevity in ant males

Abstract Across multicellular organisms, the costs of reproduction and self‐maintenance result in a life history trade‐off between fecundity and longevity. Queens of perennial social Hymenoptera are both highly fertile and long‐lived, and thus, this fundamental trade‐off is lacking. Whether social insect males similarly evade the fecundity/longevity trade‐off remains largely unstudied. Wingless males of the ant genus Cardiocondyla stay in their natal colonies throughout their relatively long lives and mate with multiple female sexuals. Here, we show that Cardiocondyla obscurior males that were allowed to mate with large numbers of female sexuals had a shortened life span compared to males that mated at a low frequency or virgin males. Although frequent mating negatively affects longevity, males clearly benefit from a “live fast, die young strategy” by inseminating as many female sexuals as possible at a cost to their own survival.

Cardiocondyla males are therefore capable of mating with larger numbers of female sexuals (Heinze & Hölldobler, 1993), males of all other social Hymenoptera are sperm-limited because their testes degenerate before or shortly after adult emergence (Hölldobler & Bartz, 1985). Therefore, wingless Cardiocondyla males are a suitable model to investigate the interrelation between reproduction and longevity in ant males. In contrast to queens, ant males do not actively contribute to the social life of the colony and also receive less care than the queens. We therefore hypothesize that, because males are less well integrated in the society, their costs of reproduction will not be borne fully by the workers and increased mating activities will therefore shorten their life span.

| MATERIAL AND METHODS
We set up 115 experimental colonies by separating a mated queen, 15-20 workers, and a wingless male pupa from our laboratory stock colonies (collected in Ilhéus, Brazil) of the ant Cardiocondyla obscurior (Wheeler, 1929) in new nest boxes. The ants were kept in the laboratory under near-natural conditions in climatic chambers with 28/23°C temperature and 12/12 h day/night cycles (for details see Cremer & Heinze, 2003). Experimental colonies were checked daily for the eclo- The life spans of males still alive at the end of the experiment were treated as censored (two HR males, one LR male). Likewise, male life span was censored in three colonies, which had to be terminated prematurely because of the emergence of a female sexual (2 V groups) or invasion by ants from another colony (LR group). In the 41 HR and LR colonies from 2014 and six HR colonies from 2006 (HR, N = 26; LR, N = 21), potentially mated young queens were removed after 5-7 days in the colony and subsequently dissected under a binocular to investigate whether they have been inseminated. If so, sperm was clearly visible inside the female sperm storage organ (spermatheca).
To check whether any differences in male life span might result from behavioral differences or different treatment by workers we set up additional colonies in 2014 (V, N = 6; LR, N = 7; HR, N = 7) as described above and observed them for 3 min twice per day over 11 consecutive days. Depending on life span, we performed three to 21 observations per male. We recorded active (e.g., running, antennating, allogrooming, self-grooming) and passive behavior (e.g., resting, being antennated or groomed) of the males and the male's location in the nest (center, periphery, outside).
All statistical analyses were performed in IBM SPSS Statistics (Version 21).

| RESULTS
Dissection showed that HR males mated with more female sexuals Males of the different mating groups were similarly treated by workers and also did not differ in their location in the nest (all p ≥ .095). However, HR males tended to be more active than LR and F I G U R E 1 Life span (days) of Cardiocondyla obscurior ant males dependent on the availability of mating partner (V: no female sexuals available, LR: 1-3 female sexuals available per week; HR: 6-60 female sexuals available per week; censored data are indicated by cross-hairs). HR males suffer a reduced life span relative to LR and V males F I G U R E 2 Life span and the number of females inseminated by males of the ant Cardiocondyla obscurior are positively correlated in both the LR and HR treatment