Mating with immature females is an alternative tactic for brown widow males, since adult females cannibalize mating males & immature females do not; but males approached and preferred to mate with adult females

Alternative mating tactics in a cannibalistic widow spider: do males prefer the safer option? Lenka Sentenská, Gabriele Uhl, Yael Lubin. Animal Behaviour, Volume 160, February 2020, Pages 53-59. https://doi.org/10.1016/j.anbehav.2019.11.021

Highlights
•    Mating with immature females is an alternative tactic for brown widow males.
•    Adult females cannibalize mating males, but immature females do not.
•    Males approached and preferred to mate with adult females over receptive immatures.
•    Males did not show a preference for unreceptive versus late-stage receptive immatures.
•    Close-range cues in webs of immatures may indicate receptivity.

Abstract: Mating generally occurs with adult females, which undergo a suite of changes in morphology, physiology and behaviour during maturation. In the brown widow spider, Latrodectus geometricus, however, males can mate with immature females during a short period before they moult to the adult stage. Mating with immature females seems beneficial for males, because they are not at risk of being cannibalized, whereas cannibalism inevitably occurs in matings with adult females. We conducted choice experiments to elucidate male preference, courtship and mating behaviour with immature and adult females of different ages. We controlled for age of the females’ webs to provide males with potential web-borne attractants of similar age. We tested whether males distinguish immature females that are ready to mate (late subadult stage) from adult females and from immature females that do not mate (early subadults), and we examined male response to young versus old adult females. Males approached and mated with adult females more frequently than late subadult females, but there were no differences in the frequencies of approach to early and late subadults or to adult females of different ages. Once on the web, however, males attempted to mate with the late subadults. We suggest that web-borne volatile cues, typical of adult females, may be reduced or lacking in late subadult females, yet less volatile cues may indicate receptivity.

Discussion

Male L. geometricus mated with late subadult females, which would seem to be highly advantageous for the male due to the reduced courtship combined with consistently successful copulation with both palps and lack of sexual cannibalism (Biaggio et al., 2016, Waner et al., 2018; this study). However, when we presented males with late subadult and adult females simultaneously, they typically chose adult, cannibalistic females over the noncannibalistic subadult females. Males showed this preference even when we controlled for web age, a confounding factor that might have affected their choice for adult females observed in a previous study (Waner et al., 2018).

Sexual Attractants

Male L. geometricus might not recognize subadult females as potential mates due to a lack of sex pheromone (Waner et al., 2018). For example, Fisher et al. (2018) suggested that subadult females of the false black widow spider, Steatoda grossa, do not produce a sexual-attractant pheromone. However, male widow spiders are often found on webs of subadult females in nature (Y. Lubin, personal observation). Furthermore, males readily approached the subadult females in their webs, when we exposed the males to late subadult females against an empty control and they did so as rapidly as when they approached adult females. These observations suggest that even before the final moult females produce cues that act as sexual attractants by which males recognize subadult females as potential mates.

In our experiments, we found no evidence that the age of adult females influenced the male's response in choice tests, contrary to previous findings (Waner et al., 2018) where males preferred older adult females to young adults and late stage subadults. The result of Waner et al. (2018) was probably due to the presence of a stronger pheromone cue that had accumulated in the older web. We also observed no difference in the male's approach towards late subadult females, which are ready to mate, and early subadult females, which are not. This suggests that males cannot identify from a distance whether the subadult females are ready to mate. However, after contacting their webs, males courted the early subadult females only very briefly and then remained in their webs without any further courting, but males immediately courted and then mated with late subadult females. Thus, our observations agree with other studies on spider chemical communication, which suggest that airborne chemicals provide less specific information than chemicals detected by contact with the web (reviewed in Gaskett, 2007, Uhl, 2013). Similarly, when presented with adult and late subadult females, males seemed to recognize the female stage only upon contacting the web because only then did they begin to add silk when with an adult female or vibrate when with a late subadult female.

The observed preference of males for adult rather than subadult females could be due to quantitative differences in male-attracting signals or to different cues emitted by subadult and adult females. Virgin adult females of many spider species attract males by producing pheromones that signal readiness to mate (Gaskett, 2007, Kasumovic and Andrade, 2004, Riechert and Singer, 1995, Roberts and Uetz, 2005, Stoltz et al., 2007, Uhl, 2013). Subadult mating in L. geometricus occurs during a 4-day period before the final moult (late subadult stage). In general, adults seem to produce more pheromone and thus provide a stronger signal than immature females and often only adults produce pheromones (Gaskett, 2007, Uhl and Elias, 2011, Uhl, 2013, Fischer et al., 2018). Then, the observed preference for adult females represents rather an attraction to a stronger signal than a preference for a certain stage per se. A few pheromones have been chemically characterized for virgin adult females (e.g. in the genus Latrodectus; Jerhot, Stoltz, Andrade, & Schulz, 2010) but none for subadults. Although males often cohabit with subadult females (Jackson, 1986), these females might not produce sex pheromones, and males may identify them by unintentionally produced chemical cues (Fischer, 2019). Therefore, olfactory cues produced by subadult females may be qualitatively different from pheromones of adult females, potentially allowing males to differentiate between the two stages. If these chemicals are distinguishable by males, the choice of adult females over late subadults might indicate that the former are perceived as higher-quality mates, even though mating with them limits males to a single copulation and sometimes even to a single insertion (Segoli, Arieli, et al., 2008).

Costs to Males of Mating with Subadult Females

There may be fitness costs to males adopting the subadult mating tactic. After mating, subadult-mated females still have to undergo a final moult to adulthood and may have a lower probability of surviving to oviposition than adult-mated females. The moulting process itself is a sensitive period due to the risk of predation on moulting or freshly moulted spiders, the risk of desiccation or an inability to release the old cuticle (e.g. Horner and Starks, 1972, Jones, 1941, Tanaka, 1984). Thus, males may prefer to mate with adult females due to the overall greater probability of successful reproduction.

Costs of mating with late subadult females could also arise from the specific mating behaviour and the mechanisms of copulation, sperm transfer and sperm storage. When mating with subadult females, L. geometricus males do not somersault and are not cannibalized. In the congener L. hasselti, cannibalism reduces the likelihood of a female remating (Andrade, 1996). Thus, the lack of cannibalism might lead to a greater probability of remating in subadult-mated females and consequently to paternity loss for the first male. Furthermore, during courtship with adult females, the male removes a large part of the female's web and adds his own silk. Webs of adult female L. hasselti that were thus altered by males attracted fewer suitors (Scott, Kirk, McCann, & Gries, 2015), a phenomenon observed also in other web-building species (reviewed in Scott, Anderson, & Andrade, 2018). By contrast, we showed here that webs of subadult-mated females were not altered by the male; male courtship was brief, the web remained intact and the male added little silk. A subsequent male might thus have no indication of a previous visitor. It is unclear whether subadult-mated females remain attractive to males and whether these females will remate after maturing to adults (Biaggio et al., 2016, Waner et al., 2019). Finally, although mating with a subadult female enables the male to seek an additional female, high mortality during mate search (more than 80% of L. hasselti males die without finding a mate; Andrade, 2003) may reduce the benefit of such matings.

A male mating with subadult females may have lower paternity than expected for the first male in adult matings due to unfavourable sperm storage conditions or incorrect placement of sperm in the subadult female spermathecae, or to lower competitive ability of his sperm against a second male's sperm. If the internal genitalia of late subadult females are not fully developed, sperm storage conditions may differ from those in adult females and might result in a lower paternity share for a male's sperm when competing with ejaculates of other males. It is possible that the mating plugs cannot be placed correctly or can shift when subadult females moult and thus may be a less effective barrier to remating. Additionally, the lack of somersaulting while mating with late subadult females may mechanically alter the insertion mechanism and affect where sperm is deposited within the female's reproductive tract. The location of deposited sperm and the storage conditions, together with potentially insufficient plugging in subadult matings, may yield lower reproductive success. In the congener, L. hasselti, the first of two males mating with an adult female achieves approximately 80% paternity (Snow & Andrade, 2005). However, the paternity share may differ if a subsequent male inseminates a female mated as subadult after she has moulted. These potential costs of mating with a subadult female can be revealed through paternity assessments in double-mating trials.

Our observation showed that L. geometricus males mate with late subadult females, but do not attempt to do so with younger subadults. Despite this, they did not show any preference for late over early subadult females. Cohabiting with subadult females and then mating with them when they moult is known for many spider species (Jackson, 1986) including the widow spiders (Biaggio et al., 2016, Segoli et al., 2006). Although in L. hasselti and L. geometricus the encounter with late subadult females often leads to immediate mating, cohabiting with early subadult females and waiting for them to mature may be another mating tactic in a male's repertoire. Additionally, it is likely that males cannot determine from a distance subadult females' readiness to mate.