One size fits all? Determinants of sperm transfer in a highly dimorphic orb‐web spider

Abstract

The evolutionary significance of widespread hypo-allometric scaling of genital traits in combination with rapid interspecific genital trait divergence has been of key interest to evolutionary biologists for many years and remains poorly understood. Here, we provide a detailed assessment of quantitative genital trait variation in males and females of the sexually highly dimorphic and cannibalistic orb-weaving spider Argiope aurantia. We then test how this trait variation relates to sperm transfer success. In particular, we test specific predictions of the one-size-fits-all and lock-and-key hypotheses for the evolution of genital characters. We use video-taped staged matings in a controlled environment with subsequent morphological microdissections and sperm count analyses. We find little support for the prediction of the one-size-fits-all hypothesis for the evolution of hypo-allometric scaling of genital traits, namely that intermediate trait dimensions confer highest sperm transfer success. Likewise, our findings do not support the prediction of the lock-and-key hypothesis that a tight fit of male and female genital traits mediates highest sperm transfer success. We do, however, detect directional effects of a number of male and female genital characters on sperm transfer, suggesting that genital trait dimensions are commonly under selection in nature. Importantly, even though females are much larger than males, spermatheca size limits the number of sperm transferred, contradicting a previous hypothesis about the evolutionary consequences of genital size dimorphism in extremely size-dimorphic taxa. We also find strong positive effects of male body size and copulation duration on the probability of sperm transfer and the number of sperm transferred, with implications for the evolution of extreme sexual size dimorphism and sexual cannibalism in orb weavers.

Increased insertion number leads to increased sperm transfer and fertilization success in a nursery web spider

“…In some species, male size is positively related to increased sperm quantity and increased sperm transfer (e.g. Assis & Foellmer, 2016;Ceballos, Jones, & Elgar, 2015;Wiernasz, Sater, Abell, & Cole, 2001), but we found no evidence that male size influences the initial amount of sperm stored in male's pedipalps or the amount of sperm that males transfer to female P. mira. Similar levels of variation in the number of sperm stored in virgin males' pedipalps has been observed in other spider species (e.g.…”

“…Eberhard, 1996;Herberstein et al, 2011). Indeed, several studies have failed to find a relationship between copulation duration and fertilization success (Assis & Foellmer, 2016;Gilchrist & Partridge, 2000;Mazzi et al, 2009). Thus, in attempting to understanding male-female mating dynamics and potential costs and benefits of copulation duration, it is important to directly quantify the relationships between copulation duration, sperm transfer, and fertilization success, as well as their relationship to female and male size.…”

“Where ignorance is bliss, 'tis folly to be wise”: Indiscriminate male care in a neotropical spider

“…A second possibility is that females may use the entire sperm storage from the first male to fertilize the eggs laid in the initial egg sacs, especially if the copulation duration of the first partner is low or he produces less sperm than the second one (Assis & Foellmer, 2016). Females may also dump the sperm from their spermathecae, such as observed in other spider species (Eberhard, 2009).…”

Intra‐specific variation of non‐genitalic and genitalic traits in two euophryine jumping spider species

“…Three regression methods, ordinary least squares (OLS), major axis (MA) and reduced major axis (RMA), have been widely applied for estimating allometric scaling relationships (Eberhard, 2009), and there have been extensive debates on which methods are more appropriate (Bertin & Fairbairn, 2007; Bonduriansky, 2007; Eberhard, 2009; Rodríguez et al ., 2015a). Following an approach in Assis & Foellmer (2016), we presented the results from all three methods, but relied more on the OLS estimates in the results and discussion because the other regressions confound the scatter of the measurements with variation in slope and can be biased towards isometry for the log‐log transformed variables (Voje & Hansen, 2013; Rodríguez et al ., 2015a; Assis & Foellmer, 2016). The slope estimates were tested if differed significantly from the null hypothesis of isometry (slope = 1) using the SMATR package in R (Warton et al ., 2012).…”