Mating rituals in the animal kingdom are often quite extraordinary, in particular when mating is traumatic. We here describe the exceptional traumatic mating behaviour of the currently undescribed sea slug, Siphopteron sp. 1. Similar to four congeners, Siphopteron sp. 1 routinely exhibits traumatic secretion transfer through a stylet-like penis appendage. Contrary to previous descriptions, however, prostate secretions are injected centrally into the partner's forehead, representing, to our knowledge, the first-known instance of ‘cephalo-traumatic secretion transfer’. We further provide a comparative quantification of within- and between-species variation in injection sites and derive a potential neurophysiological function of prostate secretions that are injected close to, or into, the central nervous system.
Mating rituals in the animal kingdom can appear bizarre, with striking examples found when species engage in the so-called traumatic mating. This occurs when dedicated male structures wound the female's integument during copulation, irrespective of whether or not this imposes fitness costs to the recipient . Traumatic mating often serves to inject prostate fluids or other gland secretions with putative allohormonal functions into the female [2–5]. Such ‘traumatic secretion transfer’ can benefit the male by stimulating or manipulating the female's current reproductive investment, inducing refractory periods or increasing the sperm donor's fertilization share (reviewed in  and a review of seminal fluid protein functions in ). It may arise as a secondary male adaptation, in particular, in systems where females have already adapted to manipulation by allohormones that are delivered with the regular seminal fluid (cf. reviews [7,8]).
The sites of traumatic injection are variable between-species and their location and within-species consistency can be informative about putative functions . First, injections can occur at more or less random positions on a recipient's body , then indicating that delivered fluids rather unspecifically target general physiological processes, for example via injection into the circulatory system. Second, injection sites may be consistent, but primarily determined by physical constraints, for instance when only certain body regions can be reached during mating [2,10]. These cases do not offer straightforward indications regarding the generality or specificity of injected substances. Finally, injection sites may be consistent in a given species despite a flexible injection organ that would allow reaching other body parts, indicating that the injected secretions indeed target specific organ systems. We are not aware of previous evidence for organ-specific injection in species exhibiting traumatic secretion transfer with flexible injection organs. However, some sea slugs with traumatic insemination inject semen directly into the female sperm storage organ , probably to circumvent sperm digestion, a fate sperm often face when transferred into a female reproductive tract that also contains sperm-resorbing mechanisms .
In the sea slug genus Siphopteron (Opisthobranchia, Gastropteridae; ), mating interactions are characterized by traumatic secretion transfer that occurs at species-specific injection sites, including different regions of the mating partner's foot, parapod (a wing-like protrusion covering the animal dorsally) and visceral hump [14–16]. This specificity occurs despite the fact that secretions are delivered via a stylet that is positioned on a flexible and extendable penile appendage . Traumatic injection occurs prior or parallel to regular insemination, where the penis is inserted into the partner's gonopore as the female receptive organ to deliver sperm [14,16]. Earlier observational suggestions that traumatic secretion transfer enforces copulatory roles  could not be confirmed experimentally . However, experimental exposure to different mating frequencies showed that elevated mating rates cause a decline in female fecundity .
We here describe the unique ‘cephalo-traumatic’ mating ritual of Siphopteron sp. 1. We continue by speculating about a putative neurophysiological function of traumatic secretion transfer based on the location and consistency of injection sites in comparison with four congeneric Siphopteron species.
2. Study species and methods
Siphopteron sp. 1 is a close relative of Siphopteron quadrispinosum. Both species superficially look similar (figure 1) but form reciprocally monophyletic groups on a concatenated molecular phylogeny using cox1 mtDNA, 16S mt rRNA and 28S rRNA (R. Lange & N. Anthes 2012, unpublished data) that are also distinct from other previously studied Siphopteron species [13–17].
Specimens were collected on SCUBA diving during field trips to Lizard Island, Queensland, Australia in 2009–2011 (Great Barrier Reef Marine Park Authority permit no. G09/30973.1 and Australian Government Department of the Environment, Water, Heritage and the Arts wildlife trade permits WT2009-2331, WT2010-10641 and WT2011-4924). The 2–4 mm long specimens were collected in 0.5–12 m water depth on coral sand. After collection, slugs were individually transferred into translucent 75 ml screw top plastic vials that contained 40 ml of 1 µm millipore-filtered seawater, and were cleaned and refilled every second day. Individuals were weighed to the nearest 0.1 mg wet weight after collection. Mating groups were size-matched whenever possible. Animals were released upon completion of the experiments or used for morphological and molecular genetic analyses that are not part of this study.
(a) Siphopteron sp. 1 mating behaviour
We established eight independent mating groups containing four individuals each (i.e. 32 individuals in total). Individual recognition within groups was based on distinct colour patterns. Starting one day after collection, individuals were grouped three times per day (at 09.00, 12.30 and 16.00) in one well of a 1.25 ml 6-well plate for 1 h or until ongoing copulations had finished; mating trials lasted 6 days, resulting in 18 trials per mating group. The following data were recorded during each mating trial: time when a mating interaction started, mating roles of each mating partner, times of penile bulb insertion and retraction, location and duration of penile stylet injection and mating duration. Most matings were also videotaped and photographed using a Canon Powershot G9 digital camera mounted on an Olympus SZ stereomicroscope. Individuals were kept at 26°C, natural day length, and no direct exposure to sunlight. To study potential effects of traumatic secretion injection on immediate egg laying behaviour, we checked for egg masses daily at noon for 10 consecutive days and counted the number of eggs per egg mass.
(b) Injection sites in the genus Siphopteron
For comparison, we assembled between seven and 16 mating groups of four individuals each of Siphopteron pohnpei , S. quadrispinosum , Siphopteron sp. 2 (= Siphopteron sp. in ), and Siphopteron sp. 3 (= Siphopteron sp. 1 in ), respectively. Mating trials were performed in the same manner as described earlier for Siphopteron sp. 1, but we here only report the data on stylet injection sites.
(a) Siphopteron sp. 1 mating behaviour
Across all 144 mating sessions, we observed 16 copulations in total; those involved 20 of 32 individuals, with single individuals mating up to three times. Larger individuals were significantly more likely to mate at least once (binary logistic regression, n = 32, χ12 = 7.732, p = 0.005). We collected 13 egg masses (from 13 different individuals), each containing 533±93 s.d. eggs on average. Whether or not individuals mated during the study affected neither their likelihood of egg laying (likelihood ratio test, n = 32, χ12 = 0.426, p = 0.513) nor the number of eggs they laid (Wilcoxon test, n = 13, Z1 = −0.694, p = 0.488).
A video showing the typical copulatory sequence is available in the electronic supplementary material. Similar to S. quadrispinosum, mating was initiated by intertwining (figure 3d), which lasted for 2.64±2.34 min (mean±s.d.). During intertwining partners everted their penises, and showed frequent pre-copulatory biting into the partner's posterior visceral hump (seen in total 10 times prior to copulation plus 10 times without subsequent copulation). Intertwining was usually followed by exclusively reciprocal penile bulb insertion (penile bulb depicted in figure 2a) into the female genital opening (figure 3d), which is situated beneath the right parapod. Penile bulb intromission commenced after an average mating latency of 15.13±13.69 min. Seconds thereafter, the penile stylet (depicted in figure 2b) was inserted into the mating partner's foreheads, where it usually stayed until close to the end of penis intromission (n = 32). In two instances, one partner retracted its stylet during copulation, but reinserted after 3 and 6 min, respectively. Stylet insertion always occurred in direct vicinity of the mating partner's eyes with the stylet deeply inserted (figure 3b,c and video in the electronic supplementary material). Fluids were transferred through the stylet into the cephalic shield in slow pulses throughout insertion (see video in the electronic supplementary material). Stylet insertion and fluid injection provoked no discernible reaction by the recipient, and both mating partners remained still during copulation. Penile bulb intromission lasted for 42.2±4.89 min on average. Penile bulb retraction was typically asynchronous between partners (14 of 16 cases, figure 3d), with a mean time lag of 6.07±8.90 min. We observed one instance of a mating circle, where three individuals reciprocally inseminated each other.
(b) Stylet injection sites in the genus Siphopteron
Our comparison of five Siphopteron species revealed substantial between-species variation: (i) in the exact location of stylet insertion, and (ii) in the degree of within-species consistency (figure 4 and table 1). Siphopteron sp. 1 and Siphopteron sp. 3 invariably targeted a single and well-defined position, namely the partner's forehead and next to the gonopore, respectively. Target sites were more variable in S. pohnpei, S. quadrispinosum and Siphopteron sp. 2, where the posterior right parapod or the anterior foot region represented the main sites of stylet injection.
The mating behaviour of Siphopteron sp. 1 represents, to the best of our knowledge, the first instance of traumatic mating where gland secretions are injected directly into the mating partner's head, a mechanism that we term ‘cephalo-traumatic secretion transfer’.
Transferred secretions were shown to derive from the prostate gland in S. quadrispinosum , the closest known relative of Siphopteron sp. 1, and this probably applies to all stylet bearing Siphopteron species. In other gastropods, prostate fluids are known to contain bioactive proteins functioning as allohormones [5,19]. Traumatic secretion transfer is generally thought to manipulate recipients , as found in blowflies [20,21], earthworms , salamanders  and snails . In other Siphopteron sea slugs, earlier work  detected overall costs of copulation when mating rates exceeded a threshold, but failed to link those directly to traumatic injection . The latter work also rejected a role of traumatic injection in manipulating female mating behaviour and short-term reproductive output, proposing manipulation of paternity success or the recipient's subsequent male mating behaviour as plausible candidates for further study. Similar to these earlier findings, reproductive output of Siphopteron sp. 1 slugs in our current study was not affected by copulatory activity. Combined, the available evidence does therefore not corroborate the hypothesis that traumatic secretion transfer in the genus Siphopteron manipulates the partner's fecundity, irrespective of whether this may be to its benefit (if containing reproductive stimulants) or disadvantage (if invoking sexual conflict).
We further found substantial variation among five Siphopteron species in the injection sites and the degree to which they are consistent within each species, allowing some speculation regarding their function . Siphopteron sp. 3 consistently injected next to the female genital opening; the secretions may therefore directly enter the sperm receiving or sperm digesting organs and there serve to increase donor paternity success, a manipulation previously described from stylommatophoran land snails [3,24]. By contrast, S. pohnpei, S. quadrispinosum and Siphopteron sp. 2 showed far lower consistency in target sites, offering two competing interpretations. First, secretions are injected into the body cavity to enter the circulatory system and affect the recipient's overall physiology. Second, the injections target specific organs but reach them equally well from different injection locations.
Siphopteron sp. 1 exhibited remarkably consistent traumatic injection into the forehead directly above the central nervous system. While we cannot entirely exclude that the fluids are simply injected into the body cavity, we consider such generalized injection unlikely: under this scenario, we cannot explain why the long, extendable and movable penile appendage used for traumatic injection (see video in the electronic supplementary material) is never inserted anywhere else on the partner's body. Even though the partner's forehead may be a particularly ‘convenient’ location for stylet placement or the result of some other constraint, this behaviour of Siphopteron sp. 1 differs strikingly from other Siphopteron species with a comparably long penile appendage; for three such species, our data document a much less specific selection of injection sites (table 1). Instead, we propose that the consistent site selection for penile stylet placement in Siphopteron sp. 1 is adaptive and directed towards the neural ganglia. Assuming that future studies indeed confirm ganglia as the target of traumatic injection, cephalo-traumatic secretion transfer makes us speculate about the central nervous system as a target for sophisticated neurophysiological manipulation (perhaps similar to that found in plethodontid salamanders ). First, in general, nervous control typically allows faster physiological responses than endocrine control , even though this needs confirmation in the context of reproductive manipulation. Second, direct manipulation of the central nervous system may be hard to counter-adapt to, because small changes here will probably affect many other essential systemic processes.
Astoundingly, fine-scaled external control over the central nervous systems is indeed known from many parasites (reviewed by Adamo ). They use their host's neurological adaptations to alter their own behaviour and physiology according to the perceived environment . Parasites thus achieve neuronal manipulation despite their genome differing massively from that of their hosts. Hence, the evolution of neural manipulations by conspecifics is clearly plausible, even more so in simultaneous hermaphrodites, which per definition have access to all possibly required sex-specific hormones .
The German Science Foundation (DFG) provided funds for fieldwork and manuscript preparation (AN549/2-1 and ZUK 63).
We thank Anne Hoggett, Marianne and Lance Pearce, and Lyle Vail from Lizard Island Research Station for continuous support during fieldwork. Carolyn Groves and Marissa Henderson assisted during fieldwork. Comments by Klaus Reinhardt and two anonymous referees helped to improve earlier versions of this manuscript.
- Received September 17, 2013.
- Accepted October 22, 2013.
- © 2013 The Author(s) Published by the Royal Society. All rights reserved.