Female control of mate plugging in a female-cannibalistic spider (Micaria sociabilis)

Mating behaviour

After first contact, the males chased the vigorously moving females and tried to mount
them. Spiders mated in the type III posture 18], meaning that the male approached the female from the front, climbed on the female
prosoma, and reached around her opisthosoma to insert one of his pedipalps. His right
palp was inserted into her left copulatory duct or vice versa. When the male achieved
the correct mating position the female always continued to move, on average for (mean?±?SE)
8.7?±?1.9 minutes (N?=?16). Attempts to couple the pedipalp to the female genital
plate (i.e. genital coupling) were always accompanied by palpal chewing (the male
moved a pedipalp between his chelicerae), brief pedipal coupling (i.e. flubs), and
lateral palpal movements (for a description of these behaviours, see Additional file
1), all occurring repeatedly before the coupling of the pedipalp to the female genitalia.
Once the pedipalp was coupled to the female genitalia, rhythmic full expansions of
the membranous part (haematodocha) of the sperm transfer organ followed. During copulation,
the pedipalp was coupled for an average of 77.7?±?13.0 min (N?=?16). Before mating
ended, a sequence of brief haematodochal expansions was performed in 50% of cases
(N?=?16).

In 74.3% of cases (N?=?35), copulation ended after a single insertion. Accordingly,
both pedipalps were used in succession in only 25.7% of cases. The copulation ended
when the female began to move vigorously or when the male left the female without
any previous struggling. In the latter case we assume male driven termination. The
sequential behaviours observed during mating trials are listed and defined in the
Additional file 1.

Mating trials with females without plugs

Descriptive data on the mating procedure with females without a previous plug are
given in Table 1. Immediately after mating, a complete plug (i.e. covering the whole atrium) was found
in 40% of cases (N?=?35). In 20% of these cases an incomplete plug was found, i.e.
plug material was found in the atrium in various amounts but it did not cover both
copulatory openings. In the remainder of the cases (40%) no plug material was found
in the atrium of the female. The plug material appeared no earlier than after 37 minutes
of copulation. The copulation lasted on average (mean?±?SE) 29.4?±?19.7 min when no
plug was produced, 97.9?±?25.7 min when an incomplete plug was produced, and 137.1?±?21.6 min
when a complete plug was produced. Right after copulation the plug resembled a whitish,
gelatinous substance. After one day, the substance had hardened and had the appearance
of a black solid mass. One day after copulation, only complete plugs persisted in
the female genital atrium, whereas incomplete ones had disappeared.

Table 1. Descriptive data on separate mating trials with females without plugs and plugged
females and statistical comparison of the mating behaviours between the two groups

Analyses of the video recordings (N?=?16) revealed that plugs were probably produced
by a sequence of brief haematodochal expansions at the end of the mating process.
These expansions followed the period of distinct full haematodochal expansions and
occurred shortly before genital uncoupling. The duration of one brief expansion was
less than a second, while long haematodochal expansions lasted on average 2.6?±?0.4 s
(N?=?16). The number of brief expansions in a sequence was on average 6.1?±?1 (N?=?16).
Brief haematodochal expansions were observed in half of the video recorded trials
(N?=?16) and when these expansions occurred plug material was found in the female
atrium (N?=?8). In the rest of the trials (N?=?8) no brief expansions were observed
and no plug material was detected in the female atrium. The amount of plug material
in the atrium was significantly affected by the presence of brief expansions (ANOVA,
F_{1, 14}?=?36.2, P??0.001), but did not significantly depend on their number (Spearman’s r?=?-0.44,
P?=?0.27).

The amount of plug material found in the female atrium was related to increasing time
in the mating posture (GLS, F_{1, 33}?=?26.4, P??0.001, Figure 1), which was negatively affected by female restlessness (GLM-g, F_1,14?=?14.4, P?=?0.001). Neither male size (GLS, F_{1, 27}??0.1, P?=?0.89) nor female size (GLS, F_1,27??0.1, P?=?0.99) nor the size ratio in the pair (GLS, F_{1, 23}?=?0.04, P?=?0.83) taken as PCA1 scores affected the occurrence or size of the plug.
The amount of plug material was negatively affected by the duration of the genital
coupling phase (N?=?12, r_s?=?-0.77, P?=?0.003, Figure 2). When males were able to couple their pedipalp quickly, a complete plug was found
after mating. When they fixed the pedipalp later, plugs were incomplete or missing.
After males had mounted the females, the females moved for 8.7?±?1.0 minutes on average.

Figure 1. Plug quantity (proportion of plug material in the female atrium) in relation to time
in mating position. Estimated logit model is shown.

Figure 2. Plug quantity (proportion of plug material in the female atrium) in relation to duration
of palp fixation. Estimated logit model is shown.

Mating trials with plugged females

Descriptive data on the mating procedure with plugged females are given in Table 1.

Removal of a complete plug only occurred in 7% of cases (N?=?42), and was not accompanied
by conspicuous male behaviour. In two of these cases, full haematodochal expansions
occurred after removal of the plug, followed by plug production in one case (N?=?24).
In the remaining cases, partial haematodochal expansions (i.e. flubs) were observed,
which did not result in plug production.

As described above (in the previous section), females who had an incomplete plug in
the atrium after copulation (N?=?10) bore no plug after one day. In subsequent mating
trials, after three days, the second males who achieved copulation (N?=?9) used the
same palp and thus the same copulatory duct as the previous male in 56% of cases and
long haematodochal expansions were always performed.

Comparison of mating trials with females with and without plugs

The behavioural components of the mating procedure observed in the experiments described
above were compared between the two groups (Table 1). The probability of males mounting the female did not differ between trials with
females with and without plugs. However, the length of time the couple remained in
the mating posture was significantly shorter in plugged females than in females without
a plug. When paired with plugged females, males performed significantly fewer insertions
and performed significantly more lateral palpal movements and palpal chewing movements
per minute than with females without plugs. Long haematodochal expansions occurred
in all trials with females without plugs, but only in two cases in trials with plugged
females and only after removal of the plug. In the rest of the cases (N?=?22) only
partial haematodochal expansions were observed (i.e. flubs).

Terminations of copulations with females without plugs were preceded by the female
struggling vigorously, which resulted in the male being shaken off; only in one case
did the male end the copulation (N?=?35). Interactions with plugged females were ended
by the female in 52% of cases (N?=?42); in the remaining 48% of cases, males ceased
attempting to fix their palp after 9.7?±?1.5 minutes (N?=?20) and left the females.
Therefore, males were significantly more likely to end interactions when trying to
mate with plugged females (Table 1, Figure 3).

Figure 3. Comparison of relative frequency of males and females terminating the copulation in
experiments with females with and without plugs.

Male genitalia

Males of M. sociabilis possess a relatively simple copulatory organ consisting of an apically tapered, spoon-like
cymbium and an egg-shaped genital bulb (palpal organ) (Figure 4A). The actual sperm transfer structure, the embolus, projects from the genital bulb.
The tip of the embolus is bent and bears an opening (Figure 4B) through which the ejaculate is released during copulation. The embolus opening
connects to a closed tube, the spermophor, which is coiled inside the bulbus and in
which the seminal fluid is stored (Figure 4C).

Figure 4. Morphology of the male genital bulb ofM. sociabilis. A – B: SEM; C: Volume and surface reconstruction. A. Ventral view of genital bulb. B. Detail of the bulbus gland opening with remnants of plug material (arrow). Asterisk
marks tip of the embolus bearing spermophor opening. C. Ventral and prolateral view of genital bulb and its longitudinal section. The gland
reservoir encircled by loops of spermophor opens up below embolus (arrow). Cy – cymbium,
E – embolus, RTA – retrolateral tibial apophysis, T – tegulum, Ti – tibia.

Two glands were found in the genital bulb: a spermophor gland (Figures 5A, B) and a bulbus gland (Figures 4C, 5B, sensu Suhm et al. [29]). The spermophor gland is attached to the spermophor and runs along
its wall. The massive bulbus gland takes a central position in the genital bulb, surrounded
by the loops of the spermophor (Figures 4C, 5B). A single-layered epithelium surrounds the lumen of the bulbus gland (Figure 5B). Both glands exhibit numerous microvilli (Figures 5C, D). The suspected plug material is present inside the bulbus gland lumen.

Figure 5. Morphology of the male genital bulb ofM. sociabilis. A – B: LM; C – D: TEM. A – B: Semi-thin cross sections of the genital bulb showing duct (A) and reservoir (B) of the bulbus gland. Asterisk marks the homogenous plug material in duct lumen. C – D: Ultrathin section showing secretion in the lumen of the bulbus gland. BG – bulbus
gland, BGL – bulbus gland lumen, bH – basal haematodocha, E – embolus, Mi – mitochondria,
Mv – microvilli, N – nucleus, S – spermophor, Sec – secretion, SF – seminal fluid,
SG – spermophor gland, SGL – spermophor gland lumen, V – vesicles.

The spermophor gland releases a secretion into the spermophor lumen through its porous
wall. Even though the bulbus gland is surrounded by the spermophor, it is not connected
to it (Figures 5A, B). The glandular lumen protrudes into a short duct which opens at the base of
the embolus (Figures 4B, C). The opening is slit-like and covered by membranous folds of the apical part
of the tegulum. The SEM investigation of the pedipalp revealed remnants of plug material
covering the slit (Figure 4B).

Female genitalia

In females of M. sociabilis, the epigynal plate forms a trapezoidal cavity, an atrium, with two laterally-situated
copulatory openings (Figures 6A, B). The openings lead to copulatory ducts, each connecting to a tubular spermatheca
(Figure 6A). Near the proximal region of the copulatory duct, the fertilization duct leaves
the spermatheca in the direction of the uterus. In all inspected plugged females,
the plug material covered the whole atrium (N?=?5 and N?=?51 from the two experiments,
respectively, Figure 6A) and extended also to both copulatory ducts up to the spermathecae, as revealed
by X-ray microscopy (Figure 6A) and histology (Figure 7).

Figure 6. Morphology of female genitalia ofM. sociabilis. A. Ventral and dorsal view of surface reconstruction of female epigynal plate with
and without plug. B. SEM of female epigyne without plug. Arrows show copulatory openings. CD – copulatory
duct, FD – fertilization duct, Spth – spermatheca.

Figure 7. Semi-thin cross sections of plugged female genitalia ofM. sociabilis.A. Schematic drawing of the female genitalia with a line showing the position of the
cross section. B. Cross sections. Atrium and copulatory ducts are filled with plug material. Inset
shows detail of spermatheca filled with seminal fluid. CD – copulatory duct, P – plug,
Spth – spermatheca.