The status of the indigenous Southeast Asian apple snails belonging to the genus Pila is of concern due to their fast rate of population decline, possibly as a result of multiple factors including habitat loss or disturbance and the introduction of globally-invasive apple snails, Pomacea spp. Conservation actions, including captive breeding of the native Pila species, have been suggested as urgent remedial practices, but the lack of knowledge regarding the fundamental reproductive biology of indigenous Pila spp. makes such practices difficult. In the present study, observations on the mating and egg-laying behaviour of an economic valuable apple snail native to Southeast Asia, P. virescens, were conducted using video recording to examine and describe their reproductive behaviour under a laboratory condition. A total of 15 types of mating and seven egg-laying behaviour were recorded. The mating sequence which subsequently resulted in egg laying was comprised of seven types of major sequential behaviour: mate probing, mounting, shell circling, positioning, insemination posture, sheath withdrawal and dismounting. Rejection of mating attempts by the female was frequently observed. Egg laying occurred during either day or night. A sequence of seven distinct types of behaviour were performed during oviposition: climbing, positioning, forming a temporary tube, mucous secreting, egg depositing, leaving and resting. Overall, these results provide an understanding of the egg-laying behavioural process and highlight its complexity in P. virescens. In addition, detailed ethograms of mating and egg-laying behaviour were derived. These will promote further systematic comparative studies of the reproductive behaviour of apple snails.

Ampullariidae, apple snails, behavioural observation, copulation, Pila, reproductive biology, Southeast Asia, spawning

The Southeast Asian apple snail Pila virescens (Dehayes, 1824) is distributed in several countries of Southeast Asia, including Thailand, Vietnam, Cambodia, Laos, Malaysia, Indonesia, Myanmar and Philippines (Keawjam 1986, 1990; Cowie 2015; Ngor et al. 2018; Ng et al. 2020). Historically, P. virescens was reported to occur in various freshwater habitat types, both standing and moving water bodies (as Pila polita in Keawjam (1986, 1990); Thaewnon-Ngiw et al. (2003a, b)). However, in the last few decades, many populations of P. virescens have declined and disappeared from where they were formerly known to be present (Ng et al. 2020). On the other hand, those habitats have often been invaded by apple snails from the genus Pomacea (Ng et al. 2020), two species of which have invaded globally, including in Thailand (Cowie et al. 2012). Moreover, in some areas, populations of P. virescens have entirely been replaced by Pomacea snails (Marwoto et al. 2020; Ng et al. 2020).

Although the direct impact of Pomacea snails on native Pila snails is unknown, many studies have indicated that populations of Pila spp. decline after the introduction of Pomacea spp. (Chaichana and Sumpan 2014; Ng et al. 2019, 2020; Marwoto et al. 2020). Habitat loss or disturbance and pollution, for instance, the use of molluscicides to remove Pomacea snails, may be additional factors that threaten native Pila species (Ng et al. 2020). Therefore, P. virescens is under strong threat. Although P. virescens is categorised in the ‘Least Concern’ category by the IUCN, this assessment was based on decades-old data that require updating (as Pila polita in Madhyastha and Sri-aroon (2012)). In contrast, more recent data suggests that P. virescens may be in risk of extinction (Ngor et al. 2018; Marwoto et al. 2020; Ng et al. 2020) and deserves urgent conservation.

Ng et al. (2020) suggested captive breeding of native Pila species from different localities should be encouraged in order to enhance conservation of wild populations as they are consumed widely in Southeast Asia, especially for P. virescens in Thailand (Keawjam 1986; Ng et al. 2020), where it is known as hoy pang. Such aquaculture by captive breeding may help decrease overharvesting of wild populations. However, culturing P. virescens is difficult (Keawjam 1986). The lack of knowledge on its reproductive biology is an important obstacle to captive breeding. Understanding the reproductive behaviour of this species would contribute to resolution of practical difficulties of captive breeding and conservation of P. virescens.

There have been several studies on the apple snail genus Pila in Southeast Asia, most of which mainly focused on the taxonomy of this group (Annandale 1920; Keawjam 1986, 1987a, b, 1990; Thaewnon-Ngiw et al. 2003a, 2004; Low et al. 2013; Tan et al. 2013; Ng et al. 2014, 2020; Cowie 2015), parasitology (Punyagupta 1965; Tesana et al. 2008; Komalamisra et al. 2009) and ecology (Chaichana and Sumpan 2014, 2015; Pradabphetrat et al. 2017; Ng et al. 2019). However, few details on their reproductive biology have been studied. The morphology and anatomy of the reproductive system and egg mass characteristics of five Pila species: P. celebensis, P. gracilis, P. pesmei, P. turbinis and P. virescens that are distributed in Thailand have been described (Keawjam 1987a; Ng et al. 2020). Egg mass characteristics, clutch size, incubation period and reproductive seasons were reported in P. celebensis (as P. ampullacea in Djajasasmita (1987)) and P. pesmei (Lamkom and Phosri 2017). For Southeast Asian apple snails, mating behaviour is only known for P. turbinis (as P. ampullacea in Panha (1985)). Egg and egg mass characteristics of P. virescens in the Philippines were described more than a century ago (Semper 1862). Thus, the biology, ecology and behaviour of reproduction of Pila in Southeast Asia, which are crucial to understand the evolution of diversity in apple snails (Hayes et al. 2009), mostly remain to be investigated. The aims of this study were to describe the mating and egg-laying behaviour of P. virescens, with a detailed ethogram of mating and egg-laying behaviour for future comparative studies amongst apple snails.

Based on the video recording analysis of 62 isolated mating pairs and 10 egg-laying occasions, detailed ethograms of the mating behaviour (Table 1) and egg-laying behaviour (Table 3) of the Southeast Asian apple snail, P. virescens, were extracted. These ethograms allowed us to identify the sequences of mating and egg-laying behaviour of P. virescens leading to an improved understanding of the reproductive process of this species. The results of the current study indicated high variation in the sequence of mating behaviour of P. virescens and revealed its previously unknown unique egg-laying sequence.

The mating process of P. virescens was classified into two types following the success of the male-mating attempt: a complete and an incomplete mating process. The complete mating process was divided into four phases: pre-courtship, courtship, copulation and post-copulation, similar to the mating process that was previously reported for the golden apple snail, Pomacea canaliculata (Burela and Martín 2009). The overall mating pattern: male approached female, mounted on the female shell, crawled on the female shell, adhered on the right side of the female’s last whorl in order to mate and dismounted, respectively, follows the general mating pattern described in other apple snails (Bahl 1928; Panha 1985; Burela and Martín 2009; Tiecher et al. 2014; Hayes et al. 2015; Gurovich et al. 2017); however, details of the mating behaviour performed in each mating phase were variable.

During the pre-courtship phase, two snails were introduced into the mating aquarium and so the first behaviour observed in this study was sheltering as a result of human manipulation. They sank to the bottom of the mating aquarium at random positions, but began moving shortly afterwards by slowly opening their aperture, extruding their tentacles and cephalopodium and then walked around. In most cases (40/62), snails moved directly to the water surface in order to breath. After that, the first contact between two snails happened while they were moving around in the aquarium. However, in a few cases (12/62), the snails sunk into positions which were close to each other. The first contact occurred between them immediately after they began moving. Consequently, the pre-courtship phase in this study does not represent mate-searching behaviour which may exist in P. virescens, as has been previously reported in the freshwater snails Pomacea canaliculata (Takeichi et al. 2007) and Biomphalaria glabrata (Trigwell et al. 1997). However, the result indicated that males very actively performed mating behaviour. Many males (53/62) performed mate probing after the first contact. This was commonly followed by courtship behaviour, such as shell mounting and shell circling.

The courtship phase began shortly after two snails were introduced into the mating aquarium following their first contact. The male mounted the female’s shell at inconsistent positions and crawled on it in either a clockwise- or counter-clockwise direction. Clockwise shell circling was frequently observed (198/327). The positioning and insemination posture (copulation phase) was mostly observed after shell circling in the clockwise direction. This is in contrast to previous reports in other freshwater snails that male frequently performed counter-clockwise shell circling (De Boer et al. 1996; Burela and Martín 2009). It has been assumed that shell circling is an assessment of the partner size in Pomacea canaliculata (Burela and Martín 2009) and Lymnaea stagnalis, (Koene et al. 2007). We are unable to verify this assumption as we did not examine the relationship between snail size and mating success. However, another or additional possibility is that males may perform shell circling in order to locate the position of the female gonopore as we noticed that males started circling from different initial locations on the female’s shell, but stopped circling at the right side of female’s last whorl where they could begin positioning and enter the copulation phase. In addition, more importantly, all occurrences of positioning behaviour observed in this study were performed after shell circling. This indicates that it is necessary for the male of P. virescens to perform shell circling for mating success.

The movement of the male snail on the female snail’s shell in order to reach the female gonopore has also been reported in Physa acuta (Ohbayashi-Hodoki et al. 2004) and Lymnaea stagnalis (Barraud 1957). Positioning took place when the male reached the right side of the female’s last whorl. The male adhered on the right side close to the rim of the female’s last whorl and then contracted its foot, slowly coiled its tentacles and moved slightly forward until its aperture overlapped with the female’s aperture. In the absence of mating rejection by the female, the pair then entered the copulation phase.

Mating rejection in apple snails was previously known only in Pomacea canaliculata. Burela and Martín (2009) reported three types of mate rejection behaviour in Pomacea canaliculata females: swinging, withdrawal-wrestling and operculum closing. In the same way, we also found these types of behaviour in P. virescens. In addition, we newly found two other types of behaviour, jerking and retracting, which could have prevented the mating process in P. virescens. Retracting was the strongest mate rejection observed in this study, as in most of the mating pairs, the courtship behaviour by a male ended after the female’s retraction without continuing to the copulation phase. However, retracting includes the two types of general retracting and strong retracting. General retracting is the retraction of a snail when it was touched by another snail and released soon after, which is possibly a common behaviour in this species in response to physical stimulus because it was commonly observed in both sexes when touching each other and it never led to cessation of mating behaviour. In contrast, in the case of strong retracting, retraction continued when it was touched by another snail until the two snails were separated from each other. This strong retraction was generally performed by the female and stopped mating attempts by the male in almost all cases (252/277). Similarly, different types of retraction have been reported in Lymnaea stagnalis: general retraction in response to disturbances and violent retraction of the cephalic part in response to penis sheath intromission that then disrupted the intromission of penis sheath in this species (Barraud 1957).

Withdrawal-wrestling and operculum closing are other strong rejections that stop mating in almost all the cases in which females showed these types of behaviour. Jerking and swinging were observed in many mating pairs and, frequently, both types of behaviour were performed. Although swinging was previously reported as a type of mate rejection behaviour in several freshwater snails (DeWitt 1991; Facon et al. 2006; Burela and Martín 2009; Tiecher et al. 2014), swinging in P. virescens resulted in mating failure in only a few couples. Likewise, jerking was reported as a type of mate rejection behaviour in Physa gyrina (DeWitt 1991).

Thus, mate rejection in P. virescens can be categorised as two levels: (i) strong mating rejection: retracting, operculum closing and withdrawal-wrestling and (ii) weak mating rejection: jerking and swinging. Most of the incomplete mating attempts that ended in the courtship phase and resulted from the female’s mate rejection behaviour, mainly by retracting or operculum closing and, to a lesser extent, by jerking, swinging or both. There were only a few cases (6/42) of female rejection of the male during the copulation phase. In the early stage of the copulation phase, females performed retracting, jerking, swinging and withdrawal-wrestling behaviour. Five females performed withdrawal-wrestling behaviour. Four of those ceased the mating sequence, whereas only two of 19 females that performed jerking and/or swinging behaviour ceased the mating sequence.

Copulation began after the male positioned itself on the right side of female’s shell and then introduced the penis sheath into the female gonopore. The sheath intromission process could not be observed in this study as the male mounted on the rim of female’s shell above her aperture that overlapped with the male aperture (Fig. 7). Thus, the part of the male shell which overlapped with the female aperture completely obscured the process of penis sheath intromission. Although males mounted at the right side of the female’s last whorl above the opening of the female gonopore, similar to that described in other apple snails, such as Asolene pulchella (Tiecher et al. 2014), P. globosa (Bahl 1928), Pomacea canaliculata (Andrews 1964; Albrecht et al. 1996; Burela and Martín 2009) and Pomacea americanista (Gurovich et al. 2017), their penis sheath was invisible throughout the copulation phase, whereas the penis sheath of other apple snails was visible. Thus, the insemination posture of P. virescens is possibly different from that in other apple snails. Although penis sheath intromission could not be seen, we recognised the copulation phase, based on the combination of the male insemination posture and the female reaction in response to penis sheath intromission – a few minutes after the male took the insemination posture, the female showed a strong retraction followed by releasing and entering the passive stage. This behavioural procedure occurred in all mating pairs from which females subsequently laid eggs. It is, thus, reasonable to assume that this repertoire is a response to penis sheath intromission.

Figure 7. 

Insemination posture of P. virescens. Abbreviations: lp, labial palps; oc, operculum; si, siphon; t, tentacle.

During copulation, the male was passive, without movement, adhered firmly on the female’s shell with coiled tentacles and moved only its siphon for breathing, similar to the type of behaviour reported in other freshwater snails (Burela and Martín 2009; Soldatenko and Petrov 2012; Tiecher et al. 2014; Gurovich et al. 2017). In contrast to males, females performed various types of behaviour, such as passive, moving, jerking and swinging. Females spent most of the time during copulation in a passive state. They could move around in the mating aquarium. However, in almost all cases, those movements were to crawl up to the water surface in order to breath. Jerking and swinging behaviour performed in copulation phase was indistinguishable actions from those that rejected courtship in some cases. Burela and Martín (2009) reported swinging in Pomacea canaliculata only during courtship and early copulation (penis sheath intromission). We observed that some P. virescens females performed jerking and/or swinging several times throughout the copulation period while they were passive or moving. At the end of copulation, the male withdrew the penis sheath from the female gonopore, which we recognised, based on the movements of the male combined with the female’s retracting.

After penis sheath withdrawal, some males (11/20) still remained mounted on the female’s shell and crawled or passively adhered on to it for few minutes before dismounting, whereas other males immediately dismounted from the female’s shell. The behaviour of males that continued to adhere on to the female’s shell after penis sheath withdrawal was described as mate guarding behaviour in Pomacea canaliculata (Burela and Martín 2011).

Oviposition took place several days after copulation, but was widely variable, ranging from 4–51 days (average 14 ± 15 days, n = 10), which is much longer than that within the 1 or 2 days after copulation reported in P. globosa (Bahl 1928). Females laid white spherical eggs in one mass on the aquarium wall at various positions, but mostly near the water surface. In nature, P. virescens lays eggs on various substrates (e.g. on the ground, floating wood and plants) away from the water (Semper 1862; Ng et al. 2020). The average clutch size was 104.3 ± 54.0 eggs (range from 50–205 eggs, n = 10), which is similar to clutch sizes of 70–80 and 100–150 eggs per mass collected from natural habitats (Semper 1862; Ng et al. 2020).

Oviposition is known to be frequent during the night or early morning in Pomacea canaliculata (Albrecht et al. 1996; Heiler et al. 2008; Gilal et al. 2015), Pomacea americanista (Gurovich et al. 2017) and P. globosa (Bahl 1928). Nocturnal oviposition was suggested to be an avoidance from predators, heat and desiccation (Estebenet and Martín 2002; Gurovich et al. 2017). Oviposition of P. virescens occurred during both the day and night in this study conducted indoors. It is possible that different conditions may be the reason for the observed difference in P. virescens; however, further research would be necessary to explore the differences. In addition, we observed that P. virescens in collective tanks in outdoor conditions frequently laid eggs after rain. This suggests that P. virescens may naturally lay eggs after rain as an alternative means to avoid heat and desiccation. This could similarly be the case in other apple snails that have been known to lay eggs typically in the night-time.

The egg-laying behavioural process of P. virescens is similar to that of P. globosa (Bahl 1928). However, P. virescens was never observed to dig a hole to deposit its eggs as was reported in P. globosa. In addition, several new types of behaviour were identified and reported for the first time in this study. The egg-laying process was comprised of seven major types of behaviour (Table 3), which could be split into the three phases of pre-egg depositing, egg depositing and post-egg depositing. The first sign of oviposition is climbing. In the same manner as other aerial egg-laying apple snails (genus Pomacea and Pila) (Bahl 1928; Andrews 1964; Cowie 2002; Gurovich et al. 2017), P. virescens females move above the water to lay eggs. It was reported that P. globosa forms the foot as a dome shape to arrange eggs in a compact mass (Bahl 1928), whereas we found that P. virescens contracts the entire foot forming an oval or heart shape (seen from the ventral view) instead of the dome shape. Subsequently, the right edge of the foot forms a tube shape used for transferring the eggs from the gonopore to the aquarium wall. Formation of this temporary egg transferring tube has also been reported in P. globosa (Bahl 1928). In other genera (e.g. Pomacea and Asolene), females form a nuchal groove at the right side of the dorsal foot surface for transferring the eggs from the gonopore to the substrate (Andrews 1964; Albrecht et al. 1996; Tiecher et al. 2014; Gurovich et al. 2017). The formation of an egg transferring tube in apple snails has only been reported within the genus Pila which may suggest the evolution of distinct egg-laying behaviour in this genus.

After forming the egg-transferring tube, instead of depositing eggs directly on the (glass) substrate, P. virescens secreted a smear of white milky mucus. The mucus appeared from the female gonopore and was then transported through the egg-transferring tube and became smeared on the aquarium wall by moving the ventral foot surface. This process has never been reported before in any apple snails. Eggs were laid one by one, appearing at the opening of the gonopore and transported through the egg transferring tube to the middle area of the ventral foot surface, similarly to that described in P. globosa (Bahl 1928). Subsequently, eggs were transported from the middle to the top-left edge of the ventral foot surface. Instead of laying flaccid eggs directly as reported in other apple snails (Bahl 1928; Andrews 1964; Albrecht et al. 1996; Gurovich et al. 2017), P. virescens laid up to the first 10 eggs with the milky mucus and the later eggs with little or no milky mucus. The egg mass was formed by arranging the eggs at top-left edge of the ventral foot surface and then expanding to the middle and lower edge. If egg mass grew larger than the ventral foot surface could hold, the female moved slowly backwards and simultaneously continued egg laying.

The behaviour of P. virescens after leaving the water until laying the last egg was consistently observed amongst egg-laying events of different females. The behaviour varied only during the post-egg-deposition phase. Two females adhered on the egg mass, one for 15 minutes and the other for 23 minutes, after laying the last egg and then left the egg mass by slowly sliding backwards into the water. Three females immediately dropped into the water after laying their last egg, while the rest (n = 5) slowly slid backwards to the water after laying the last egg. These post-egg depositing variations were also reported in Pomacea canaliculateta (Albrecht et al. 1996). Resting and refreshing seem to be important for females after finishing laying eggs. Almost all the females (n = 9) entered the resting stage, possibly resulting from energy expenditure for laying eggs.

Behavioural characteristics identified in this study increase our understanding of the diversity of reproductive behaviour in a Southeast Asian apple snail. Although types of mating behaviour largely followed those previously recorded in South American confamilial snails (Burela and Martín 2009; Tiecher et al. 2014; Hayes et al. 2015; Gurovich et al. 2017), our results reveal some variations. In addition, our results on egg-laying behaviour highlight the distinct aerial egg-laying sequence in P. virescens. Further research, including comparative studies on reproductive behaviour of different species would provide new insights into the adaptation and evolution of the reproductive process of apple snails (Hayes et al. 2015).

We thank members of Animal Systematics Research Unit of Chulalongkorn University for assistance in the laboratory; and reviewers for valuable suggestions that improved the manuscript. We acknowledge the Thailand Research Fund for financial support through the Royal Golden Jubilee Ph.D. Program (Grant no. PHD/0206/2559) to SA and SP and National Research Council of Thailand & Centre of Excellence on Biodiversity (N35E660138).