A time-lapse observational study of the activity patterns and feeding behaviour of urban terrestrial gastropods on carcasses of the Asian house shrew (<i>Suncus murinus</i>)

Urban land gastropods, snails and slugs, particularly introduced species, are notorious for causing damage to crops and plants. They are also known for their adaptability and ability to survive in unlikely environments, often feeding on other invertebrates and animal faeces (Speiser, 2001). Despite this, their roles as scavengers on animal carcasses have received relatively little attention in the feeding ecology of land molluscs and forensic science (Galvão et al., 2015). Previous observations of land gastropods scavenging on animal carcasses have been opportunistic and limited to a single point in time. Our knowledge of possible interactions with other species, particularly other land molluscs, and feeding activity patterns, including scavenging on carcasses at different stages of decomposition, remains poor. Although the scavenging behaviour of land gastropods on animal carcasses is known (Graham, 1955; Zamora & Gomez, 1996), data are available only for particular mollusc species scavenging on animal carcasses, mainly amphibians (Tan, Lim & Chua, 2015; Pustilnik, 2020; Yadav et al., 2021; Ayres, 2022; Kalki, Rittenburg & Pierson, 2022). While snails have been recorded on mammalian carcasses, the exact roles of these snails are not well understood (Payne, 1965; Galvão et al., 2015). This study aimed to observe the activity of five species of land gastropods near and on the carcass of an Asian house shrew (Suncus murinus) over an 80-h period in an urban environment. The activity patterns of the land gastropods varied over the observation period, and this is the first recorded scavenging event involving five scavenging mollusc species and a small mammal carcass. On the morning of 2 June 2020, a carcass of an Asian House Shrew (S. murinus) was found by the second author in the garden of a house in Ipoh, Perak, Malaysia (4°36′N, 101°6′E) (Fig. 1A). The total body length of the adult shrew was c. 150 mm and was presumed to have been killed by a cat a few hours before the discovery. A camera was set up to record the time-lapse of scavengers and decomposers visiting the carcass (Fletcher & Baylis, 2020). Photographs were taken at 3-min intervals between 11.33 on 2 June 2020 and 19.58 on 5 June 2020, for a total duration of 80 h (Fig. 1A–D). The project focuses on the biodiversity of invertebrates involved in the decomposition of the shrew. As the shrew was not intentionally killed for the purposes of the project and its death was due to circumstances beyond our control, ethical concerns regarding animal welfare are not relevant in this case (DVS, 2019). Time-lapse photos showing the decomposition process of an Asian House Shrew (Suncus murinus) carcass during the observation period of this study. A. Fresh/bloated stage. B, C. Decaying stage. D. Dry stage. Photographic equipment consisted of a Nikon D7200 camera protected from the weather in a 4.5-l Lock & Lock plastic storage box fitted with a 72-mm UV filter window. Exposures were manually fixed at 1/250 s and f16 at ASA800, with illumination provided by an off-camera Nikon SB-700 Speedlight at 1/16 power triggered wirelessly by a Vello FreeWave Fusion Pro. Images were taken in RAW (6,000 × 4,000 pixels) format, converted to JPG (Nikon Capture NX-D) and deflickered (GBDeflicker 2.5.1.0) before being cropped to 2,000 × 1,150 pixels (aspect 16:9) for this study. The entire carcass was included in the centre of each photo. Due to unavoidable interruptions during the time-lapse photography, such as changing camera and flash batteries or flash malfunctions, a total of 124 photos were missing from the 1,608 photos that could be potentially captured in total for the study duration (Supplementary Material Table S1). Almost all the missing photos were for 3 June 2022 (43 photos, c. 9% of total data for the day) and 5 June 2022 (75 photos, c. 20% of total data for the day) (Supplementary Material Table S2). It should be noted that almost all cases of missing images occurred during daylight when the snails and slugs were inactive. Photographs were examined by the first author to collect data on the land gastropod species, including the number of individuals of each species that made direct contact with the carcass and those that appeared around the carcass but did not directly come into contact with it. The examination and annotation of the photos were done using LabelImg v. 1.8.1 (LabelImg, 2022), and annotations were saved as XML files in PASCAL VOC format. Each annotation that represented an individual slug or snail was extracted as an independent record, with attributes including the species name, whether the individual made direct contact with the carcass (e.g. ‘Lissachatina fulica—animal’) or appeared in the photo but did not come into contact with the carcass (e.g. ‘Lissachatina fulica—grass’), the date and time the photo was taken and the photo filename. For example, as shown in Figure 2A, six records of land gastropods were obtained from the same photo and saved in an XML file for the photo. Then, the data in the XML file were extracted and tabulated, as shown in Figure 2B, using a customized R script. Annotation and extraction of data. A. The programme LabelImg was used to annotate the images, and the annotated data were saved in an XML file. B. The data in the XML file were tabulated for further time-series analysis of the occurrence of land gastropods on and around the shrew carcass. Time-series activity patterns in terms of the number of individuals captured in each image were plotted against the date and time, separately for each species recorded on the carcass and around the carcass. We also matched the occurrence of land gastropod species with four stages of decomposition. The names and definitions for these are as follows (Reed, 1958; Al-Mekhlafi, 2021): ‘fresh’, stage until bloating occurs; ‘bloated’, duration of bloated stage; ‘decaying’, stage between dissipation of bloating and drying; ‘dry’, stage when carcasses are dry and there is no evidence of insect presence. The study observed five species of land gastropods, two of which, Bradybaena similaris and Tanychlamys indica, were rarely found near the carcass, with B. similaris appearing in only 16 images and T. indica appearing in just one image out of 1,482 images taken over 80 h (Supplementary Material Table S3). Therefore, only the activity patterns of the remaining three species will be described in detail. The most frequently recorded species throughout the observation period was Laevicaulis alte, the only slug species, which appeared in 439 images, 210 of which were on the carcass. This was followed by Subulina octona, which appeared in 255 images, 160 of which were on the carcass, and then L. fulica, which appeared in 125 images, 81 of which were on the carcass. All three species were active during the night, from around 20.00 to 07.00 the next day. However, none of the species were found on the carcass after the second night (Fig. 3). The activity patterns of these three species varied over the observation period. The number of individuals of each species on the carcass showed a steep decrease for L. alte and S. octona after the first night, and neither of these species were found on the carcass after the second night (Fig. 3). The activity patterns of L. fulica on the carcass did not show clear changes for the first two nights, but, similar to the other two species, none were recorded on the third night. However, despite no land gastropods being found on the carcass on the third night, they were still active on the surrounding grass over the whole observation period. Activity patterns of three land gastropod species captured in time-lapse photos taken at 3-min intervals between 11.33 on 2 June 2020 and 19.58 on 5 June 2020, on and around the carcass of an Asian House Shrew (Suncus murinus). The four decomposition stages, as based on the definitions by Reed (1958) and Al-Mekhlafi (2021), are annotated in the chart. A. Subulina octona. B. Laevicaulis alte. C. Lissachatina fulica. This study provides rare time-lapse data. As noted by Speiser (2001) and Bailey & Wedgwood (1991), direct observation using time-lapse is the most straightforward method to determine the feeding behaviour and food preferences of land gastropods. All five land gastropod species recorded in this study were common urban species, and all are not native to Malaysia (Lim & Heyneman, 1965) but are also found worldwide (Cowie et al., 2008; Raheem et al., 2014; Nurinsiyah et al., 2016). These species are typically known to feed on plant matter, fruits and other forms of organic waste, and are known as agricultural pests. Of the three species found on the shrew carcass in this study, L. fulica is one of the top 100 invasive species in the world, and its impact on agricultural crops and plants is well known (Lange, 1950; Raut & Ghose, 1983; Santos et al., 2018; Pratiwi et al., 2022). It is omnivorous in its feeding habits and was observed as a scavenger, preferring human and animal faeces (Lange, 1950). Subulina octona is the second recorded species that scavenges on fresh meat (i.e. shrew carcasses in this study) and was previously recorded gathering in large numbers, with almost 400 individuals in a pitfall trap baited with lamb meat (Schilthuizen, 2011). Lastly, this is the first recorded instance of L. alte scavenging on small mammal carcasses, as it is mainly known to feed on plant materials (Raut & Panigrahi, 1990). The study also found that different land gastropods species may differ in feeding behaviour, and this may be influenced by competition for resources. Subulina octona are small (c. 10 mm in shell height), so they may not compete directly for space with larger land gastropods feeding on the same carcass. However, L. fulica and L. alte could potentially compete with each other, as they are relatively larger in size and were observed to stay on the carcass for up to 4 h. The state of decomposition of the shrew carcass may trigger or deter scavenging behaviour in land gastropods. It is worth noting that small mammal carcasses go through several stages of decomposition, and the quality of food sources may change over time, affecting the attractiveness of the carcass to land gastropods (or other scavengers). For example, the emission of volatile organic compounds, released from the carcass at different stages of decomposition (Dekeirsschieter et al., 2009), may act as attractants or cues for the gastropods, triggering their response and causing them to forage or engage in other related activities. On the third night, the presence of the snails and slugs on the grass in proximity to but not on the carcass suggests that decomposition may have rendered the carcass unpalatable. Further studies of the specific chemical compounds and their effects on gastropod behaviour could provide valuable insights into the mechanisms underlying the attraction of slugs and snails to decaying animal material. Overall, these findings suggest that the activity of land gastropods in scavenging on carcasses is complex and influenced by multiple factors, including the state of decomposition of the carcass and competition for resources among different mollusc species. Further studies could explore the feeding behaviour of different species of land gastropods in more detail to gain a better understanding of their roles in urban ecosystems. Finally, it is worth noting that many of these species, including L. fulica (Lim & Heyneman, 1965; Lim, Ow-Yang & Lie, 1965; Lim & Omar-Ahmad, 1969, 1970; Lim et al., 1976), L. alte (Lim & Heyneman, 1965; Lim et al., 1965, 1976; Lim & Omar-Ahmad, 1969; Lim, 1970) and S. octona (Alicata & McCarthy, 1964; Alicata, 1965; Lim & Heyneman, 1965; Lim et al., 1965), act as intermediate hosts for metastrongyloid nematodes. These nematodes are known to be ingested by land gastropods that feed on rat faeces (Cowie, 2013). However, our study suggests that there may be an additional way for these nematodes to be transmitted from animal carcasses to land gastropods. We thank Junn Kitt Foon and Mohd. Effendi Marzuki for the discussion. We also thank an anonymous reviewer for helpful suggestions. The authors declare that they have no competing interests. Data are uploaded as supplementary files.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.