Mantises are among the most iconic ambush predators in the insect world. Roughly 2,400 species are known globally, and South Africa alone hosts 157 species in 64 genera across 14 families. They typically rely on crypsis and specialized raptorial forelimbs to snatch smaller invertebrates that wander within reach. Yet the order has a well-documented taste for the unusual: mantises have been recorded eating small frogs, lizards, snakes, mice, fish, and even small birds. The Afrotropical species Polyspilota aeruginosa has a particularly colorful record, including a documented encounter with a juvenile songbird and startle-threat displays used to deter an approaching dark-capped bulbul. What remained missing from this catalogue of dietary audacity was any confirmed instance of a mantis eating a snail or slug in the wild.

The new study changes that. The researchers observed a Polyspilota aeruginosa nymph feeding on a silky tailwagger snail, Kerkophorus inunctus, in foliage at St Lucia. Both predator and prey were collected and accessioned into the KwaZulu-Natal Museum, where the snail was identified by museum malacologist Dr. I.V. Muratov using standard taxonomic keys. To test whether the behavior was a one-off accident, the team also reared a P. aeruginosa nymph in captivity and offered it common garden snails, Cornu aspersum. The captive mantid readily accepted the snails, grabbing each one before it could fully retract into its shell, securing it firmly with its raptorial forelegs, and maintaining constant pressure to prevent escape. Feeding focused on the exposed tissue, and once that was consumed, the mantid discarded the snail and its shell entirely.

Wild observations were not limited to that single St Lucia encounter. Of nearly 1,200 iNaturalist observations of Polyspilota mantises in South Africa as of February 2026, five showed mantises feeding on snails or slugs, spanning three P. aeruginosa individuals, one male P. caffra in George in the Western Cape, and an undetermined Polyspilota nymph in Kelso, KwaZulu-Natal. These citizen science records came from urban gardens, vegetation patches between agricultural fields, and semi-natural urban green spaces, suggesting that gastropod predation may be a recurring rather than exceptional behavior within the genus. The study also documents two striking vertebrate kills: a female P. aeruginosa photographed capturing and consuming a tropical house gecko in Leisure Bay in May 2025, and, four days earlier, a Polyspilota cf. caffra female catching, dismembering, and eating a marbled leaf-toed gecko. The researchers note this appears to be the first documented case of reptile predation by P. aeruginosa in South Africa.

Beyond the feeding records, the study assembled a comprehensive picture of where these mantises live. The team compiled 466 records of Polyspilota from South Africa, Eswatini, and Lesotho, drawing on museum collections from eight South African institutions, the Global Biodiversity Information Facility, and iNaturalist. Most records, 351 of them, came from citizen science observations, while 101 originated from museum specimens and 14 uniquely from GBIF, including seven DNA sequences in the BOLD database. Of the iNaturalist observations, 140 could be identified as P. aeruginosa and 73 as P. caffra, with another 138 confirmed only to genus because photo angles prevented reliable measurement of the diagnostic characters.

That diagnostic work relied on a morphometric approach. The principal feature separating the three South African Polyspilota species is the relative length of the pronotal metazone, expressed as a ratio to the prozone, the front portion of the pronotum. The metazone-to-prozone ratios clearly separated P. aeruginosa from P. caffra without any overlap between the sexes: males of P. aeruginosa ranged from 2.73 to 3.08, females from 2.50 to 2.79, while P. caffra males ranged from 2.18 to 2.41 and females from 2.05 to 2.31. The ratio of pronotum length to width provided additional discrimination. A third species, P. magna, is known only from a single female holotype collected at the Cape of Good Hope, and none of the museum specimens or iNaturalist records examined in this study matched its description. The authors note it remains unclear whether P. magna is a valid species at all, since the original description may have been based on a simple anatomical anomaly.

The mapped distributions reveal a striking biogeographic pattern. Polyspilota species are widespread along the subtropical eastern seaboard and in low-lying regions east of the eastern Great Escarpment, with only five records from the northeastern interior around Gauteng. The two better-known species are largely allopatric: P. aeruginosa dominates the northeastern coastal region of KwaZulu-Natal and the northern parts of Limpopo and Mpumalanga, accounting for 70 percent of its records in KwaZulu-Natal, while P. caffra is concentrated in the southeastern and southern coastal regions of the Eastern and Western Cape, with 87 percent of its records from those provinces. Remarkably, the boundary between the two species corresponds closely to the Umtamvuna River gorge on the KwaZulu-Natal–Eastern Cape border, a recognized biogeographical barrier that limits gene flow in a range of taxa. Vegetation differences may reinforce the split: KwaZulu-Natal supports more extensive and continuous forests, including scarp, mistbelt, and coastal-dune forests, thanks to higher rainfall, whereas the Eastern Cape holds smaller, fragmented forest patches within a landscape of grassland and thicket. Most P. caffra records come from the Western Cape, a region dominated by the Fynbos biome.

Why would a mantis eat a snail at all? The answer may lie in the peculiar sensory rules of mantid predation. Mantises are sit-and-wait hunters that strike in response to visual cues, particularly prey movement, and they attempt captures only when several stimulus parameters are met, including prey size, direction of movement, and apparent speed. Slow-moving targets generally elicit fewer strikes, which makes snails and slugs, on the face of it, poor candidates for mantid dinners. Previous laboratory work across six mantid species found that P. aeruginosa was the most generalist feeder, accepting 75 percent of all prey species offered, including two slug species and one snail, while the other five genera accepted between 43 and 69 percent and showed particularly low acceptance of gastropods. But recent research on P. aeruginosa strike mechanics offers a clue: this species adjusts its strike speed and tibial extension to prey movement, using slower tibial extension and more frequent pauses when targeting slower prey. This behavioral flexibility, reported in only three other mantid species, could allow Polyspilota to time its final downward tibial sweep precisely enough to capture slow-moving mollusks.

There may also be a nutritional incentive. Snails are far more than mucus and moisture: the common garden snail averages 78 to 83 grams of water, 11 to 17 grams of protein, and only 0.4 to 1.2 grams of fat per 100 grams, with highly digestible protein and meaningful amounts of calcium, sodium, potassium, magnesium, and iron. Terrestrial snails are generally abundant in moist environments, which corresponds well with the humid, low-altitude, coastal distribution of Polyspilota in South Africa. Yet the strategy carries risks. Snails serve as intermediate hosts for various parasitic nematodes and trematodes, and while the only parasite known to infest P. aeruginosa is a horsehair worm, Chordodes ferox, the authors caution that little is known about mantid parasites in general and that snail-eating could expose these predators to new infection routes.

Whether snail-eating is pure opportunism, a response to prey scarcity, or a genuinely overlooked component of the Polyspilota diet remains an open question. One hypothesis suggests snails should be unattractive prey because their mucus coating reflects light, and most studied mantises are less likely to track and strike at glossy targets, yet something in Polyspilota evidently overrides that aversion. What is clear is that these mantids are opportunistic feeders capable of exploiting novel locally available food resources, and that their diet composition is likely shaped strongly by local resource availability and the climatic conditions that promote gastropod abundance. For a genus as conspicuous as Polyspilota, remarkably little is known about its ecology and behavior, and the authors hope these observations will open new avenues of research into mantid feeding strategies, the trade-offs of mineral-rich but parasite-laden prey, and the broader ecobiology of South Africa’s poorly studied mantis fauna.

Subject of Research: Novel snail-eating behavior and distributions of Polyspilota praying mantids in South Africa

Article Title: Snacking on snails: Novel prey and the distributions of Polyspilota praying mantids ( Mantodea , Mantidae ) in South Africa

Article References: Greyvenstein, B., Janse van Rensburg, P., Bellingan, T., van den Berg, J., Jordaens, K., Stals, R., & Midgley, J. (2026). Snacking on snails: Novel prey and the distributions of Polyspilota praying mantids (Mantodea, Mantidae) in South Africa. Journal of Orthoptera Research, 35(2), 251-258. https://doi.org/10.3897/jor.35.187916

Image Credits: AI Generated

DOI: 10.3897/jor.35.187916

Keywords: praying mantis, Polyspilota, Mantodea, snail predation, South Africa, iNaturalist, citizen science, biogeography, feeding behavior, KwaZulu-Natal, museum collections, opportunistic predation