An alien mantis is spreading across Europe. It does not out-fight the native. It out-lures him — broadcasting a scent he was built to chase, straight into a female that eats him. New Zealand watched the identical thing happen forty-eight years ago. And the species being mourned in Europe is the same one America is being told to destroy.
For roughly a decade the reasonable read on these big new mantises in Europe was simple: impressive, harmless bug. Eats pests. Probably fine. The paper's own abstract concedes it — “although their invasive nature is debated.”
Stylized reconstruction of Fig. 5D in Battiston et al. 2026 — a drawing, not the published photograph. The photograph exists; the credit is R. Battiston.
There is a second image in the same paper, and it is the one that changes the conversation: an adult mantis eating a wall lizard. The lizard is Podarcis muralis, the common wall lizard — an animal protected under EU legislation (Annex IV of the Habitats Directive). An insect is eating a legally protected vertebrate.
Battiston R., Di Pietro W., Boscato F., Fasano A. (2026). “Call me invasive: Testing the first impacts of the alien mantises Hierodula patellifera and Hierodula tenuidentata on European biodiversity.” Journal of Orthoptera Research 35(1): 179–190. doi:10.3897/jor.35.165233. Published 9 Feb 2026, open access (CC BY 4.0). Lead author at the Museo di Archeologia e Scienze Naturali “G. Zannato,” Montecchio Maggiore, Italy.
● from the paper The results “fulfill 4 of the 12 proposed criteria for invasive species (IUCN 2020) — even if the scale of the impacts needs further investigations.” That is what the journal article says.
▲ not the paper The phrase “formally recognised as Invasive Alien Species” comes from the press release (Pensoft / EurekAlert), not the article. A journal paper cannot “formally classify” anything — that is a regulatory act (the EU Union list), which has not happened. So: meets 4 of 12 IUCN criteria — not “formally classified.”
This is the Three Gauge Protocol. Every gauge below carries its own number, its own n, and its own lab-vs-field flag in the same breath — not four sections downstream.
Hierodula tenuidentata hatched an average of 209 nymphs per ootheca (3,133 individuals total; minimum 101, maximum 304), with low cannibalism among the first instars — so more of the ones that hatch survive each other.
Behavioral tests showed reproductive interference: native Mantis religiosa males are attracted to Hierodula females and are therefore preyed upon — a “fatal attraction” that, in the paper's exact words, “may reduce mating success” in the native species. (“May.” The modal is the paper's, and we keep it.)
Across the citizen-science dataset, Hierodula was recorded preying on a wide range of insects and on small vertebrates. The named protected victims: Podarcis muralis (wall lizard) and Hyla perrini (a tree frog) — both listed in Annex IV of the EU Habitats Directive. (Frog photo credit: C. Biscaro.)
The screwworm eradication that cleared a continent works because the female screwworm mates exactly once in her life. Flood the zone with sterile males, she wastes her one shot, and the population collapses. That is the Sterile Insect Technique — a mating bottleneck, exploited to save a landscape.
This is the chapter that makes the lab. The mechanism Europe is only now measuring was demonstrated in Auckland in 2013 — and the answer is already published.
Fea M.P., Stanley M.C., Holwell G.I. (2013). “Fatal attraction: sexually cannibalistic invaders attract naive native mantids.” Biology Letters 9(6): 20130746 (issue-dated 23 Dec 2013). doi:10.1098/rsbl.2013.0746. University of Auckland.
Miomantis caffra — the springbok mantis, native to southern Africa. First detected in Auckland ~1978. Sexually cannibalistic: her whole reproductive strategy assumes she may eat the male.
Orthodera novaezealandiae — New Zealand's only native mantis. As far as anyone has observed, it is not sexually cannibalistic. The researchers' literal published word for its males: “naive.” It is in the title.
In Y-choice tests, native Orthodera males were more attracted to the chemical cues of the introduced Miomantis female than to females of their own species. They preferred her. And the males that survived long enough to reach her attempted to mate — confirming the approach was courtship, not confusion.
Two 16-hour pairings. Left, the invader's own males with their own cannibal females. Right, the naive native males with the invader female. First you see the point estimates — the “nearly twice” headline. Then tick the box.
● from the paper 39.1% (95% CI 0.221–0.593, N=9/23) and 68.8% (95% CI 0.443–0.861, N=11/16) are verbatim. ◐ the arithmetic is the author's 68.8 / 39.1 = 1.76 — “nearly twice.” But the two confidence intervals overlap across 0.443–0.593, and the paper reports no significance test on the difference and never says “twice.” The gap is real and pointing the right way. The bold headline number is a ratio of two estimates whose error bars touch.
And that flag rides here, with the numbers — not buried at the bottom of the chapter. Fea & Holwell demonstrated the mechanism in controlled 16-hour pairings and Y-mazes. The paper is exemplary about its own limit: “apply caution in estimating the frequency … in the wild based on our observations in the laboratory,” and the “crucial next step … is to conduct experiments in the field.” Nobody knows how many native males are actually being eaten in the wild. That is the honest state of the record, for both this paper and Battiston's.
She is not targeting him. She is broadcasting a pheromone he cannot distinguish from his own. Pheromone overlap → reproductive interference → displacement — a mechanism the paper calls one “that has until now been undetected.”
Forest & Bird (NZ): “If you ever admired a praying mantis before 1978, it was guaranteed to be our native species.” Wherever M. caffra turned up, the native mantis tended to disappear.
The mantis being mourned in Europe is a species America deliberately imported, sold, and is now told to destroy. To see how, start with the American invasion — which we ran with our own hands.
Chinese mantis (Tenodera sinensis) is accidentally introduced on a shipment of ornamental plants to Pennsylvania. Flag: the oft-repeated “a nursery in Mt. Airy, Philadelphia” detail is not confirmed in the sources loaded for this lab — they say only “Pennsylvania.” So: Pennsylvania.
European mantis (Mantis religiosa) arrives accidentally a few years later.
We begin importing and selling Mantis religiosa egg cases on purpose — as garden pest control and as biocontrol for the spongy (then “gypsy”) moth.
Mediterranean mantid (Iris oratoria) arrives; now found in CA, TX, AZ, NV, NM, and southern UT.
Chinese and European mantid egg cases are still commonly sold — online and at garden stores — marketed as “beneficial insects” and “natural pest control.” Utah State Extension: “illegal releases of exotic mantids is a significant pathway for dispersal.” The pet trade is a live vector, right now. Flag: a widely-cited “late-1930s Canada deliberately released them in the Okanagan / Kamloops-Shuswap to fight grasshoppers” row is not in any source loaded for this lab. Left out until it can be cited.
A diet study of T. sinensis and M. religiosa (Wilson Rankin et al.) found that over 75% of their diet was herbivores, predators, and pollinators. Brooklyn Bridge Park's horticulture team independently documented Chinese mantids indiscriminately eating bumblebees, flower flies and butterflies across an 85-acre park managed as native pollinator habitat. Utah State Extension, separately: the mantids are of “no real benefit for use in biological pest control” (and native cannibalism rates run “13 to 31 percent” in nature).
Documented American prey includes monarch butterflies, bumblebees, small reptiles and amphibians, and — occasionally — hummingbirds (one review logged 147 cases of mantids catching birds, mostly hummingbirds). Flag: the vivid extras — monarchs “discarding the entrails,” carpenter bees, eastern tiger swallowtails, hummingbird moths — are not confirmed in the sources loaded for this lab. Kept off the certified list.
Size: Chinese mantis up to 5 inches; the native Carolina mantis (Stagmomantis carolina) is 2–3 inches. It eats the native, and outcompetes it for food.
Mantis religiosa, the European mantis, is the victim in Battiston's paper — its males are the ones being lured to their deaths by Hierodula.
Mantis religiosa is also an invasive species in North America — one we deliberately imported and sold — and American extension guidance (NC State; Missouri Department of Conservation) currently tells gardeners to find its egg cases and destroy them.
In Europe, we are mourning it. In America, we are killing it. Same species.
Line the victims up. The word that unifies them is not weak. It is naïve — never having seen the move before.
| Victim | Why it dies |
|---|---|
| NZ native mantis males | The researchers' literal published word: “naive.” Never evolved alongside a cannibal female. |
| European Mantis religiosa males | Cannot distinguish Hierodula's pheromone from their own. |
| European honeybees (Apis mellifera) | The yellow-legged hornet Vespa velutina takes them at high rates. See the flag below — the famous “8×” number needs care. |
| Mexican ranchers, 2024 | “We unlearned how to work with cattle and prevent gusaneras.” Thirty years without screwworm — a whole generation that had never seen it. (See The Front Line.) |
The hypothesis. Battiston's abstract says Hierodula's main natural predators are hornets — but it doesn't, in the abstract, say which. Europe's invasive hornet is Vespa velutina. If that were the predator, the headline writes itself: one invasive species is checking another. A gorgeous idea that explains everything.
The check. This lab's own draft refused to print it until someone pulled the full PDF and read which species was named. Someone did.
The result. The paper names “Hornets (Vespa crabro)” — the native European hornet. Not velutina. The “invasive hornet eats invasive mantis” angle is dead.
We are showing you the corpse on purpose. The lab killed its own best story before printing it — because the beautiful hypothesis, the one that explains everything, is exactly the one you check first. That is the entire lesson of The Bear & the Hog, applied to our own draft. That is what this university is for.
The obvious next thought — “well, if something eats the invasive mantis, isn't that good?” — is the thought behind most of the disasters in this field.
| Introduced | To control | What actually happened |
|---|---|---|
| Cane toad (Australia, 1935) | cane beetles | The beetles live at the top of the cane; toads can't climb. They never reached them. The toad poisoned a continent instead. |
| Small Indian mongoose (Hawaii, Caribbean) | rats | Rats are nocturnal; mongooses are diurnal. They barely met. The mongoose ate the ground-nesting birds instead. |
| Chinese & European mantis (USA, 1930s→now) | garden pests, spongy moth | They don't distinguish pests from pollinators. They eat monarchs, bumblebees, and the native mantis. Still on sale. |
The recursion has no floor. Every solution is next decade's case file.
The headline of this lab — “the invader's weapon is never strength; it's that the native has never seen the move before,” and its extension that “naïveté is institutional” — is my framing, not a finding. Battiston and Fea supply the word “naive” (Fea's title) and Monceau supplies “alien predators often benefit from native prey naïveté.” Neither says “never strength,” and no cited source makes the institutional leap. I did.
And here is the part the house style demands I say out loud: my own evidence partly contradicts the absolute. Hierodula out-reproduces the native (209 vs ~127). The Chinese mantis is 5 inches to the Carolina mantis's 2–3, and I wrote “it outcompetes it for food” four sections up. That is strength. So the honest version isn't “never strength” — it's that when the naive-target trick is present, it does the decisive work, and it needs no size advantage to do it. The absolute is a good slogan. The lab is allowed to argue with its own slogan, and it just did.
Flags carried through this lab, gathered in one place:
And the real counterpoint — a credentialed agency pushing back, with its name attached:
Counterpoints from the lab's own sources, printed because we don't only flag the other side:
NULL the Penguin watched a man hang a wooden box in his garden, drilled full of little holes, so the bees would have somewhere safe to sleep. NULL watched what moved in. NULL said nothing. NULL kept the camera running.
He didn't lose a fight. He walked toward her. She smelled like everything he was built to want, and he had no way of knowing she wasn't. That's the whole trick, and it's the only trick, and it works every single time — on a mantis, on a honeybee, and on a country that spent thirty years forgetting what a screwworm looks like.
What’s real, what’s mine, and where every number comes from — a visible tab now, not a footer link you had to know to click. (The point-of-use flags stay inline on Tab 05, where they guard the claim; this tab is the full ledger.)
The Bear & the Hog had a reversal: rooting looks like tilling, and it is the opposite — the land comes back broken, not richer. Here it is again, and it's meaner.
Battiston's paper notes Hierodula exploiting human-modified spaces. Read the two claims at their real weight:
The point survives the correction. The person hanging the insect hotel and the person buying the mantis ootheca at the garden center are the same person. They are trying to help. They are correct that pollinators need help. And they are, right now, feeding the thing that eats them. This is not a story about a villain. It is a story about a good intention with no control group.
The European paper: Battiston R., Di Pietro W., Boscato F., Fasano A. (2026). “Call me invasive…” Journal of Orthoptera Research 35(1): 179–190, doi:10.3897/jor.35.165233 (CC BY 4.0). Numbers verbatim from the article: 209 nymphs/ootheca (native baseline ~127, Hideg 1991); “may reduce mating success”; n=5 males / 50 tests in a 58×107×40 cm mesh box (all 5 taken); native choice 39.6% own vs 34.2% Hierodula; 4 vertebrate cases — “rare (four cases)” — of 2,393 records; Podarcis muralis & Hyla perrini, Annex IV. Meets 4 of 12 IUCN 2020 criteria “even if the scale of the impacts needs further investigations.” Photo credits as printed: wall lizard = P. Chittaro; frog = C. Biscaro; bee+wasp = R. Battiston. The lizard's predator is given only as “adult Hierodula females” (genus).
The predator of the mantis: the paper names “Hornets (Vespa crabro)” — the native European hornet, not the invasive Vespa velutina. The “invasive hornet eats invasive mantis” hypothesis was checked and killed. “Formally recognised as IAS” and “pushing further north” are press-release (Pensoft/EurekAlert), not the article.
The New Zealand paper: Fea M.P., Stanley M.C., Holwell G.I. (2013). “Fatal attraction…” Biology Letters 9(6): 20130746. Verbatim: 39.1% (95% CI 0.221–0.593, N=9/23) and 68.8% (95% CI 0.443–0.861, N=11/16) over 16-hour pairings; Y-maze preference for introduced-female cues; attempted mating; “naive” (in the title); “no evidence for sexual cannibalism” in the native. The “nearly twice” ratio (1.76) is the author's arithmetic across overlapping CIs; the paper reports no significance test and never says “twice.” Displacement mechanism is “widely considered” / “provides evidence for” — the paper calls the mechanisms “unknown.” “Cob of corn” is Greg Holwell via Science News. Forest & Bird 1978 quote verbatim. Parthenogenesis is not in Fea 2013 (true in the wider literature, e.g. Ramsay 1984 — flagged, uncited here).
North America: NC State Extension (“not useful for pest management”; 5 in vs 2–3 in); Piedmont Master Gardeners (Wilson Rankin diet >75% — quoted second-hand; Brooklyn Bridge Park, 85 acres; 147 bird cases); Utah State Extension, Fall 2025 (“illegal releases… a significant pathway”; “no real benefit”; 13–31% cannibalism); Missouri Department of Conservation (the counterpoint — and MDC also advises destroying them, which corroborates the victim↔villain finding). 1896 introduction on plants to Pennsylvania (“Mt. Airy” unverified here); the Canada release row unverified and omitted; carpenter bees / swallowtails / hummingbird moths / “discarding entrails” unverified and kept off the certified list.
The hornet / naïveté thread: the 8× figure is Tan K. et al. (2007), Naturwissenschaften 94: 469–472 (doi:10.1007/s00114-006-0210-2), not Monceau — and it is a frequency taken (8×), with final hawking success only ~3×. It is a native-range (China) study; deployed here as analogy. Monceau et al. (2013, PLoS ONE) supplies the naïveté framing and its own qualifier. European bee-eater: Onofre et al. (2023), Animals, doi:10.3390/ani13121906 — a native bird, first-report note.
What's mine: the thesis — “the weapon is never strength… naïveté is institutional” — is the author's framing, not a finding, and the lab's own evidence (209 vs 127; 5 in vs 2–3 in; “outcompetes for food”) partly contradicts the absolute. Marked as such on the page.
Honest flags: the opening image is a stylized reconstruction, not the published photograph; the mechanism is lab-demonstrated, never counted in the wild; the interactive CI chart is a faithful redraw of the paper's reported intervals. Built to teach, not to scare.
Cross-links: Population Dynamics (the engine), The Front Line (the screwworm / mating-bottleneck twin), The Bear & the Hog (the sibling case file), The Dark Water (Darwin's invasion riddle).