Interspecific Altruism in Humpback Whales

Humpback whales (Megaptera novaeangliae) deliberately interfere with attacking killer whales (Orcinus orca) — and in the large majority of documented cases the animal being rescued is not another humpback but a member of an entirely different species that can offer nothing in return. The pattern is real, well-documented, and still not cleanly explained. Whether it counts as “altruism” depends on which definition you bring, and that definitional fight is the interesting part.

The documentation

The foundational dataset is a 2017 review in Marine Mammal Science by Robert Pitman (NOAA Southwest Fisheries Science Center) and thirteen co-authors, which compiled 115 humpback–killer-whale interactions recorded between 1951 and 2012 by at least 54 observers across multiple ocean basins. The headline numbers:

• Humpbacks initiated 57% of the interactions — they swam toward the killer whales, not away.
• 95% of the orcas involved were mammal-eating forms (MEKWs) rather than fish-eaters — humpbacks reliably distinguished the two ecotypes and only engaged when the orcas hunted warm-blooded prey.
• When humpbacks approached orcas that were attacking something, 89% of the time the prey was not a humpback — it was a seal, sea lion, porpoise, grey whale calf, minke whale, or (in one case) a sunfish. Ten distinct species in all: three cetaceans, six pinnipeds, one teleost fish.
• The interventions were costly and protracted: humpbacks traveled up to 2 km (in some accounts over 7 km), approached bellowing and trumpeting, positioned themselves between predator and prey, tail- and flipper-slashed, and sometimes stayed for hours — one gray-whale-calf event held humpbacks on scene for a minimum of 437 minutes. In at least 8 recorded cases the prey escaped.

The visceral origin story is Pitman and John Durban’s own January 2009 Antarctic encounter, later written up for Natural History magazine: a Weddell seal washed off its ice floe by wave-washing type-B orcas swam to a humpback, which rolled belly-up, cradled the 400-pound seal on its chest, arched out of the water, and — when the seal began to slip — gently nudged it back on with a five-meter, one-ton flipper. The BBC Frozen Planet crew happened to be filming. Pitman’s own account: after reviewing the footage, “we knew it was no accident.”

The authors classify the humpback approach-and-harass pattern as mobbing behavior — the same functional category as songbirds pestering a hawk — and note these 31 mobbing accounts exceed all previous cetacean mobbing reports combined, and are the first for any baleen whale.

The explanatory problem

Helping unrelated members of other species carries real costs (energy, time, some injury risk) and no obvious benefit — a seal does not reciprocate, and a sunfish “does not, in any meaningful sense, participate in a reciprocal ecosystem of mutual aid” (Athar’s phrasing). The paper walks the standard evolutionary toolbox:

• Kin selection — could explain communal defense of related humpback calves, but says nothing about rescuing seals.
• Reciprocity — needs stable social units; humpback associations are loose, and the beneficiaries are other species, so there is no return loop.
• Acoustic-trigger spillover — Pitman’s own favored, deliberately deflationary hypothesis. Mammal-eating orcas hunt silently but vocalize during and after an attack; humpbacks (shown by Curé et al. 2015 playback experiments to react strongly to MEKW calls) likely home on those attack vocalizations without knowing the prey species until they arrive. The selected rule is something like “hear a mammal-eating orca attacking, go break it up” — adaptive because it sometimes saves your own calves — and the seal-rescues are the rule firing past its trigger. Norris and Dohl (1980) called this kind of thing a “‘spillover’ of an intraspecific pattern into the domain of more distant relationships.”

The paper’s actual conclusion is unusually careful: “Interspecific altruism, even if unintentional, could not be ruled out.” That is not a claim that humpbacks are altruists. It is a claim that after 115 interactions, the pattern resists clean reduction to a simpler mechanism.

Where the debate actually sits

The spillover/rule account and the “something closer to intention” account are both live in the literature:

• For the deflationary rule: Pitman — humpbacks “may indeed be altruistic, but for selfish reasons in the end.” The acoustic-trigger evidence (93% selectivity for MEKWs, approaches from beyond visual range, hydrophone records of MEKW vocalizing before humpbacks arrived) is genuinely strong.
• Against full deflation: co-author Alisa Schulman-Janiger notes some intervening humpbacks were themselves attacked by orcas in the past — suggesting remembered experience, not just a hard-wired reflex. The Monterey Bay coordination (14 humpbacks converging from different directions, hours of sustained station-keeping over a dead calf) is awkward for a pure fire-on-trigger rule. And Pitman himself concedes inadvertent altruism “doesn’t quite rule out the possibility that something closer to intention is also in the mix.”

Nathan’s engagement with this (2026-09-09 highlight note) lands on the intention side with a theory-of-mind argument: the seal-rescue account shows humpbacks tracking that a seal has different needs than a whale — arching to keep it out of the water, nudging it back when it slips — and an animal smart enough to model another species’ needs is plausibly smart enough to know what it’s doing. Deflationary rules don’t usually come with that much situational tailoring. The counterpoint the paper would raise: situational tailoring after arrival is compatible with an indiscriminate trigger — the rule decides whether to go, not what to do once there.

Why this page exists in this wiki

This is the cleanest documented case of apparent cooperation that crosses the species boundary — the stress test for every account of cooperation in the wiki. mutual-aid (Kropotkin) and the Hamilton/Trivers machinery on cooperation-and-defection explain intra-species altruism via kin selection and reciprocity; the humpbacks are the case those mechanisms explicitly fail to cover, forcing the question of whether “altruism” names a behavior (fitness cost to actor, benefit to recipient — de Waal’s definition, which the paper quotes) or a motivation. evolutionary-game-theory frames strategy as what selection locks in; the spillover hypothesis is a claim that a locked-in strategy can leak cooperative outcomes into domains selection never priced. And the intent question — does rescuing a seal require modeling the seal — is the ethological edge of the virtue-ethics / kantian-indirect-duty cluster: whether care-like behavior without a moralizer behind it still counts, or whether, as with carefulness, the disposition is the object of interest regardless of what is metaphysically inside the practitioner.

Open questions the wiki should track: whether playback experiments can decouple the trigger from the tailoring; whether individual humpback histories (prior orca attack) predict intervention; and whether any other baleen species shows the pattern as ocean populations continue recovering from whaling-era depletion (the paper’s “shifted baseline” caveat — we may be watching a behavior still re-emerging).

Sources

• Kiran Athar (Space Daily) 2026 — Humpback whales are interrupting orca hunts to rescue completely different species — Kiran Athar’s Space Daily long-form piece (the highlighted article; Nathan’s note on the deflationary passage)
• Pitman, R. L., Deecke, V. B., Gabriele, C. M., et al. 2017 — Humpback whales interfering when mammal-eating killer whales attack other species: Mobbing behavior and interspecific altruism? — Pitman et al. 2017, Marine Mammal Science 33(1): the 115-interaction dataset and the kin/reciprocity/spillover analysis
• Robert L. Pitman and John W. Durban 2009 — Save the Seal! — Pitman & Durban’s first-person Natural History account of the 2009 seal-on-chest event