Vital Zone Size by Species — 61 Animals, With Sources
The short answer. A mature whitetail's broadside vital zone is about 9 inches tall by 12 inches wide, centred 27 inches above the ground, behind about 11 inches of chest. An elk is 15 × 19 at 41 inches, behind 17 inches of chest. A Cape buffalo is 16 × 18 at 44 inches, behind 22 inches. Full table for all 61 species below, with every shoulder-height source linked.
Here is the thing every other page answering this question leaves out: nobody publishes these numbers. We looked. There is no state wildlife agency, no zoological reference and no peer-reviewed paper that prints a per-species broadside heart-lung width or a vital-centre height above ground. Those are bowhunting constructs. What is published, widely and by credible institutions, is shoulder height — and shoulder height is the landmark everything else on this page is built from.
So this table is honest about two different kinds of number. 54 of 61 species carry a chest width and vital-centre height derived from a cited shoulder height through a ratio model we state in full below — 49 distinct sources, every one linked in the table. 7 species are labelled judged: they are ours, with nobody's name on them. And the vital-zone height × width column is entirely ours across all 61 rows. We say so in the column header rather than letting a table format imply a citation that does not exist.
North American species — 31 rows
Shoulder heights come from Animal Diversity Web (University of Michigan Museum of Zoology), the National Park Service, and state wildlife agencies in Washington, Colorado, Michigan and Massachusetts, plus the University of Arizona and Oregon State. Where a source gives a range, we take it by the one rule described under Which shoulder height.
| Species | Shoulder height (in) | Body plan | Chest through-width (in) | Vital centre above ground (in) | Vital zone H × W (in) | Basis | Source of shoulder height |
|---|---|---|---|---|---|---|---|
| Moose (Alaska-Yukon) | 78 | Cursorial ungulate | 23 | 55 | 18 × 23 | derived | NPS |
| Moose (Canadian) | 72 | Cursorial ungulate | 21 | 50 | 17 × 21 | derived | MassWildlife |
| Moose (Shiras) | 72 | Cursorial ungulate | 21 | 50 | 16 × 20 | derived | CPW |
| Bison | 69 | Heavy bovid | 24 | 48 | 16 × 20 | derived | ADW |
| Elk | 59 | Cursorial ungulate | 17 | 41 | 15 × 19 | derived | ADW |
| Polar bear | 59 | — | 17 | 38.9 | 13 × 15 | judged | Sparrow judgement |
| Musk ox | 53 | Heavy bovid | 19 | 37 | 14 × 18 | derived | Ultimate Ungulate |
| Caribou (Quebec-Labrador) | 47 | Cursorial ungulate | 14 | 33 | 12 × 16 | derived | ADW |
| Caribou (barren-ground) | 47 | Cursorial ungulate | 14 | 33 | 12 × 16 | derived | ADW |
| Caribou (central Canada) | 47 | Cursorial ungulate | 14 | 33 | 12 × 16 | derived | ADW |
| Caribou (mountain) | 47 | Cursorial ungulate | 14 | 33 | 12 × 16 | derived | ADW |
| Caribou (woodland) | 47 | Cursorial ungulate | 14 | 33 | 12 × 16 | derived | ADW |
| Whitetail deer | 39 | Cursorial ungulate | 11 | 27 | 9 × 12 | derived | ADW |
| Brown / Grizzly bear | 37.5 | — | 16 | 26 | 12 × 14 | judged | Sparrow judgement |
| Mule deer | 37 | Cursorial ungulate | 11 | 26 | 10 × 13 | derived | ADW |
| Bighorn sheep | 36 | Cursorial ungulate | 10 | 25 | 10 × 13 | derived | San Diego Zoo |
| Columbia blacktail | 36 | Cursorial ungulate | 10 | 25 | 9 × 12 | derived | Oregon Sea Grant |
| Dall's sheep | 36 | Cursorial ungulate | 10 | 25 | 9 × 12 | derived | NPS |
| Desert sheep | 36 | Cursorial ungulate | 10 | 25 | 10 × 13 | derived | San Diego Zoo |
| Stone's sheep | 36 | Cursorial ungulate | 10 | 25 | 9 × 12 | derived | NPS |
| Mountain goat | 35 | Cursorial ungulate | 10 | 25 | 9 × 12 | derived | Ultimate Ungulate |
| Pronghorn antelope | 34 | Cursorial ungulate | 10 | 24 | 8 × 10 | derived | ADW |
| Wolf | 34 | Canid | 8 | 24 | 8 × 10 | derived | Michigan DNR |
| Black bear | 31.9 | — | 13 | 20 | 9 × 11 | judged | Sparrow judgement |
| Coues whitetail | 30 | Cursorial ungulate | 9 | 21 | 7 × 9 | derived | U. of Arizona |
| Cougar | 30 | Big cat | 10 | 21 | 8 × 10 | derived | WDFW |
| Coyote | 25 | Canid | 6 | 18 | 6 × 8 | derived | WDFW |
| Bobcat | 23 | Big cat | 7 | 16 | 5 × 6 | derived | ADW |
| Javelina | 20 | Suid (pig) | 8 | 13 | 6 × 8 | derived | ADW |
| Sitka blacktail | — | — | 9 | 24 | 8 × 10 | judged | Sparrow judgement |
| Wild turkey | — | — | 5 | 12 | 4 × 4 | judged | Sparrow judgement |
African species — 30 rows
African shoulder heights lean on the San Diego Zoo Wildlife Alliance Library and Ultimate Ungulate (B. Huffman, BSc Zoology, Lead Keeper of Mammals at the Toronto Zoo and a contributor to Handbook of the Mammals of the World and Bovids of the World), with Animal Diversity Web and Encyclopaedia Britannica filling gaps.
| Species | Shoulder height (in) | Body plan | Chest through-width (in) | Vital centre above ground (in) | Vital zone H × W (in) | Basis | Source of shoulder height |
|---|---|---|---|---|---|---|---|
| Elephant | 130 | Columnar pachyderm | 59 | 93 | 18 × 21 | derived | San Diego Zoo |
| Giraffe | 130 | Long-legged | 30 | 99 | 18 × 23 | derived | San Diego Zoo |
| Rhino (white) | 73 | Short-legged heavy | 38 | 46 | 16 × 20 | derived | San Diego Zoo |
| Eland (common) | 71 | Cursorial ungulate | 21 | 50 | 17 × 21 | derived | Ultimate Ungulate |
| Hippo (land) | 65 | Short-legged heavy | 34 | 41 | 16 × 20 | derived | San Diego Zoo |
| Cape buffalo | 63 | Heavy bovid | 22 | 44 | 16 × 18 | derived | ADW |
| Roan antelope | 63 | Cursorial ungulate | 18 | 44 | 13 × 16 | derived | Ultimate Ungulate |
| Kudu (greater) | 59 | Cursorial ungulate | 17 | 41 | 13 × 16 | derived | ADW |
| Hartebeest (red) | 57 | Cursorial ungulate | 17 | 40 | 11 × 14 | derived | Ultimate Ungulate |
| Wildebeest (blue) | 57 | Cursorial ungulate | 17 | 40 | 13 × 16 | derived | Ultimate Ungulate |
| Zebra (Burchell) | 57 | Cursorial ungulate | 17 | 40 | 14 × 18 | derived | ADW |
| Zebra (Hartmann) | 57 | Cursorial ungulate | 17 | 40 | 13 × 16 | derived | ADW |
| Sable antelope | 56 | Cursorial ungulate | 16 | 39 | 12 × 15 | derived | Ultimate Ungulate |
| Tsessebe | 51 | Cursorial ungulate | 15 | 36 | 10 × 13 | derived | Ultimate Ungulate |
| Gemsbok / Oryx | 49 | Cursorial ungulate | 14 | 34 | 12 × 15 | derived | ADW |
| Lion | 48 | Big cat | 15 | 34 | 10 × 12 | derived | San Diego Zoo |
| Nyala | 47 | Cursorial ungulate | 14 | 33 | 9 × 12 | derived | Ultimate Ungulate |
| Wildebeest (black) | 47 | Cursorial ungulate | 14 | 33 | 12 × 15 | derived | Ultimate Ungulate |
| Lechwe | 43 | Cursorial ungulate | 12 | 30 | 9 × 12 | derived | Ultimate Ungulate |
| Impala | 37 | Cursorial ungulate | 11 | 26 | 8 × 10 | derived | Ultimate Ungulate |
| Reedbuck | 36 | Cursorial ungulate | 10 | 25 | 8 × 10 | derived | Ultimate Ungulate |
| Bushbuck | 33 | Cursorial ungulate | 10 | 23 | 7 × 9 | derived | ADW |
| Springbok | 31 | Cursorial ungulate | 9 | 22 | 6 × 8 | derived | ADW |
| Leopard | 28 | Big cat | 9 | 20 | 8 × 9 | derived | Britannica |
| Warthog | 28 | Suid (pig) | 11 | 18 | 8 × 10 | derived | San Diego Zoo |
| Duiker (blue/common) | 24 | Cursorial ungulate | 7 | 17 | 4 × 5 | derived | Ultimate Ungulate |
| Klipspringer | 24 | Cursorial ungulate | 7 | 17 | 4 × 5 | derived | Ultimate Ungulate |
| Steenbok | 24 | Cursorial ungulate | 7 | 17 | 4 × 5 | derived | Ultimate Ungulate |
| Blesbok | — | Cursorial ungulate | 11 | 27 | 8 × 10 | judged | Sparrow judgement |
| Crocodile (Nile) | — | — | 26 | 8 | 4 × 5 | judged | San Diego Zoo |
The ratio model, stated in full
Two coefficients per body plan, both with a plain anatomical meaning, and one shared constant. That is the entire model. There are no per-species fudge factors, and if there ever are, this stops being a model and starts being a pile of tuned constants.
- b — brisket fraction. Height of the chest floor above ground, divided by shoulder height. Plain English: how much of the animal is leg.
- r — width-to-depth. Side-to-side rib-cage width divided by rib-cage depth. Plain English: how round the barrel is.
- Vital centroid fraction = 0.40, shared by every plan. The heart-lung mass sits low in the chest cavity; its centroid sits about 40% of the chest depth above the brisket. Standard bowhunting placement puts the heart roughly a third of the way up from the brisket and the lung centre higher; 0.40 sits between them.
From those three numbers:
chest depth = (1 − b) × shoulder height chest through-width = r × chest depth vital centre height = [ b + 0.40 × (1 − b) ] × shoulder height
| Body plan | Species using it | b — brisket fraction | r — width ÷ depth | Why these two numbers |
|---|---|---|---|---|
| Cursorial ungulate | 40 | 0.5 | 0.58 | The classic running-ungulate conformation: leg length ≈ chest depth (the brisket sits at about half the withers height), and a deep, laterally compressed rib cage a little over half as wide as it is deep. Covers every antelope on this list, both zebras, every North American deer and moose, caribou, pronghorn, and the North American mountain sheep and mountain goat. r = 0.58 is what v194's independently-set whitetail, mule deer, elk and caribou figures imply when divided by those species' published shoulder heights. The two members whose real conformation this fits least well are named in situ: pronghorn (proportionally longer-legged than the plan assumes) and mountain goat (stockier). |
| Canid | 2 | 0.5 | 0.48 | Wolf and coyote. Leg length ≈ chest depth, the same proportion as a deer — a canid is a cursorial animal and the brisket sits at about half the withers height — on the narrowest barrel on this board. r = 0.48 comes from this module's own prior reasoning: the v295 note that used to sit on these two rows said the felid coefficient (r = 0.64) would read a canid "about a third too wide through the chest", and 0.64 ÷ 1.33 is 0.48. That is Sparrow's judgement made explicit rather than a measurement, and it is the least-supported number in this table — it is stated here so it can be argued with instead of hiding inside a row. It affects chest through-width only; the vital centre and the drawn ellipse depend solely on brisketFraction. AND THEN IT CHECKED OUT, which is the part that turns it from a guess into a coefficient: r = 0.48 reproduces BOTH v194 hand-set chest figures EXACTLY — wolf 0.48 × (0.5 × 34") = 8.16" against v194's 8", coyote 0.48 × (0.5 × 25") = 6.0" against v194's 6". Two independent field estimates set years before this plan existed, both hit on the nose. That is the same evidence that justified r = 0.70 against v194's bison figure. THE OTHER CHECK: b = 0.5 puts the vital centre at 0.700 × shoulder height, and Sal set both canids to that height by eye against the calibrated photograph on 2026-08-07, before being shown what the model produced. The two agreed. That is the same kind of independent agreement that justified r = 0.58 and r = 0.70 above. |
| Heavy bovid | 3 | 0.5 | 0.7 | Wild cattle (Bovini): the same leg proportion as a deer, on a markedly rounder barrel — a large rumen and a heavy brisket push the rib cage wide. Applied to Cape buffalo, bison and musk ox. r = 0.70 is what v194's shipped bison figure (24" through-chest) implies against a published bison shoulder height — and v295 closed that loop: ADW's published MALE bison shoulder height puts the derived through-chest at exactly the 24" v194 set by hand. Musk ox is not a bovine (Ovibos is Caprinae) and sits here on conformation rather than taxonomy; it is flagged in situ as the one non-Bovini member of this plan. |
| Big cat | 4 | 0.5 | 0.64 | Felids: deer-like leg proportion, chest rounder than a cervid but far narrower than a bear. r = 0.64 is what v194's shipped cougar figure implies against a published cougar shoulder height; lion and leopard are the same body plan at different scales. v295 extends it DOWN in scale to the bobcat, which is the largest extrapolation this plan carries and is flagged in situ. It is deliberately NOT used for wolf or coyote: a canid rib cage is markedly narrower for its depth than a cat's, and we have nothing calibrated to say how much. |
| Suid (pig) | 2 | 0.4 | 0.65 | Pigs and peccaries: short-legged and low-slung (brisket at ~40% of withers height) with a barrel chest. Calibrated against v194's shipped javelina figures, which this plan reproduces to within an inch on both axes. Applied to warthog and to javelina (collared peccary, Pecari tajacu — not a wild boar, which is a different animal at a different size). |
| Short-legged heavy | 2 | 0.38 | 0.85 | Hippo and white rhino: very low belly clearance on a massive, near-cylindrical barrel. This is the plan with the least calibration behind it — there is no North American animal shaped like this to check it against — so both species built on it are listed among our least-confident rows. |
| Long-legged | 1 | 0.6 | 0.58 | Giraffe, and giraffe alone. Proportionally the longest legs of any ungulate — the brisket sits at roughly 60% of withers height rather than 50% — over a rib cage that is deep but no wider, relatively, than an antelope's. This plan is the direct fix for the 38" vital centre a giraffe used to inherit, which sat well BELOW its own brisket. |
| Columnar pachyderm | 1 | 0.52 | 0.95 | African elephant: columnar limbs holding the belly line at just over half the shoulder height, over a chest very nearly as wide as it is deep. r = 0.95 is the coefficient we are least able to defend from anything published and it is flagged as such — but the resulting ~59" through-chest matches the commonly quoted 1.5-1.8 m body width of a mature bull, and it is not in the same universe as the 14" this animal used to inherit from a leopard. |
Why you should believe any of this
The model reproduces figures that were set independently, by field judgement, two years before the model existed. In 2024 we hand-set chest widths and vital-centre heights for the common North American species with no ratio model in sight. In 2026 we sourced published shoulder heights for 23 of the 31 North American species and could check thirteen of those hand-set pairs against the model. Thirteen species, twenty-six numbers, worst error 3 inches, and nine of the thirteen chest widths unchanged to the inch.
| Species | Published shoulder height (in) | Model: chest / vital centre (in) | Field judgement, 2024 (in) | Shoulder-height source |
|---|---|---|---|---|
| Whitetail deer | 39 | 11 / 27 | 11 / 26 | Animal Diversity Web |
| Columbia blacktail | 36 | 10 / 25 | 10 / 26 | Oregon Sea Grant |
| Mule deer | 37 | 11 / 26 | 12 / 28 | ADW, male midpoint |
| Coues whitetail | 30 | 9 / 21 | 8 / 22 | University of Arizona |
| Pronghorn antelope | 34 | 10 / 24 | 9 / 22 | ADW, male mean |
| Mountain goat | 35 | 10 / 25 | 11 / 26 | Ultimate Ungulate |
| Musk ox | 53 | 19 / 37 | 20 / 36 | Ultimate Ungulate |
| Elk | 59 | 17 / 41 | 17 / 40 | Animal Diversity Web |
| Caribou | 47 | 14 / 33 | 14 / 34 | Animal Diversity Web |
| Moose (Shiras) | 72 | 21 / 50 | 20 / 50 | Colorado Parks and Wildlife |
| Moose (Alaska-Yukon) | 78 | 23 / 55 | 24 / 58 | National Park Service (Denali) |
| Cougar | 30 | 10 / 21 | 9 / 20 | Washington DFW |
| Javelina | 20 | 8 / 13 | 8 / 14 | Animal Diversity Web |
That is not proof the model is right and we do not claim it is. It is evidence that a stated ratio model and an experienced eye agree, which is the strongest check available for a number nobody publishes. Two caveats we will state before somebody else does:
The check nearly became circular. The test asserting this agreement originally compared the model against the live table — a real check only while those rows still held the hand-set numbers. Converting them to derived values would have turned it into |x − x| ≤ 2, a test that passes forever and asserts nothing, while this very paragraph kept citing it as the evidence. The 2024 figures are now frozen as literals in the test file and must never be updated to match the model.
Bison is a real outlier and stays one. The model reads 48 inches for the vital centre where field judgement set 44. A bison's published shoulder height is taken over the shoulder hump, so it overstates the withers and the model reads high. The same source's male figure reproduces the 24-inch through-chest exactly, which tells us the coefficient is right and the input is what is tall. Noted rather than tuned away.
Which shoulder height, when a source gives a range
Published ranges usually span females to males. A bowhunter is pointing at a mature male, so we want the mature-male figure, and we take it the same way every time:
- A published male mean → use it. (Giraffe, Hartmann's zebra, pronghorn, cougar.)
- A published male range → use its midpoint. (Elephant, warthog, bushbuck, reedbuck, mule deer, bison, bighorn sheep.)
- A single published adult range spanning both sexes → use its top, because that end of the range is the male. (Everything else.)
You can tell a rule is being applied rather than shopped when it costs you. Rule 2 pulls mule deer down from the 12″/28″ we had set by field judgement to 11″/26″, when taking the top of the male range would have reproduced our own figure exactly. It pulls bighorn sheep from 12″/28″ to 10″/25″, when the other available source's both-sex range would have given 11″/27″. We kept the rule.
This biases the geometry slightly large, and the effect on a shot solver is mixed, not uniformly conservative: a larger chest width increases modelled exit drop, which is conservative, while a higher vital centre shallows the down-angle from a treestand, which is not.
The six species we will not derive
Shoulder height — height at the withers — is not a universal landmark, and this page does not pretend otherwise.
- Wild turkey. A bird does not have withers. Vital zone treated as small and roughly round, 4 × 4 inches, rather than elongated.
- Nile crocodile. A crocodile lies belly-down and has no withers. It is also the one species here whose posture is flagged as something other than standing, because promising a “relaxed, standing animal” would describe a pose it never adopts.
- Sitka blacktail. No subspecies-specific shoulder height published. We will not borrow Columbia blacktail's and call it sourced.
- Blesbok. Demoted on the day this page was published. Its 39″ shoulder height rested on a single Ultimate Ungulate account for Damaliscus pygargus that now returns 404, and Animal Diversity Web's account for the same species publishes head-and-body length but no shoulder height. There is nothing to derive from, so the row is ours until a published figure turns up.
- Black bear, brown/grizzly bear, polar bear. Published shoulder heights exist and we list them. But a bear's shoulder height is taken to the top of a standing quadruped's shoulder mass rather than to true withers, and no body plan on this page describes an ursid chest honestly. Rather than assign one and inherit its coefficients, these three keep hand-set geometry and carry nobody's name but ours.
Every one of the 61 species has an explicit row. There is no class fallback any more, and the lookup function throws rather than returning a default — so “silently inherits a class average” is not something this data can express. That matters because it used to happen: all 30 African species once took a fallback, which put a 14-inch chest and 22-inch vital centre on a six-tonne elephant. That is leopard geometry. The overlay was not slightly off; it was drawing a different animal.
How to use these numbers
Chest through-width is how far a broadside arrow travels from near-side entry to far-side exit. It sets how far the arrow drops inside the animal on a steep downhill shot — drop is roughly width × tan θ. On a Cape buffalo at 22 inches through, a 30° downhill costs about 12.7 inches of drop across the body. On a whitetail at 11 inches, the same angle costs 6.4 inches. This is why the same aiming error is forgiving on deer and lethal on buffalo.
Vital centre height sets the down-angle from a treestand: θ = atan((stand height − vital centre) ÷ horizontal distance). Aiming at the ground instead of the vitals overstates the angle. From an 18-foot stand at 20 yards on a whitetail, the true angle is about 24°, not 27°.
Vital zone height and width set the target ellipse. The zone is wider than it is tall because the lungs extend rearward behind the shoulder, which is the mechanical reason a horizontal miss is more forgiving than a vertical one, and the reason ranging error — which is purely vertical — is the error that costs you animals.
What is sourced and what is ours
| Column | Status | Who owns the number |
|---|---|---|
| Shoulder height | published | The cited institution. Verbatim figure, converted to inches. |
| Chest through-width | derived | Input theirs, model ours. 55 of 61 rows. |
| Vital centre height | derived | Input theirs, model ours. 55 of 61 rows. |
| Vital zone height × width | judged | Sparrow, all 61 rows. No published source exists for any species. The width is currently the height times a class ratio, which means it is not an independent estimate of anything — it is the height wearing a multiplier. We are saying that out loud because the previous wording implied a source it never had. |
The honest gap: rib-cage width is a far better basis for a broadside vital width than vital height is, and we now have a per-species, cited rib-cage width in the chest column. Rebuilding the vital-zone width off that is the next change to this data. We have not made it yet because replacing one unsourced ratio with another unsourced ratio is motion rather than progress, and because it moves every hit-probability number in the app, which is a separate change with a separate blast radius.
Sources
49 distinct source URLs across 61 species, linked per row in the tables above. Every one of them was fetched immediately before this page was published, and every one returns HTTP 200. That check is the reason the blesbok row says “Sparrow judgement” rather than carrying a citation: its source went dead, we found out because we looked, and we changed the row rather than the link. The institutions:
- Animal Diversity Web — University of Michigan Museum of Zoology (19 species)
- Ultimate Ungulate — B. Huffman, BSc Zoology; Lead Keeper of Mammals, Toronto Zoo (16 species)
- San Diego Zoo Wildlife Alliance Library (8 species)
- U.S. National Park Service (3 species)
- Washington Department of Fish and Wildlife (2 species)
- Colorado Parks and Wildlife, Michigan Department of Natural Resources, Massachusetts Division of Fisheries and Wildlife, Oregon Sea Grant (Oregon State University), University of Arizona College of Agriculture & Life Sciences, Encyclopaedia Britannica (1 species each)
- Viljoen, Webb, Myburgh, Truter & Myburgh 2023, Frontiers in Animal Science — crocodile girth-to-length ratio
Frequently asked questions
How big is a deer's vital zone?
A mature whitetail buck's broadside heart-lung zone is about 9 inches tall and 12 inches wide, centred roughly 27 inches above the ground on a standing animal. The chest measures about 11 inches side to side, so an arrow that passes broadside travels roughly 11 inches of tissue between entry and exit. Those figures come from a 39-inch published shoulder height (Animal Diversity Web) run through the ratio model on this page. The 9 x 12 vital zone itself is Sparrow's field judgement, not a published measurement — nobody publishes one.
How big is an elk's vital zone?
About 15 inches tall and 19 inches wide, with the centre roughly 41 inches above the ground. Chest through-width is about 17 inches. Input shoulder height 59 inches, from Animal Diversity Web.
What is the vital zone size of a Cape buffalo?
Roughly 16 inches tall and 18 inches wide, centred about 44 inches up. The through-chest is about 22 inches — double a whitetail's — which is why buffalo is a penetration problem before it is an energy problem. Shoulder height 63 inches, from Animal Diversity Web's figure of 140-160 cm for Syncerus caffer caffer.
How high off the ground is a deer's heart?
About 27 inches on a standing, relaxed whitetail, using the heart-lung centroid rather than the heart alone. This number matters most from a treestand: the down-angle to the animal is atan((stand height − vital centre height) / horizontal distance), so an 18-foot stand at 20 yards gives roughly 24 degrees, not the 27 degrees you get if you aim at the ground.
Does anybody actually publish vital zone dimensions?
No. We could not find a published per-species broadside heart-lung width or a published vital-centre height for any species, from any wildlife agency, zoological reference or peer-reviewed source. Shoulder height is published widely; vital-zone geometry is a bowhunting construct that nobody measures. That is why every number on this page carries a label saying whether it was derived from a cited figure or is our judgement.
Why is bison's number different from the model?
A bison's published shoulder height is measured over the shoulder hump, which overstates the withers, so the model reads about 4 inches high on vital-centre height (48 inches modelled against 44 inches judged). The same source's male figure reproduces the 24-inch through-chest exactly, so the coefficient is right and the input is what is tall. We flagged it rather than tuning the model to hide it.
Which shoulder height do you use when a source gives a range?
One rule, applied to all 61 species with no exceptions. A published male mean is used as-is. A published male range is used at its midpoint. A single adult range spanning both sexes is used at its top, because that end of the range is the male. This costs us agreement on mule deer and bighorn sheep, where the other choice would have matched our field judgement exactly. We kept the rule.
How many species does this cover?
61 — 31 North American and 30 African. 54 carry a chest width and vital-centre height derived from a cited shoulder height. 7 are labelled as Sparrow's judgement: wild turkey and Nile crocodile (neither has withers), Sitka blacktail (no subspecies figure published), blesbok (our source page no longer resolves and no substitute publishes a shoulder height for Damaliscus pygargus), and black, brown/grizzly and polar bear (published shoulder heights exist, but no body plan on this page describes an ursid chest honestly).
Do you check that your own sources still work?
Yes, mechanically, and it changes the table when they fail. Every source URL behind this page was fetched before publication. 49 of the 50 returned HTTP 200. One did not: the Ultimate Ungulate page for Damaliscus pygargus, which was the sole basis for blesbok's 39-inch shoulder height, now returns 404. Animal Diversity Web's account for the same species publishes head-and-body length but no shoulder height, so there is no substitute. Rather than leave a citation pointing at a dead page — or quietly keep the number and drop the link — we demoted blesbok to judged. A dead citation is not a weaker citation. It is no citation.
Sparrow Expeditions publishes the arithmetic behind its tools rather than the conclusions alone. If a number here is wrong, it is wrong in a way you can check — the input is cited, the model is stated, and the ones that are ours say so.