Ecological indices computed from filtered observation data.
How to read this tab
Species richness (S) is just a count of species. On its own it can mislead, so this tab
checks it against two properties: (i) Effort-corrected tells you if the richness
is real or just the result of heavy sampling; (ii) Completeness tells you how much
of the true total is still unseen. The rare species, seen only once or twice, are the evidence
behind the completeness estimate.
Species inventory
97% found
range 91–100% (95% CI)
188 species recorded · ~195 estimated across the selected reefs
≈ 7 species likely still undetected
OBSERVED
the raw counts, before any correction
Species Richness (S)
188
Total Individuals (N)
391,969
EFFORT-CORRECTED
is the richness real, or just a product of heavy sampling?
Margalef (SMg)
14.52
Menhinick (SMn)
0.30
ESTIMATED RICHNESS & RARE TAIL
how much of the true total is still unseen?
Chao Estimated (SChao)
≈195
95% CI 188–206
Rare species (S1 / S2)
9 / 6
singletons / doubletons
EFFORT-CORRECTED
is the richness real, or just a product of heavy sampling?
Effort-corrected richness
Margalef by Reef
(S − 1) / ln N
Effort-corrected richness
Margalef over Time
by year
Effort sensitivity
Margalef vs Menhinick by Reef
colour = N (effort)
COMPLETENESS
how much of the true total is still unseen, and where to survey next?
Inventory completeness
Rarefaction Curve
by individual
Inventory completeness
Species Accumulation
Chao2 · by transect
Abundance structure
Rank–Abundance (Whittaker)
log abundance vs rank
Sample completeness
Observed vs Estimated Richness by Reef
S → Chao · 95% CI · % found
Where to survey next
Sampling Priority
effort vs completeness · size = unseen
How to read this tab
Richness counts species. Diversity also asks how evenly the individuals are spread
among them: two reefs with the same species count differ if one is dominated by a
single fish. This tab shows that three ways. (i) The cards give one
number each, most as an effective number of species: how many equally
abundant species would give the same value. (ii) The Hill numbers
recount the same community at a series of weightings, set by the order
α. (iii) The charts show where those numbers
come from, which sites differ despite uneven survey effort, and how the ranking shifts
with α.
CLASSICAL INDICES
how much diversity is in the pooled sample? The Hill row below restates two of these on one common scale.
Shannon (H′)
2.71
exp(H′) = 15.1 = H₁
Simpson (DSimpson)
5.50
effective species · same value as H₂
Hurlbert PIE (DHurlbert)
0.818
≈ 1 − D at this sample size
DIVERSITY NUMBERS (HILL)
the same community counted at four weightings, on one scale. The weighting is the order α, a dial for how much rare species count: at α = 0 every species counts once however rare, so H₀ is just the species total; at α = 1 each counts in proportion to its abundance; by α = 2 the common species dominate and the rare tail barely registers; at α = ∞ only the single commonest matters. Hill numbers unify the classical indices: H₁ and H₂ are the two cards above, re-expressed as effective species.
Richness (H0)
188
= S, the species count
Shannon (H1)
15.1
the Shannon card, rescaled
Simpson (H2)
5.5
identical to the Simpson card
Berger-Parker (H∞)
2.5
p₁ = 0.397
WHAT THE INDICES ARE READING
the one distribution every number above compresses
Ranked abundance
Where each index looks
cumulative share by rank
BY SITE
where is an encounter most likely to be with a different species?
By site · 23 dive sites
Chance of an interspecific encounter, by site
Hurlbert PIE · truncated axis
DIVERSITY PROFILE
how the species count changes as rare species stop counting, pooled and then reef by reef
Hill numbers
Diversity profile across orders of α
Hα · α = 0 to 4
By island · 8 locations
Diversity profiles by island
crossing curves cannot be ranked
How to read this tab
Evenness asks whether the individuals are spread among the species or concentrated in a
few. Perfect evenness never happens in the wild, so this tab reads it
two ways: (i) the cards give six measures, all but the RAD slope scored
0 to 1 where 1 is perfectly even; (ii) the charts ask
which islands are most even, and whether that ranking survives cutting every island to the
same number of fish. The measures differ wildly in value because each divides by the
species count differently, but they agree almost exactly on the ranking.
BOUNDED INDICES (0 TO 1)
how far the community sits from perfectly even. These are not on a comparable scale: each divides by the species count differently, so at S in the hundreds they read very differently for the same community
Shannon (J′)
0.518
Simpson (ESimpson)
0.029
Camargo (ECamargo)
0.088
Smith & Wilson (Evar)
0.087
SLOPE AND DIVERSITY-NUMBER RATIOS
evenness read off the rank-abundance line, and as a share of the species count. H2/H0 is not shown separately because it is identical to the Simpson card above
RAD slope (ENHC)
-9.25
0 would be perfectly even
H1 / H0
0.080
exp(H′) as a share of S
THE SLOPE BEING MEASURED
the line whose steepness is the RAD evenness
Rank abundance
The rank-abundance line and its slope
β = -9.25
How to read this tab
Dominance asks how much of the community a few common species account for. It is roughly
the inverse of evenness, but it looks only at the crowded end, so this tab reads it two
ways: (i) the cards give three measures, the commonest species as a raw
count, then as a share, then the top two together; (ii) the charts name
which species those are and show where dominance runs strongest. Unlike the evenness
indices these compare safely across reefs of very different survey effort, because a share
of the whole does not care how much of the whole you counted.
HOW CONCENTRATED IS THE COMMUNITY
the first is a raw count and moves with survey effort; the other two are shares and do not
Absolute (CAbs)
155,572
individuals of the commonest species
Berger-Parker (CRel)
0.397
1 / H∞ on the Diversity tab
McNaughton (CMcNaught)
50.2%
share held by the top two
WHO DOMINATES
the cards say how concentrated the community is; this says in whom
Top species by share
Which species make up the community
bars = share · line = running total
How to read this tab
Rarity is the opposite end of the community from dominance: it counts how many species are
scarce rather than how much the common ones hold. Because abundance stops at 1, it is
measured by counting species in low abundance classes, so this tab reads it two ways:
(i) the cards give five measures, the shape of the whole distribution,
then the singletons, then three widening definitions of scarce;
(ii) the charts show the abundance classes themselves and where each
definition draws its line. Two of the cards are blank on small samples, because below
those sizes no species can qualify.
SHAPE OF THE DISTRIBUTION
how asymmetric the community is on a log scale, and how many species were seen only once
LogSkew (RLogSkew)
0.344
negative would mean excess rare species
Singletons (S1)
9
species seen exactly once
HOW MANY SPECIES ARE SCARCE
the same calculation three times, differing only in where the line is drawn. A blank card means the sample is too small for that definition to count anything
PctRare 1% (R1%)
92.6%
needs N ≥ 100
PctRare 5% (R5%)
97.9%
needs N > 20
PctRare N/S (RN/S)
88.3%
rare = below the mean abundance
THE ABUNDANCE CLASSES
the raw counts the five cards each summarise in a different way
Species per abundance class
How many species were seen how often
S₁ and S₂ highlighted
WHERE EACH DEFINITION DRAWS THE LINE
why the three PctRare cards land so close together on this data
Three cut-offs, one distribution
Where the rarity thresholds fall
log abundance
WHICH SPECIES ARE SCARCE
named, and grouped by how widely they were found rather than ranked, because the order among the scarcest is not measurable