Skip to main content
. 2010 Apr 23;5(4):e10294. doi: 10.1371/journal.pone.0010294

Table 6. Elephant design- and model-based abundance estimates (N and N′, respectively) with 95% confidence intervals (95% CI and 95% CI′) and percent coefficient of variation (%CV and %CV′) for each survey stratum and for the landscape.

Survey stratum N [95% CI] %CV N′ [95% CI′] %CV′1
NNNP 2,175 [1,595–2,966] 15.2 2,131 [447–3,309] 50.9
Kabo FMU 1,774 [1,160–2,713] 20.4 1,606 [217–2,869] 65.7
Pokola FMU 3,130 [1,817–5,392] 26.7 3,157 [508–5,590] 74.3
Loundougou FMU 1,406 [679–2,914] 35.8 1,230 [416–4,179] 66.5
Mokabi 59 [19–185] 57.5 95 [199–3,058] 70.0
Bailly 2,495 [1,059–5,880] 43.4 2,194 [672–4,891] 49.1
LTCR 37 [13–109]2 59.82 39 [752–2,763] 43.9
Ndoki-Likouala Landscape 11,076 [8,223–14,920] 15.0 10,452 [7,813–17,126] 22.5
1

The model-based coefficient of variation %CV′ is directly comparable to the design-based %CV, as aside from the variation from encounter rate it also includes the variation from the detection function, dung production and dung decay. The model-based 95% CI′ only include variation due to variation in encounter rate.

2

See Table 3 for methods used to estimate abundance, density and confidence intervals for the whole LTCR.