Skip to main content
. 2017 Oct 26;12(10):e0186762. doi: 10.1371/journal.pone.0186762

Table 5. Inferences for climatic variables for the middle Miocene mammalian assemblage from Somosaguas.

Quantitative models for climatic inference a Climatic inference g for Somosaguas
Number Climatic factor b Continent Body-size categories c model d r2r e r2p f r2r − r2p Inferences Confidence Interval Mean inference h
1 Tmin (°C) Palaeotropics C,D,E,G L 0.509 0.247 0.262 26.8
2 Africa G L 0.556 0.922 −0.366 25.1 17.5 − 32.6
3 G Q 0.579 0.942 −0.363 26.8 18.9 − 34.8
4 G E 0.567 0.938 −0.371 25.7 17.0 − 38.8
5 Tmax (°C) Palaeotropics B,C,D,F L 0.583 0.449 0.134
6 Asia C Q 0.528 0.337 0.191
7 Mta (°C) Palaeotropics B,C,D,E,F L 0.633 0.681 −0.048 8.4 5.4 − 11.4 9.2
8 C Q 0.522 0.534 −0.012 16.7 7.8 − 25.6
9 C E 0.505 0.533 −0.028 14.2 4.6 − 43.9
10 Asia C,D L 0.632 0.929 −0.297 13.7 8.2 − 19.2
11 C L 0.584 0.990 −0.406 18 8.9 − 27.1
12 C Q 0.584 0.984 −0.4 18.1 8.6 − 27.7
13 C E 0.698 0.882 −0.184 19.0 8.9 − 40.7
14 Africa A,D,G L 0.733 0.132 0.601
15 G L 0.663 0.713 −0.05 0.9 −4.7 − 6.5
16 G Q 0.666 0.721 −0.055 0.3 − 5.7 − 6.3
17 G E 0.644 0.665 −0.021 2.0 0.8 − 5.3
18 P (mm) Asia C,D L 0.516 0.613 −0.097 610.8 − 378.4 − 1600.0 219.3
19 C L 0.510 0.537 −0.027 389.6 − 1149.5 − 1928.6
20 C Q 0.529 0.585 −0.056 219.3 − 1365.1 − 1803.7
21 D (months) Asia C Q 0.508 0.850 −0.342 10.3 6.2 − 14.3 10.3

a according to the different regression models presented in S5 Table

b as in Table 3

c categories A–H as in Table 1

d L, linear regression models; Q, quadratic regression models; E, exponential regression models

e coefficient of determination of the quantitative bioclimatic models

f coefficient of determination of the regression between the observed and predicted values

g only for accurate models (mean ± 95% confidence interval); inference of the best fitting model for each climate parameter in each data subset is shown in bold

h mean of the best fitting models for each climate parameter in each data subset