International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
ile Class MAXENT GARP GLM DA
-2 4 Area (ha) % Area (ha) % Area (ha) % Area (ha) %
il, Habitat 290 1.7 1406 8.2 991 5.8 1582 9.2
in Non-Habitat 16846 98.3 15730 91.8 16145 94.2 15554 90.8
pa Sum 17136 100 17136 100 17136 100 17136 100
as Table 3 SS-1: distribution statistics of predictive maps of the CF potential habitat generated from the models
fs with elevation, slope, and terrain position variables.
E Class MAXENT GARP GLM DA
NT Area (ha) % Area (ha) % Area (ha) % Area (ha) %
M Habitat 374 22 998 5.8 1006 5.9 1527 8.9
er, Non-Habitat 16762 97.8 16138 94.2 16130 94.1 15609 91.1
les Sum 17136 100 17136 100 17136 100 17136 100
jle Table 4 SS-3: distribution statistics of predictive maps of the CF potential habitat generated from the first
St, data-merged models with elevation, slope, and terrain position variables.
as
to Class MAXENT DA
: ass
se Area (ha) 0o Area (ha) %
ly Habitat 4395 15.4 7549 26.46
to Non-Habitat 24130 84.6 20976 73.54
dy Sum 23525 gg 28525 100
ed Table 5 SS-4: distribution statistics of predictive maps of the CF potential habitat
generated from the two models with elevation, slope, and aspect variables.
ed
of Class MAXENT DA
1d Area (ha) % Area (ha) %
an Habitat 2924 10.3 7567 26.53
In Non-Habitat 25601 89.7 20958 73.47
ed Sum 28525 100 28525 100
he Table 6 SS-5: distribution statistics of predictive maps of the CF potential habitat generated
in from the second data-merged models with elevation, slope, and aspect variables.
ne
no (a) (b) (c) (d)
In
he
ly
he
ly,
%
les
Figure 2 SS-3: four models for mapping the potential habitat of cycad-ferns in the Huisun study area, (a) MAXENT; (b) GARP;
(c) GLM; (d) DA.
(a) (b)
ex
Figure 3 SS-5: two models for mapping the potential habitat of cycad-ferns in the
Huisun and Tong-Mao Mountain study area, (a) MAXENT and (b) DA.
