Academia.eduAcademia.edu

Figure 1 – uploaded by Hum Kala Rana

See full PDFdownloadDownload figure
Figure 1 Variables contribution in prediction of F. cirrhosa and L. nepalense. Moran’s I test conducted for binary presence/ absence data showed significance for both F. cirrhosa (0.155; p=0.00046 at significance level 0.01) and L. nepalense (0.161; p=0.000251 at significance level 0.01). The positive Moran’s I value for both species indicates a tendency towards clustering. The MaxEnt modelling successfully delineated the potential distribution of F. cirrhosa and L. nepalense in Nepal Himalaya using nine variables (Table 2). The relative contribution of two variables Bio8 (60.5%) and Aspects (8.5%) to the model was more for F. cirrhosa whereas, Bio8 (46.6%) and Bio3 (43.53%) for L. nepalense (Figure 1). The MaxEnt models’ Jackknife test of variable importance showed that the Bio8 for both species has the highest training gain when used in isolation (Figure 2). The area under Receiver Operating (Figure 2). The area under Receiver Operating

Figure 1 Variables contribution in prediction of F. cirrhosa and L. nepalense. Moran’s I test conducted for binary presence/ absence data showed significance for both F. cirrhosa (0.155; p=0.00046 at significance level 0.01) and L. nepalense (0.161; p=0.000251 at significance level 0.01). The positive Moran’s I value for both species indicates a tendency towards clustering. The MaxEnt modelling successfully delineated the potential distribution of F. cirrhosa and L. nepalense in Nepal Himalaya using nine variables (Table 2). The relative contribution of two variables Bio8 (60.5%) and Aspects (8.5%) to the model was more for F. cirrhosa whereas, Bio8 (46.6%) and Bio3 (43.53%) for L. nepalense (Figure 1). The MaxEnt models’ Jackknife test of variable importance showed that the Bio8 for both species has the highest training gain when used in isolation (Figure 2). The area under Receiver Operating (Figure 2). The area under Receiver Operating

Related Figures (10)

Table 1 Pearson’s correlation matrix among the environmental variables (Var.) Table 2 The subset of least correlated explanatory variables used to calibrate the model of focal species
Table 1 Pearson’s correlation matrix among the environmental variables (Var.) Table 2 The subset of least correlated explanatory variables used to calibrate the model of focal species
product of the probabilities of the sample locations (Phillips et al. 2006). The Gibbs distribution takes the form of:
product of the probabilities of the sample locations (Phillips et al. 2006). The Gibbs distribution takes the form of:
Figure 2 Average regularized gains for each variable calculated from 50 subset models produced for a) F. cirrhosa and b) L. nepalense.
Figure 2 Average regularized gains for each variable calculated from 50 subset models produced for a) F. cirrhosa and b) L. nepalense.
Table 3 Statistical approach for validating the model of two selected species Note:*10 percentile training presence logistic threshold; TSS =True Skill Statistics; AUC=Area Under Curve.
Table 3 Statistical approach for validating the model of two selected species Note:*10 percentile training presence logistic threshold; TSS =True Skill Statistics; AUC=Area Under Curve.
Figure 3 Response curves of the selected significant predictive variables showing Logistic output (probability of presence) for F. cirrhosa (a-d) and L. nepalense (e-h). Temperatures are expressed in °C and precipitation in mm; Aspects: 1(North), 2(North-East), 3(East), 4(South-East), 5(South), 6(South-West), 7(West), 8(North-West); Slope: 1(-1° - 3°), 2(3° - 8°), 3(8° - 13°), 4(13° - 18°), 5(18° - 23°), 6(23° - 28°), 7(28° - 33°), 8(33° - 38°), 9(38° - 43°), 10(43° - 48°), 11(48° - 53°). 3 Discussion
Figure 3 Response curves of the selected significant predictive variables showing Logistic output (probability of presence) for F. cirrhosa (a-d) and L. nepalense (e-h). Temperatures are expressed in °C and precipitation in mm; Aspects: 1(North), 2(North-East), 3(East), 4(South-East), 5(South), 6(South-West), 7(West), 8(North-West); Slope: 1(-1° - 3°), 2(3° - 8°), 3(8° - 13°), 4(13° - 18°), 5(18° - 23°), 6(23° - 28°), 7(28° - 33°), 8(33° - 38°), 9(38° - 43°), 10(43° - 48°), 11(48° - 53°). 3 Discussion
Figure 4 Predicted potential distribution of a) F. cirrhosa and b) L. nepalense under current bioclimatic conditions and location of occurrence used for modelling and c) combined habitat suitability of focal species. The gray area in the map represents the area below 10" percentile training presences regarded as no suitable).
Figure 4 Predicted potential distribution of a) F. cirrhosa and b) L. nepalense under current bioclimatic conditions and location of occurrence used for modelling and c) combined habitat suitability of focal species. The gray area in the map represents the area below 10" percentile training presences regarded as no suitable).
Table 4 The projected areas of F. cirrhosa and L. nepalense under current and future 2050 climate change scenaric (four trajectories) categorizing low, medium and high suitability value. (Note: area below 1oth percentile training presences regarded as no suitable) plays a useful role in conservation management due to the interrelationship between the size of species geographic range and species extinction risk (Purvis et al. 2000; Cardillo et al. 2008). For that purpose, SDM is one of the important too for determining species’ niches (Lozier et al. 2009) and the modelling algorithms such as MaxEnt perform well with a limited number of presence- only data to produce distribution ranges (Elith et al. 2011) (Ranjitkar et al. 2014a). With the use of bioclimatic variables at macro-scales, these modelling algorithms are producing successful
Table 4 The projected areas of F. cirrhosa and L. nepalense under current and future 2050 climate change scenaric (four trajectories) categorizing low, medium and high suitability value. (Note: area below 1oth percentile training presences regarded as no suitable) plays a useful role in conservation management due to the interrelationship between the size of species geographic range and species extinction risk (Purvis et al. 2000; Cardillo et al. 2008). For that purpose, SDM is one of the important too for determining species’ niches (Lozier et al. 2009) and the modelling algorithms such as MaxEnt perform well with a limited number of presence- only data to produce distribution ranges (Elith et al. 2011) (Ranjitkar et al. 2014a). With the use of bioclimatic variables at macro-scales, these modelling algorithms are producing successful
Table 5 The Retracted, Stable and Expanded areas of future (2050) under four trajectories with respect to tt current distribution of F. cirrhosa and L. nepalense
Table 5 The Retracted, Stable and Expanded areas of future (2050) under four trajectories with respect to tt current distribution of F. cirrhosa and L. nepalense
Related topics:
GeographyClimate ChangeSpecies Distribution ModelsSpecies Distribution ModellingMountain Science
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved