Dear Modelers
Warm greetings
I am trying to evaluate the habitat suitability for tigers in one of the national parks of Nepal. I used Prey species presences point, NDVI, Environmental variables Elevation, Slope and Aspect, which I think is enough to cover tiger ecology. Before running the final tiger model, I first developed a prey species model using the prey species presence points with all the environmental variables listed above. Then later, the Prey species ASCII files along with Environmental variables were used for the final tiger model. I am afraid using environmental variables more than one time will create an overfitting problem in the model. Next, I tried to develop a habitat suitability map for individual prey species, but due to few presence points, this was not possible. Therefore, I would like to know whether the approach I have considered for developing the model is correct or not. If you were a part of this project, will you please elaborate on how would you run the model?
Have a nice day.
Sincerely
Kamal Ghimire
Using the prey species maps, first, (1) classify the habitat for these species (prey).
Then enter the variables that are more priority (2) for the predator(tiger) in the final model and run Maxent based on these data (1&2).
Dear Abbas, Thank you for your suggestion for developing a final tiger model. I still have a doubt about classifying the habitat for the prey species. To classify the habitat for prey species, you mean I have to run a model with all environmental variables mentioned above (those variables that are not correlated using Pearson's correlation or any other methods) and the prey species presence points. And then the output files (consider ASCII files of prey species as an outcome of Maxent) along with environmental variables act as entry variables to the final tiger model. While doing this, we are using environmental variables more than one time in our model which I am afraid is going to reduce the overall accuracy of the model. If you can provide me with any reference to your answer then I will be glad to go through it and clarify my doubts.
Thank you.
Dear Kamal;
That's true, to first perform a modeling for prey species using the above variables that are not correlated (correlation less than 80%). Then classify the final map obtained for the prey species in the ArcGIS. Enter this map as an environmental variable with other important data for the tiger into Maxent software.
-The second method, Map of prey species;
Using the presence points of the prey species at the area, do the interpolation for the species and classify the obtained map:(1) most presence - (2) low and (3) very low, (Classify by purpose and species distribution points in the area). Use these maps with all the environmental variables listed above (those variables that are not correlated - less than 80%) to model the tiger.
Dear Abbas:
Thank you for this explanation. I must say that it is a great help for me for the successful completion of my project. Despite having all variables ready I was stuck on the methods for generating the final tiger model. I appreciate your help.
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps