Some thoughts:

1) An important starting point is to ask which platform you wish to use to conduct your spatial analyses. While GIS systems software is one of the better know tools to model habitat suitability, it can be expensive. Alternatives to consider are freeware options, such as QGIS or conducting analyses in R with packages such as sp, raster, and ggplot2 for visualizing your data. Both of these have a rich community of users and many helpful guides, Q and A sites, and so forth. Did I mention they are free?

2) Another important consideration is to figure out what spatial layers exist for your study area. For instance, if you want to make inferences about habitat use in wetlands/grassland you might search online sources to see whether spatial layers exist for your study area at the resolution and scale over which you hope to project your results. Often times this can be a limiting step. No spatial data = no projecting your results.

3) A third consideration is the type of data you have. Sounds like you have presence/absence data. Great! Better than presence only data. If you have visited sites multiple times over the study, I would think about using an occupancy modeling approach, which account for imperfect detection of your species.

4) Finally, it sounds like you want to make a predictive model, rather than one that just explains where you species is in the area that you surveyed. If you surveyed a small area (like 10 hectares) don't try and project you results over the entire country, region, continent. Go for something smaller and more realistic that may reflect the same conditions / habitat suitability experienced by the species that you surveyed.

Hope you find this to be a helpful starting point!

Best regards,

Micah

Yes, a spatially explicit predictive model sounds best. I made one in my attached paper on bird movement :)

Yes, a spatially explicit predictive model sounds best. I made one in my attached paper on bird movement :)

ArcGIS has a tutorial online for spatial analysis of habitat suitability. Also, there are programs called MaxEnt and PRESENCE that will analyze the data and create a map. As Micah said, finding GIS layers might be difficult depending on where you are. You can make a wetland map using remote sensing software like ecognition or envi, if you know what the land cover types are and train the software. Harris geospatial has tutorials online for Envi.

Thank you Micah , Androw, & Amanda for your important suggestions!!

Dear Abebayehu, there are different model for habitat quality like InVEST, MARAXAN... So you can try it.

Thank you