Re: Creating a map of species richness

[Alicia Sutton]

Hi Colin, 

Back again. I just took the Intro to using GIS in Marine Mammal Research course here in Perth with Michael. It was great- I pretty much learnt everything I needed to get the bulk of my GIS chapter done. So thank you for putting together the course and book. 

I am coming up one little hurdle though- which I can't seem to get over.

I have created a polygon grid to cover my study area using Create Fishnet, and have removed cells that fall on land (file name: POLY_GRID_NO_LAND). I have a polygon of a species distribution and I want to populate the polygon grid cells with presence values from the species polygon (file name: B_AMBLYOPS; should also note here that it is a series of small polygon circles that are not connected so they each have their own data_value in the attributes table). If the polygon falls over even part of a polygon grid cell, I would like it to count as presence. I thought I could link the two shapefiles using the Spatial Join tool, as we ran through how to do exactly this with point data layers in the course. 

In Spatial Join, I would select :

Target Feature: POLY_GRID_NO_LAND 

Join Feature: B_AMBLYOPS     

Join Operation: One to One

Keep All target features: ticked

Match Option: Contains

I have also tried a combination of other options for the above fields with no success. I have also made sure to select SUM as the merge rule, but whether or not this is necessary because they are polygons of a single value...

Am I on the right track using Spatial Join for this?   

Cheers, 

Alicia

On Friday, December 14, 2012 10:45:14 PM UTC+8, GIS in Ecology wrote:

Hi Alicia,

 

Glad to be able to help.  If you have any further questions on this, feel free to post them again on this thread.

 

All the best,

 

Colin

 

On Thursday, December 13, 2012 1:17:06 PM UTC, Alicia Sutton wrote:

Thanks Colin, this information has been very helpful! I will give it a go, and if I get stuck it would be great to be able to ask for your help again.

Many thanks, 

Alicia 

[Alicia Sutton]

Hi Colin, 

I persisted again today with my query and I think I have it worked out. So for now, please disregard my message sent yesterday.

[GIS in Ecology]

Hi Alicia,

 

Good to hear you managed to get along to the course in Perth and that you found it useful, and that you have almost managed to get your GIS chapter finished. 

 

It's also good that you seem to have got this worked out in the end. Sometimes with GIS it is a matter of taking a break and coming at it with fresh eyes, or just starting again in a new project.

 

If you have any further problems with this, just post in this thread.

[Alicia Sutton]

Hi Colin, 

So I was able to spatially join a species distribution polygon to a polygon grid of the study area and it works well. However, with this process, it seems like I may have to do this individually for all of the 69 species polygons, which means 69 new spatially joined shapefiles. Which is fine, but ideally at the end of it all, I would like to combine all of these 69 into one table with the columns: cell ID number, lat,longs, total species richness (which I could calculate using field calculator) and then columns for each of the species with an indication if they are present at that cell number or not. Is this something that is possible using ArcGIS, or is there a speedier way of being able to create one big table of all these species polygons with a polygon grid. 

I also had a go at using the program Spatial Analysis for Macroecology (SAM), to populate the grid with species and get a total richness, but for some reason SAM won't pick up on some tiny poylgons, and also misses some species altogether. So if I can do this in ArcGIS, or even add in a tool to do this, that would be great.  

Cheers, 

Alicia       

[Alicia Sutton]

Sorry Colin, think I found the solution again. I ended up using UNION to combine all the species polygons, creating an extra column in the attributes with a sum of the species, then used spatial join to combine this union poly with a poly grid.

Cheers,

Alicia

[GIS in Ecology]

Hi Alicia,

 

Sounds like you're getting the hang of working these things out (which is great),

 

The method you've worked out with the UNION tool should certainly work. However, while it will give you the total number of species in each grid cell, you will lose the ability to work out which species are present in it. This might not be important to you, but I figured I'd let you know a way of doing this (since I posted something similar on this yesterday - see https://groups.google.com/forum/#!topic/gis-in-ecology-forum/xjFjUH4iw28).

 

Presumably at the moment you have a value of 1 (for presence) for each species so that when you add the numbers together during the UNION, you get a total number of species in each polygon. However, if you code each species separately you can also generate a code which tells you which species is present in each polygon (and so grid cell). To do this, you give the first species a value of 1 to indicate its presence, the next a value of 10, the next 100, the next 1000 and so on. Then when you add the values together you generate not a total number of species but a unique binary code whcih tells you which species are present in ant given polygon/grid cell. For example, if all four of the above species were present, you'd geta summed value of 1111, of the first species was absent you'd get a code of 1110, if only the second speceis was present you'd get a value of 0010 (the first two values would be missing so it would read as 10). 

 

The only limitation to this is the fact that if you try to enter too many digits in a field, ArcGIS will then shorten the number to an exponential function (i.e. 1.45+14 for a number with 15 digits) and you will lose information. The way round this is to break your species down into smaller groups, such as individual taxa (such as family or genera) where the total number of species is small enough for the field format to handle and create a field with a binary code representing which species are present for each smaller group. In ArcGIS 9.3 the maximum number of digits you can have in an long interger field before it starts shortening them with an exponential function is 14, this means that each of your smaller groups would have to have 14 or less species for this to work. However, that would only mean having to break your 69 species down into 5 groups.

 

Anyway, this might not be of interest, but I thought I would mention it in case it was.

 

As always, any question s about this, or if you need helpm with anything else, feel free to post again.

[amar...@ucr.edu]

Hello Colin,

 

I need to make a species richness map, and I found this on the web. I followed all your instructions until the end of this email and it all worked fine. I only need to know how to transform the species richness column in the grid table to a raster map. I would appreciate if you could help me with this last step into making a species richness map.

 

Regards,

Alejandra

 

On Wednesday, December 12, 2012 12:39:08 AM UTC-8, Alicia Sutton wrote:

Hi, 

In short I would like to create a grid cell design (~300*200km cells) to span the Indian Ocean data, where I can then populate each of those cells with a number representing how many species occur within that individual cell. Would I need to somehow design an excel sheet and then import it in? Or apply the grid cell design, export it, then populate the cells with fnumbers?

I have the study area, created in IDRISI, but I've been told IDRISI doesn't have the option to apply the grid cell design. I've been trying to understand how to do this in ArcGIS but have had no luck so far.

Is anyone able to suggest what I might do?

Cheers, 

Alicia

[GIS in Ecology]

Hi Alejandra,

 

If you wish to convert the polygon grid into a raster data layer of species richness, this is just a matter of using the POLYGON TO RASTER tool (in the ArcGIS toolbox window, select CONVERSION TOOLS> TO RASTER> POLYGON TO RASTER). For the VALUE FIELD select the field in your attribute table which has the species richness value in it. You will need to make sure you use exactly the same cell size and the same extent for your raster data layer as you did for your polygon grid data layer.

 

If you have any problems with this, or if it doesn't work, just let me know.

 

All the best,

 

Colin

[P Siri]

Hi Colin,

Not sure if this forum is out of date but I need to create a map of species richness and I'm coming across some problems, so I was wondering if you could advise me. Currently, I have rasters of individual species distributions with presence values 1, and a main raster of madagascar as my baseline map with grid values of 0. 

I was thinking of using raster calculator tool to add the individual species distributions with the baseline map for all my species and then adding them on top of each other again to create a final richness map. 

The only issue I have is that when i add the individual species distribution to the baseline map, the 0's from the madagascar raster disappears and the result is just a raster that is the same as my original. Since my species distributions are very specific and patchy, I am unable to add it all at once. The cell size/coordinate system is the same. 

Do you have any advice as to why that might happen?

Any help would be appreciated.THank you.

Siri

[GIS in Ecology]

Hi Siri,

 

This sounds like a relatively simple issue, and it is related to the way that the RASTER CALCULATOR tool works. Specifically, when you add any two raster data layers together, any cell which has a 'no data' value in any data layer will be assigned a 'no data' value in the output raster data layer (and thus will appear to disappear from it).

 

The way to deal with this is that for each species raster data layer, you need to use the RECLASSIFY tool to change the 'No Data' values in your species raster data layers to zeros before you try to add them together in the RASTER CALCULATOR tool.

 

You can find the full details of this in my book 'An Introduction To Using GIS In Marine Biology: Supplementary Workbook One - Creating Maps Of Species Distribution', but it is relatively easy to work out how to do this yourself with the RECLASSIFY tool (just use a 0 for the new value in the row that has No Data for the old value) so give that  a go be buying the book.

 

I hope this helps. If it doesn't, or if I've mis-understood the problem you're having, just report back on this thread.

 

All the best,

 

Colin

[Anahi]

Hi Colin and Alicia, and group, 

I am new on this group and I am trying to do the same task than Aicia: I have several polygons (more than 100) and I need to overlap them with a grid, and count how many polygons are overlaped in each cell of this gris, but I can´t find a solution yet. Can you help me?

Thanks in advance!

Anahí

[GIS in Ecology]

Hi Anahi,

Thanks for your post.  I'm going to assume that you're using ArcGIS here, but you'll find similar tools that will do the same things in other GIS software.

I'm also going to assume that all you are looking for is the total number of polygons that overlap each grid cell, rather than needing to identify which polygons overlap with each grid cell.

If this is the case, the simplest way to do this will be to make sure that all the polygons are in the same data layer. If they aren't all already in the same data layer, then you can use either the MERGE tool or APPEND tool to add them all into the same data layer. Once you have your data layer containing all your polygons, you need to need to create a polygon grid using the FISHNET tool  with the required cell size and coverage. Once you have created this grid, you can use the SPATIAL JOIN tool to count how many polygons overlap with each grid cell. This is because when you use the SPATIAL JOIN tool, you can set it to count how many features in one data layer overlap with the features in another data layer. This is generally done using a MERGE rule of some kind so that the tool knows exactly how you wish to join the data together.. If you need the final result as a raster data layer, you can then convert your polygon grid data layer into a raster data layer using a convert tool, and base the conversion on the field that has the count information in it.

I hope this helps, and if this doesn't give you the results you're needing/doesn't work, feel free to post back on this thread and I'll see what else I can suggest as there are a few other possible methods to do this type of thing. Similarly, if you get stuck with any of the suggested steps and need more information, just let me know.

All the best,

Colin