Assessing and ranking the relationships and contribution of environmental correlates to species richness

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Assessing and ranking the relationships and contribution of environmental correlates to species richness

ah3881
I am trying to run some analysis to determine the percentage contribution
of various factors (minimum temp in lgm, npp, etc) on determining
species richness of 171 species throughout Southeast Asia (on a km by km
basis-so over 4 million rows of cells, and about 14 columns).
I have read about various stats tests, and softwares-but I am no
statistician and I would really appreciate some advice as to the best
method/tests.
So far I have been attempting SPSS, but it does not give me the outputs I need, and a colleague suggested R might be the best way to analyse the data
Thanks in advance
Reply | Threaded
Open this post in threaded view
|

Re: Assessing and ranking the relationships and contribution of environmental correlates to species richness

Mathieu Rajerison
Hi,


How is represented your data: a regular grid: raster, irregular samples:
points, or slots/parcels: areas?

If you want to determine species richness, you can use simpson and shannon
indexes with diversity package.

On a point pattern object, use spatstat::marktable function as mentionned in
http://r-sig-geo.2731867.n2.nabble.com/density-diversity-of-points-td6355348.html

On a raster, use raster::focal

Then you may perform a linear regression with your explanatory variables.

2011/7/3 ah3881 <[hidden email]>

> I am trying to run some analysis to determine the percentage contribution
> of various factors (minimum temp in lgm, npp, etc) on determining
> species richness of 171 species throughout Southeast Asia (on a km by km
> basis-so over 4 million rows of cells, and about 14 columns).
> I have read about various stats tests, and softwares-but I am no
> statistician and I would really appreciate some advice as to the best
> method/tests.
> So far I have been attempting SPSS, but it does not give me the outputs I
> need, and a colleague suggested R might be the best way to analyse the data
> Thanks in advance
>
> --
> View this message in context:
> http://r-sig-geo.2731867.n2.nabble.com/Assessing-and-ranking-the-relationships-and-contribution-of-environmental-correlates-to-species-richs-tp6543945p6543945.html
> Sent from the R-sig-geo mailing list archive at Nabble.com.
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo
Reply | Threaded
Open this post in threaded view
|

Re: Assessing and ranking the relationships and contribution of environmental correlates to species richness

ah3881
Hi

Thanks.
I have species richness projections (made using Maxent projections of species distribution of 171 species, reclassified to give binary estimates of species distributions {10 percentile training threshold) before combining all 171 projections).
So I have a raster of species richness in addition to all the possible correlates I am examining (npp, intra-annual variability of npp, inter-annual variability of npp, distance from coast, distance from karsts, latitude, lgm minimum temperature, lgm mean temperature, temperature change in minimum temperature since lgm, t change in mean temp since lgm, actual evapotranspiration)
None of the correlates were used in species distribution projections and I have a gis raster of each variable, in addition to a database denoting the value of each for every km2 throughout Southeast Asia.

Now I want to determine how much variability in species richness can be explained by variation in each of the other variables, in addition to the relationships. But I don't know which test would be best to do this, or software (GIS or database) which could best analyse this-especially given that the database is over 4520000 rows, and a large number of columns.....
Advice would be greatly appreciated
Reply | Threaded
Open this post in threaded view
|

Re: Assessing and ranking the relationships and contribution of environmental correlates to species richness

Mathieu Rajerison
So, it seems you have one raster for your dependent variable: species
richness and one for each explanatory variable.

I have never accomplished this but I think you should make a raster stack of
all these variables using raster package: the dependent one + the
explanatory ones then fit a model, for instance using a linear regression
(lm function)

With the model, you'll get coefficients for each exp. variables and
p-values. You'll trhen see which ones are significant and contribute the
most to explain species richness.

Hope this helps..

Mathieu

2011/7/4 ah3881 <[hidden email]>

> Hi
>
> Thanks.
> I have species richness projections (made using Maxent projections of
> species distribution of 171 species, reclassified to give binary estimates
> of species distributions {10 percentile training threshold) before
> combining
> all 171 projections).
> So I have a raster of species richness in addition to all the possible
> correlates I am examining (npp, intra-annual variability of npp,
> inter-annual variability of npp, distance from coast, distance from karsts,
> latitude, lgm minimum temperature, lgm mean temperature, temperature change
> in minimum temperature since lgm, t change in mean temp since lgm, actual
> evapotranspiration)
> None of the correlates were used in species distribution projections and I
> have a gis raster of each variable, in addition to a database denoting the
> value of each for every km2 throughout Southeast Asia.
>
> Now I want to determine how much variability in species richness can be
> explained by variation in each of the other variables, in addition to the
> relationships. But I don't know which test would be best to do this, or
> software (GIS or database) which could best analyse this-especially given
> that the database is over 4520000 rows, and a large number of columns.....
> Advice would be greatly appreciated
>
> --
> View this message in context:
> http://r-sig-geo.2731867.n2.nabble.com/Assessing-and-ranking-the-relationships-and-contribution-of-environmental-correlates-to-species-richs-tp6543945p6545267.html
> Sent from the R-sig-geo mailing list archive at Nabble.com.
>
> _______________________________________________
> R-sig-Geo mailing list
> [hidden email]
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo
>

        [[alternative HTML version deleted]]

_______________________________________________
R-sig-Geo mailing list
[hidden email]
https://stat.ethz.ch/mailman/listinfo/r-sig-geo