WorldClim/ WorldClim2

[teixeirak]

Helene asks, doesn't WorldClim2 replace WorldClim? Are we using the latest version for the WorldClim variables?

[teixeirak]

Another question: did we use temperature seasonality (SD) or temperature seasonality (CV)?

[teixeirak]

Note: if we used temperature seasonality (SD) , we should divide by 100 to get the right units.

Here is a description of both:

Bio 4—Temperature Seasonality (Standard Deviation)

• Definition: The amount of temperature variation over a given year (or averaged years) based on the standard deviation (variation) of monthly temperature averages. The original AML® multi- plies the result by 100, which was designed to preserve significant digits, but in our calculations we do not do multiply by 100. • Units: Temperature (degrees Celsius) Interpretation: Temperature seasonality is a measure of temperature change over the course of the year. The AML® developed by Robert Hijmans and posted on Bioclim (http://www.worldclim.org/bioclim) calculates temperature seasonality using the standard deviation of the mean monthly temperature instead of deriving the temperature seasonality coefficient of variation. The larger the standard deviation, the greater the variability of temperature.

Bio 4a—Temperature Seasonality (CV)

• Definition: The amount of temperature variation over a given period based on the ratio of the standard devia- tion of the monthly mean temperatures to the mean monthly temperature (also known as the coefficient of variation (CV)). • Units: Percent • Interpretation: Temperature seasonality is a measure of temperature change over the course of the year. The temperature Coefficient of Variation (CV) is the ratio of the standard deviation of the monthly mean temperatures to the mean of the monthly temperatures (also known as the relative standard deviation) and is expressed as a percentage. CV therefore captures the dispersion in relative terms because standard deviation can produce two similar values while the means may be different. However, if variance is the same, an area with a lower mean temperature is distinguishable from an area with similar variance but with a higher mean temperature. The larger the percentage, the greater the variability of temperature.

[teixeirak]

to fix/ adjust based on this:

• [x] Fig. 4
• [x] Table S1
• [x] SI figs

[beckybanbury]

@teixeirak I've just seen that worldclim was updated in January - the reason some of the variables reference worldclim not worldclim2 is because when I ran the analyses not all variables were available in worldclim2 - I see that now everything is available. All the data I used was on the S drive at SCBI so I don't have access to that any more.

I'm fairly sure I used BIO4, SD.

[teixeirak]

Okay, that also matches what you put in Table S1. I also think 4a is new in v2.

So, based on that, we need to either divide T seasonality by 100 in all figures / analyses (because this will affect coefficients) OR adjust the axis labels on all figures. The former is preferable, but the latter easier, and acceptable.

It may be good to put a footnote in Table S1 indicating that WorldClim2 wasn't available at the time we downloaded the climate data.

[beckybanbury]

I've now updated T Seasonality/100 in all the figures - it doesn't affect which model is the best model or dAIC/Rsq values, so I don't think there is anything to update in the tables, but let me know if you catch anything. I believe that the units of degrees C are now correct for this?

[teixeirak]

I don't believe this affects anything in the manuscript or appendix. It would be good to make sure you don't have any outdated results in the GitHub repo (#105). I'll close this, now covered in #105.