Thanks for explaining this! Excel is my least favorite part of any study but it's so important nowadays, and you made this very simple and easy to understand. I wish you had been the one to teach my environmental statistics course!
Gracias por su opinión. Me alegra saber que esto es útil. También he grabado uno nuevo que debería mostrarse en HD, si eso ayuda. (Y mi español es rudimentario, así que estoy usando DeepL como ayuda para la traducción.)
Mathematically it should work so long as the individual per cent covers sum to 100 - be sure you're clear about percent (100) vs proportion (1.00) in calculations. BUT, be very clear in your application that you're not specifically comparing the same thing with percent cover as you are with abundance of individuals. One big leaf could cover a lot of area in a small plot. Conversely, entire individuals of various seedlings might occupy a really small area. I'd say percent cover is at best analogous to abundance. With those caveats, why not?!
Thank you very much for this video. May I ask you about missing data? If we actually know that there are 18 species in the sample, but we can find (collect) only 15 species from this. How should we calculate the Shannon diversity index? We can use only 15 species that we have, or we cannot use this sample anymore. Thanks.
If you're comparing among sites then you wouldn't want to add species to one sample from sources outside your standardized surveys. If you used multiple sources of information to generate a species list that's fine, so long as you do it the same way across all of your sites. The point is just to make sure that you're comparing among sites for which there has been similar sampling effort, but what that effort is can be up to you. One issue might be that you know a species occurs at a site and you want to include it but you have no abundance estimate for it. You could just count each species like that as a 1 and you could still run the calculation. The math would work, but you'd need to be careful about how you interpret the results you got.
A diversity index should function like any other continuous variable. You would consider the index value as your response variable and environmental variables as your predictors.
I was wondering if you could help me. I've been told to calculate shannon diversity index but we have all percentages and have been told to convert this percentage into proportions of the sample catch size (Pi).so say if we had a total of 50% and only 3 species present would I just need to divide 50 by 3?
The information you need includes the total number of species captured - that's species richness - the total number of individuals of ALL species captured, and the total number of individuals of EACH species captured. In your case I'd begin with acknowledging that if you only have 3 species as your S then you certainly don't need a biodiversity index to help you make sense of the data. Calculating a biodiversity index even for the 10 species in my example is pushing the bounds of relevance anyway. A biodiversity index is something we use to help us make sense of otherwise confusing data on many different species - say dozens to hundreds. For example, imagine you were comparing among 5 different sites with respective species richness of 32, 29, 40, 27, and 33. Shannon's index could be helpful in illuminating which if any of those sites were significantly more diverse than others, irrespective of their species richness. The site with 27 species could be more diverse than the site with 40. Now if you already have percentages/proportions of the different species in your capture data (and 50% is just 0.50), the important thing is to be sure what that 0.50 is a proportion of. For this calculation you can NOT use the 0.50 of the 3 total species, it must be a proportion of all *individuals* of all species. If you caught 200 individuals of 3 different species and 100 of those individuals came from 1 species --> that's what you need here.
@@timoconnell2206 Thank you so much for getting back to me! So the data we have it only has percentage cover for each site and then at each site we have given a percentage of each species that is present so we do not have individual counts of each species. If i wanted to compare the percentage cover between these sites, theres a total of 8, what kind of statsistics would I be able to use? I have been trying to work it out all day but there is nothing thats like use this for percentage cover. Do you have any ideas?
@@jackwarren4519 sorry for my confusion, but are you saying that you have 8 sites and at each you have proportional LAND cover? You can calculate Shannon's diversity index as "landscape diversity" within a defined area, but only with raster data for your land cover. If each pixel is coded as a particular cover type (e.g., "forest", "crop field", "open water") then you can treat the number of pixels of each type as the number of "individuals" of each "species". Computationally, it's the same as in my example. You'll get an index score that you can compare among all your sites.
@@timoconnell2206 no it’s fine. I’m very confused with this too. So what we have is 8 sites and at each site we have a structure in which we did a total percentage cover on each structure and then noted how much of a species covered that. So e.g. structure 1 80% bare mud and 10% Salicorna perennis and 10% spartina anglica. I’m not too sure if that is land cover. Also what do you mean by pixel and how to work that out? Because what I tried I just had my percentage say 10% and just used the value 10 as my count number using your example but I don’t think that’s right. Thank you for helping me! I haven’t been able to nm find much help else where
@@jackwarren4519 Aha, you want to know something about proportional cover of vegetation by species in a salt marsh. Sounds like you visually estimated percent cover by species in a standardized sampling area, e.g., a Daubenmire frame. You have two species - three if you count bare mud as a "species". Now, what do you want to *do* with those 8 sites, rank them by "diversity" or something? You can't calculate Shannon index with the data you describe but the good news is that you wouldn't want to. It'd be a silly way to try to make sense of your system with so few species to work with. What you *should* do depends on your objective. That will determine your method.
H' is an estimate index of species diversity that includes abundance of each species in the estimate. Yes, H' increases if the number of species increases, but those gains can be offset if there is great unevenness in abundance among the species. My example illustrates this with simulated data.
Correct. The Shannon diversity index tells you only something about the number of species and the abundance of species at a site. You can compare among sites using equitability, but if you want to compare abundance of the same species among multiple sites (including sites where the species does not occur at all), then you are better served with some kind of index of community similarity. Good explanation of different methods here: cals.arizona.edu/classes/rnr555/lecnotes/10.html
Sorry, I'm not familiar with that. If your data are stem counts by species that would work. If basal area or canopy coverage by species then the process would be similar in that the "abundance" of different species is considered. So long as you were clear that you'd be basing the calculations on an abundance estimate that was not the number of individuals, then you should be okay to use Shannon Diversity Index.
@@timoconnell2206 I tried to make a graph of the results. The paper I had to write, has already been handed it. If I funk it's your fault but don't worry because I was already funking the horrible, nightmare ornithology class anyway. It was awfully hard to resist saying in the paper that stinking birds aren't being very good canaries in the coal mine for us these days. At least 50% of the American people have some kind of chronic disease (like cancer, autoimmune disease, diabetes) ... I think I remember reading somewhere that 7 million people a year, world wide die from pollution.. 7,000,000 a year. Why are there any bird left at all? Evident we've switched places. The canaries are watching humans die and leaving the coal mines that our gov't has turned our country into...by failing to regulate for our health and safety. Well, that was fun... I should be studying for the Ornithology final exam... dang birds. Actually, a lot of bird species have gone extinct in a short period of time ...far more birds are dying than human... but humans are too busy watching Fox and their smart cancer screens to notice. How much longer can I procrastinate from studying...
Proportion of Hmax (equitability) ranges 0.00-1.00. If you calculated a Shannon Diversity Index score of 1.64 with an equitability of 0.60, then that is not a problem. One (admittedly simplistic) interpretation is that the diversity in your sample is ~60% of the maximum it could possibly be for the number of species considered.
Thanks for explaining this! Excel is my least favorite part of any study but it's so important nowadays, and you made this very simple and easy to understand. I wish you had been the one to teach my environmental statistics course!
I LOVE YOU!!!! THANK YOU!!! YOU SAVED MY LIFE
Hahaha -- YES, this is why I've put this out to UA-cam. I want people to find it and see how easy it can be. Congratulations!
I think you just helped me with a major hurdle in my thesis, thank you so much!
This video just saved my life!! Thank you for this!
Thank you so much for uploading this. Understanding the formula within the textbook took way longer. Excel for the win!
Plain old Microsoft Excel is a vastly underappreciated resource. I'm glad you found the video useful!
WONDERFUL!!! You have helped me a lot
Thanks a lot for sharing this on UA-cam, it si really helpfull. Regards from Buenos Aires Argentina 🇦🇷❤️
This helped IMMENSELY, thank you!!! This was really well detailed and quick!!
God bless you, sir. Best video I've found of this.
This was very helpful. Thank you so much sir!
I guess I am quite randomly asking but do anyone know of a good website to stream newly released movies online?
@Rhett Keagan I would suggest flixzone. You can find it by googling :)
@Alvaro Fletcher yup, I have been watching on FlixZone for years myself :)
@Alvaro Fletcher thanks, I signed up and it seems like they got a lot of movies there :) I appreciate it !!
@Rhett Keagan you are welcome :)
Thank you for this- you are a great teacher! 🙏🏻
Saludos, mi ingles no es perfecto pero entendí y fue fenomenal escucharlo, gracias por enseñarnos!
Gracias por su opinión. Me alegra saber que esto es útil. También he grabado uno nuevo que debería mostrarse en HD, si eso ayuda. (Y mi español es rudimentario, así que estoy usando DeepL como ayuda para la traducción.)
@@timoconnell2206 Gracias por responder en mi idioma y pueda entenderlo, valoro mucho eso, me suscribo para estar pendiente de mas videos
Perfect, very informative. Thank you so much for your effort
Thank you so much for your excellent explanation!
Sir i need guidance regarding spatial structure of forest by nearest neighbor method
Thanks a lot, you explain so well this bioestadistic problem, cheers.
Can this be used with percent cover of each species instead of total abundance?
Mathematically it should work so long as the individual per cent covers sum to 100 - be sure you're clear about percent (100) vs proportion (1.00) in calculations. BUT, be very clear in your application that you're not specifically comparing the same thing with percent cover as you are with abundance of individuals. One big leaf could cover a lot of area in a small plot. Conversely, entire individuals of various seedlings might occupy a really small area. I'd say percent cover is at best analogous to abundance. With those caveats, why not?!
Respected Sir, Thank you so much for your incredible video .....stay blessed
Can you get all the answer of In(pi) and pi[In(pi)] please !!!! I'm dying now ....
Species. / ni
A. / 3
B. / 6
C. / 5
D. / 8
E. / 2
I wonder if there is somebody who can interpret the evenness result (value) versus Shannon Wiener index (H) value and what is the relationship?
Thank you very much for this video. May I ask you about missing data? If we actually know that there are 18 species in the sample, but we can find (collect) only 15 species from this. How should we calculate the Shannon diversity index? We can use only 15 species that we have, or we cannot use this sample anymore. Thanks.
If you're comparing among sites then you wouldn't want to add species to one sample from sources outside your standardized surveys. If you used multiple sources of information to generate a species list that's fine, so long as you do it the same way across all of your sites. The point is just to make sure that you're comparing among sites for which there has been similar sampling effort, but what that effort is can be up to you.
One issue might be that you know a species occurs at a site and you want to include it but you have no abundance estimate for it. You could just count each species like that as a 1 and you could still run the calculation. The math would work, but you'd need to be careful about how you interpret the results you got.
Is it the same process when using percentage cover? You simply swap the percentage cover for number of species?
Well, a mathematician or a more quantitative ecologist might have a different opinion but yes, it still works.
Awesome video thank you!
Dear sir and everyone, Please i want to create an index for a panel data using PCA. can someone help me, please
Hello sir how to find correlation and significance value between environmental variables and diversity indices
A diversity index should function like any other continuous variable. You would consider the index value as your response variable and environmental variables as your predictors.
I was wondering if you could help me. I've been told to calculate shannon diversity index but we have all percentages and have been told to convert this percentage into proportions of the sample catch size (Pi).so say if we had a total of 50% and only 3 species present would I just need to divide 50 by 3?
The information you need includes the total number of species captured - that's species richness - the total number of individuals of ALL species captured, and the total number of individuals of EACH species captured.
In your case I'd begin with acknowledging that if you only have 3 species as your S then you certainly don't need a biodiversity index to help you make sense of the data. Calculating a biodiversity index even for the 10 species in my example is pushing the bounds of relevance anyway. A biodiversity index is something we use to help us make sense of otherwise confusing data on many different species - say dozens to hundreds. For example, imagine you were comparing among 5 different sites with respective species richness of 32, 29, 40, 27, and 33. Shannon's index could be helpful in illuminating which if any of those sites were significantly more diverse than others, irrespective of their species richness. The site with 27 species could be more diverse than the site with 40.
Now if you already have percentages/proportions of the different species in your capture data (and 50% is just 0.50), the important thing is to be sure what that 0.50 is a proportion of. For this calculation you can NOT use the 0.50 of the 3 total species, it must be a proportion of all *individuals* of all species. If you caught 200 individuals of 3 different species and 100 of those individuals came from 1 species --> that's what you need here.
@@timoconnell2206 Thank you so much for getting back to me! So the data we have it only has percentage cover for each site and then at each site we have given a percentage of each species that is present so we do not have individual counts of each species. If i wanted to compare the percentage cover between these sites, theres a total of 8, what kind of statsistics would I be able to use? I have been trying to work it out all day but there is nothing thats like use this for percentage cover. Do you have any ideas?
@@jackwarren4519 sorry for my confusion, but are you saying that you have 8 sites and at each you have proportional LAND cover? You can calculate Shannon's diversity index as "landscape diversity" within a defined area, but only with raster data for your land cover. If each pixel is coded as a particular cover type (e.g., "forest", "crop field", "open water") then you can treat the number of pixels of each type as the number of "individuals" of each "species". Computationally, it's the same as in my example. You'll get an index score that you can compare among all your sites.
@@timoconnell2206 no it’s fine. I’m very confused with this too. So what we have is 8 sites and at each site we have a structure in which we did a total percentage cover on each structure and then noted how much of a species covered that. So e.g. structure 1 80% bare mud and 10% Salicorna perennis and 10% spartina anglica. I’m not too sure if that is land cover. Also what do you mean by pixel and how to work that out?
Because what I tried I just had my percentage say 10% and just used the value 10 as my count number using your example but I don’t think that’s right. Thank you for helping me! I haven’t been able to nm find much help else where
@@jackwarren4519 Aha, you want to know something about proportional cover of vegetation by species in a salt marsh. Sounds like you visually estimated percent cover by species in a standardized sampling area, e.g., a Daubenmire frame. You have two species - three if you count bare mud as a "species".
Now, what do you want to *do* with those 8 sites, rank them by "diversity" or something?
You can't calculate Shannon index with the data you describe but the good news is that you wouldn't want to. It'd be a silly way to try to make sense of your system with so few species to work with.
What you *should* do depends on your objective. That will determine your method.
Respected Sir, more the value of H, indicates the more species diversity , is directly propotional?
H' is an estimate index of species diversity that includes abundance of each species in the estimate. Yes, H' increases if the number of species increases, but those gains can be offset if there is great unevenness in abundance among the species. My example illustrates this with simulated data.
Thank you so much!!!!!!!!❤
what if you have a count of 0 for some species at site a and then a count for that species as 1 or more etc a site b? i cant do LN on a count of 0
Correct. The Shannon diversity index tells you only something about the number of species and the abundance of species at a site. You can compare among sites using equitability, but if you want to compare abundance of the same species among multiple sites (including sites where the species does not occur at all), then you are better served with some kind of index of community similarity. Good explanation of different methods here: cals.arizona.edu/classes/rnr555/lecnotes/10.html
Well explained
well can u upload ivi value calculation which are related to forestry standard as im forestry student
Sorry, I'm not familiar with that. If your data are stem counts by species that would work. If basal area or canopy coverage by species then the process would be similar in that the "abundance" of different species is considered. So long as you were clear that you'd be basing the calculations on an abundance estimate that was not the number of individuals, then you should be okay to use Shannon Diversity Index.
Thank you very much, helped a lot!
then how do you do a graph?
You would like to make a graph of Shannon Index values?
@@timoconnell2206 I tried to make a graph of the results. The paper I had to write, has already been handed it. If I funk it's your fault but don't worry because I was already funking the horrible, nightmare ornithology class anyway. It was awfully hard to resist saying in the paper that stinking birds aren't being very good canaries in the coal mine for us these days. At least 50% of the American people have some kind of chronic disease (like cancer, autoimmune disease, diabetes) ... I think I remember reading somewhere that 7 million people a year, world wide die from pollution.. 7,000,000 a year. Why are there any bird left at all? Evident we've switched places. The canaries are watching humans die and leaving the coal mines that our gov't has turned our country into...by failing to regulate for our health and safety. Well, that was fun... I should be studying for the Ornithology final exam... dang birds. Actually, a lot of bird species have gone extinct in a short period of time ...far more birds are dying than human... but humans are too busy watching Fox and their smart cancer screens to notice. How much longer can I procrastinate from studying...
Can you get all the answer of In(pi) and pi[In(pi)] please !!!! I'm dying now ....
Species. / ni
A. / 3
B. / 6
C. / 5
D. / 8
E. / 2
Useful, thank you so much!
great explanations ! thanks !
Very good! Thanks a lot!
The value of H= 1.64 and Hmax= 0.6, dont understand to to fix this problem
Proportion of Hmax (equitability) ranges 0.00-1.00. If you calculated a Shannon Diversity Index score of 1.64 with an equitability of 0.60, then that is not a problem. One (admittedly simplistic) interpretation is that the diversity in your sample is ~60% of the maximum it could possibly be for the number of species considered.
Thanks a lot. Very helpful
Interesting 😍
thank you a lot sir...nice and easy :-)
Thank you so much.
thank you!
Thanks!!
thanks a lot
Great.... Job
Tnk u
Can you get all the answer of In(pi) and pi[In(pi)] please !!!! I'm dying now ....
Species. / ni
A. / 3
B. / 6
C. / 5
D. / 8
E. / 2
Sorry, I'm not sure I understand your question. Are you plugging in your numbers but something isn't working?
Too blur
Sorry for that, but you might be able to adjust your settings to play it back at higher resolution.
Thank you so much. super helpful