Thank you for the video. From other sources, I see that beta is shown to be gamma/alpha. Does this not apply to this method of calculating beta? Is this because in this analysis you are comparing each site to one another? Thank you. I have a similar matrix (larger) of plant data at different sites along an elevational gradient on a mountain.
Hmm. Gamma/alpha can provide an index into how much each site contributes to gamma diversity for multiple sites. A larger value would indicate that the site contributes little to gamma diversity. As such, you could do that to determine quickly if you have sites that are really different from most sites. That difference will just reflect species richness, though. To actually address beta diversity is to consider species composition. That means not just how many species are different, but *which* species those are. For that you'd need a similarity index. There are multiple that might suit your needs and some wonderful sources online, e.g., www.semanticscholar.org/paper/Comparing-Communities-%3A-Using-%CE%B2-diversity-and-%2F-to-Jacobs/1d4fe58be2e592fb9ab8251cb14a04598b2dc415. ResearchGate can be a terrific resource too. If you have discrete sampling locations along your gradient, then I assume you're seeing gamma diversity as total species richness for the entire mountainside. Each sample site, plot, etc. will have its own alpha diversity. So far so good: those values are just integers. What beta does is allow you to consider composition among those sampling sites. Yes, the answer will still be an integer, e.g., 32 is the number of species unique to site G plus the number of species unique to site H in a pairwise comparison of G and H, but that number considers which species comprise those samples. For big species lists like you probably have, I stack them in a single column and then use the Pivot Table function in Excel to tell me how many of those species occurred at both sites and how many only occurred at one of the two sites (i.e., beta diversity). I divide gamma diversity for the pair by beta diversity for a proportional index, e.g., 0.25. I then subtract from 1.00 to give me a similarity index for the pair, e.g., 0.75. If you want something like that^ and you want to do more than individual pairwise comparisons, you'll want to set it up for real and analyze in one fell swoop using a real similarity index - and real software. My clunky Excel examples are really just to illustrate principles and help my students develop some basic skills. Thanks for your comment and good luck with your research!
@@timoconnell2206 Thank you so much for the explanation! I was getting confused with multiple ways of doing the index and what they were showing. It seems the gamma/alpha is more simplistic in comparing just richness, while calculating beta using the difference in composition of those species pairwise, that shows you a bigger picture of the species change across sites. I dumped my data into PAST, and did a beta analysis using Whittaker. It gave me decimals ie 0.66, 0.40 and on for paired sites in my data matrix. It is slightly different then when I calculate manually using the method you show in the video, which I assume is something to do with the Whittaker formula, but at least now I am mentally headed in the right direction of understanding! It is quite a labor of love to learn and teach oneself multivariate statistics for ecological studies, which I feel like I am grasping at the rate of a turtle haha. At any rate, I hope to finish my thesis and graduate this Spring. Thank you!
@@thebotanator1230 Terrific resources here on ResearchGate on interpretation of various similarity indices: www.researchgate.net/post/How-to-interpete-Whittaker-b-diversity-index. Perhaps even better: cals.arizona.edu/classes/rnr555/lecnotes/10.html. Whittaker index would be a fine approach for similarity along an elevational gradient. Good luck, Botanator!
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
How you got thus input data ? For eg amy pipeline u used
Thank you for the video. From other sources, I see that beta is shown to be gamma/alpha. Does this not apply to this method of calculating beta? Is this because in this analysis you are comparing each site to one another? Thank you. I have a similar matrix (larger) of plant data at different sites along an elevational gradient on a mountain.
Hmm. Gamma/alpha can provide an index into how much each site contributes to gamma diversity for multiple sites. A larger value would indicate that the site contributes little to gamma diversity. As such, you could do that to determine quickly if you have sites that are really different from most sites. That difference will just reflect species richness, though. To actually address beta diversity is to consider species composition. That means not just how many species are different, but *which* species those are. For that you'd need a similarity index. There are multiple that might suit your needs and some wonderful sources online, e.g., www.semanticscholar.org/paper/Comparing-Communities-%3A-Using-%CE%B2-diversity-and-%2F-to-Jacobs/1d4fe58be2e592fb9ab8251cb14a04598b2dc415. ResearchGate can be a terrific resource too.
If you have discrete sampling locations along your gradient, then I assume you're seeing gamma diversity as total species richness for the entire mountainside. Each sample site, plot, etc. will have its own alpha diversity. So far so good: those values are just integers. What beta does is allow you to consider composition among those sampling sites. Yes, the answer will still be an integer, e.g., 32 is the number of species unique to site G plus the number of species unique to site H in a pairwise comparison of G and H, but that number considers which species comprise those samples.
For big species lists like you probably have, I stack them in a single column and then use the Pivot Table function in Excel to tell me how many of those species occurred at both sites and how many only occurred at one of the two sites (i.e., beta diversity). I divide gamma diversity for the pair by beta diversity for a proportional index, e.g., 0.25. I then subtract from 1.00 to give me a similarity index for the pair, e.g., 0.75.
If you want something like that^ and you want to do more than individual pairwise comparisons, you'll want to set it up for real and analyze in one fell swoop using a real similarity index - and real software. My clunky Excel examples are really just to illustrate principles and help my students develop some basic skills. Thanks for your comment and good luck with your research!
@@timoconnell2206 Thank you so much for the explanation! I was getting confused with multiple ways of doing the index and what they were showing. It seems the gamma/alpha is more simplistic in comparing just richness, while calculating beta using the difference in composition of those species pairwise, that shows you a bigger picture of the species change across sites. I dumped my data into PAST, and did a beta analysis using Whittaker. It gave me decimals ie 0.66, 0.40 and on for paired sites in my data matrix. It is slightly different then when I calculate manually using the method you show in the video, which I assume is something to do with the Whittaker formula, but at least now I am mentally headed in the right direction of understanding! It is quite a labor of love to learn and teach oneself multivariate statistics for ecological studies, which I feel like I am grasping at the rate of a turtle haha. At any rate, I hope to finish my thesis and graduate this Spring. Thank you!
@@thebotanator1230 Terrific resources here on ResearchGate on interpretation of various similarity indices: www.researchgate.net/post/How-to-interpete-Whittaker-b-diversity-index. Perhaps even better: cals.arizona.edu/classes/rnr555/lecnotes/10.html. Whittaker index would be a fine approach for similarity along an elevational gradient. Good luck, Botanator!