Watch videos with subtitles in your language, upload your videos, create your own subtitles! Click here to learn more on "how to Dotsub"

BITC / Biodiversity Diagnoses - Inventory Completeness 4

0 (0 Likes / 0 Dislikes)
[Man]: #### because with plans with no money do, it was projected to do an inventory in an area. You calculate the site of the area and do something like the sampling intensity. OK, with this site, this area, I want to take this level of sampling intensity and I put this number of plots there. So, I don't know in that case we know that we have to put 25 plots, and at the 15th plot you realize that you have most of species What do you do? Do you stop it there? Or do you continue to 25 plots? [Town]: In a result-based sampling world, you would add plots until you seem to be done. OK? Now, notice that, remember, go back to the definitions that I gave of inventory vs. sample Inventory is something that you can do only when you can detect and count and catalog everything. And you sample plots are much more in the world of sampling because you put out 50 plots, but you don't know that there is some additional habitat or additional micro-habitat that wouldn't have it more plants. So, you've kind of set it up in a sampling world, but you could turn that into an inventory context by picking, remember we need a comparable unit of effort for the X-axis of these plots Or the horizontal in that matrix over there. It's time intensive, I know, but you could use your plots as those units of efforts. And in fact, there is another level and I think Arturo has put more thought into this that I have. You could very easily make this hierarchical. So you could imagine counting effort as you know hours of work First hour, second hour, third hour, fourth hour within a plot. And then the second hierarchical level being the plots. And then the plots could be distributed within major habitat types on a local landscape. And so essentially, what you are doing is you're going up from a single, let's say 10 * 10 m sampling plot to a whole habitat to a whole landscape. And at each level, you can look at the accumulation and the completeness. And at each level you can guess it how many species are left out. How many species have I likely not found withing this plot? How many species have I likely not found within this habitat? And how many species have I likely not found within this local landscape. So, it's a hierarchical analysis, so many have thought about that but certainly not done yet. It's very doable. Sorry, more questions? OK, one last bit, we're gonna talk a little bit about Scale, and this takes us back rather quickly to the regional question. Everything I have talked to you about until now has been points. You know, for insect people points are what? square meter? And for bird people, I'm ashamed to say, but our points are usually how far can a bird person walk in the day? So, we would rather coarse. So, even then when we want to go to a regional perspective, we very quickly get into the issues of scale, OK? And resolution.This is a virtual herbarium for Mexico. The publication that I gave you was an analysis of parallel data for Brazil. And my point is very simply, you know, if I look at a country at different resolutions, I get very different answers. So, if my pixel ... this is a work from Jorge Soberon, by the way, I stole it from him. But stole with credit. If my characterization of my region is in one pixel, then I've got a very coarse scale, right? Very coarse spatial resolution, but I've got tons of data, 691000 records. But I can start dividing that up, and we can come all the way down to a pixel like that. And remember we've talked about at this level, we're very much closer to a single site OK? And you're going to hear about alpha and beta diversity. We're much more measuring alpha diversity. At this level, we've got deserts, we've got sea channels, we've got mountain ranges, high desert, lowland rainforest There is a huge amount of diversity in there. So, all of Mexico, you can see about 1300 species of birds. But the richest site in Mexico, single site, would be somewhere down here with about 220 species of birds. All the rest is between site diversity. So, what Jorge did was to do a series of completeness analysis Starting in one pixel and splitting that into four, and then splitting those into four, and then splitting those into four And so watch this, at the level of the whole country, our inventory is essentially done, it's 78% complete. Probably more than that if you had better access to data. When we go to 500,000 square Km, pretty big pixels, you can see most of the country is pretty well-known But as we come to finer and finer pixels, look at this, we started getting gaps. OK? And those are still pretty big. Those are still very large area. Let's go finer, those are 22 * 22 Km. Those are still big pixels. And notice now we have some really big gaps, OK? And now just for fun, 5 Km pixels, at this scale you know nothing. It's 5.3 on a side. Twenty and fiveish Km square. And if we took this down to one square Km, you literally know nothing. And you know nothing for two reasons. One is that they're gaps amongst what literally you do know But the other is that you are not able to aggregate data and learn from nearby points. [Man]: those are for bird species? [Town]: No, this is the plant data. Again, remember I very artfully stole this from Soberon I think it's published but I stole the slides. OK, so the points from this multi-scaler thinking is that there is really no wrong answer or right answer This is looking at local knowledge and a couple scales up is looking at kind of local landscapes or regional landscapes But you do get very very different answers at different spatial resolutions. And so, one of the things that we've been trying to do in the precursors to this course has been to try to pick an optimal resolution. And maybe there is no optimal resolution, but at least we've been trying to pick one resolution Essentially, what we wanna do is not have a bunch of empty pictures of your countries and regions, but pick a level of aggregation, you know, maybe we're talking back to this, pick a level of aggregation that can still point out the gaps. But not be so pessimistic that you just say: OK, we're starting at zero. You know, we want you all to have a short list of highest priorities of where you would go if you were trying to finish the inventory of your region And in the second half of your career, you can do the real fine details. So, just to sum up, these inventories statistics give us a quantitative approach to understanding how complete biotic inventories are They are crucial elements in understanding inventory data to the point where I would say if you read a paper that is a flora of, or a bird of, or whatever And it doesn't do some sort of analysis of the completeness, I'm talking about modern papers, it's not a complete paper and I would say it's not publishable. Right? Because, you know, somebody tells you here 100 species of birds from this place, what does that mean? You know, those sites where I was in Mongolian deserts, there aren't 100 species there. Maybe, if you stay there for 10 years and saw every last migratory species, maybe you get to 100 and that is done. Where at sites in the Congo or in the Amazon, 100 species is partial. So, those inventory graphics and statistics really give us a way of thinking about what that list means. Is it a full list? Is it a partial list? Is it getting down to the very last details, or is it just a start? So, is very key documenting which sites are assessed thoroughly. Very very important in understanding absence. Maybe there is a flora of some mountain in the Cameroon mountains, Moses. And it doesn't mention species. Now, you know those plants, maybe you can say: well, any botanist who walks into a forest will see this species. So, it must not be there. But it also would help to see some accumulation curves and some statistics that might suggest, no there is still 100 species left to discover there or this done. So, maybe I can believe the absence of that species. Certainly, the statistics allow us to optimize our use of resources in future inventory efforts. and certainly, they are crucial to conservation efforts where if you are going to invest in a protected area, you probably ought to know clearly which species that area protects and which species that area doesn't protect.

Video Details

Duration: 13 minutes and 5 seconds
Language: English
License: Dotsub - Standard License
Genre: None
Views: 0
Posted by: townpeterson on Jul 26, 2016

This talk was presented in the course on National Biodiversity Diagnoses, an advanced course focused on developing summaries of state of knowledge of particular taxa for countries and regions. The workshop was held in Entebbe, Uganda, during 12-17 January 2015. Workshop organized by the Biodiversity Informatics Training Curriculum, with funding from the JRS Biodiversity Foundation.

Caption and Translate

    Sign In/Register for Dotsub above to caption this video.