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 - Gaps 3

0 (0 Likes / 0 Dislikes)
This is an example of what we need taxonomists to fill taxonomical gaps. This is our recent work by Angelica Harding Comparing what an expert knew about chameleons and what the dataset told about chameleons. The difference in taxonomists where ##### make ##### to discover that data that were supposed to be well-known after all they were #### were not really at all known. I will give you another example of what we need taxonomists to fill taxonomical gaps and also specific example of the knowledge transfer gap or gaps that arise from deficient knowledge transfer This example deals with the biosphere reserves that you know ####(they are spread all over the world) And which are places where assumably biodiversity should be rather well-known because they can be protected areas. Those are areas where traditional activities are kept and some other activities are not allowed because we want to preserve those areas as they are In the entire world, there are more than 600 biosphere reserves In 120 countries around the world And in these biosphere reserve areas, there are policy implementations, there is research, a lot of things are done there to preserve the areas as they are. We've studied the biosphere reserves of Mexico and Spain. I'll give you an example which is Mexico. One of the richest countries in terms of biosphere reserves. There are a lot of them. Those that are described in green. And what we did was to setup a workflow trying to see whether there was missing knowledge About the biosphere reserves by looking at how many species we knew from different sources. So, the general workflow was like this. We select biosphere reserves, a network of biosphere reserves. And we extracted information from different sources. For instance, we go to the management plans of each reserve Each reserve must have a management plan that explains what you can do, what you cannot do, how often you have to do to conduct a survey and etc. And this is a document we are supposed to prepare specifically for each reserve So, assumably it should contain the most #### (affordable/affirmative) set of knowledge about the reserve. We #### (result) to scientific literature. We scan the literature and we look for data that coincide with a biosphere reserve It can be data about biosphere reserve or it can be data about the area in the reserve or old data that happen in the same place as the biosphere reserve is. And we can go also to datasets that have been made available through GBIF which may or may not have been published. So, we compile lists. Basically, we want to get 3 lists. A list coming from management plans A list coming from literature and a list coming from a dataset. We compared the list. We reviewed the taxonomy, so we tried to #### (homogenize) names that is a very large exercise And quite complicated. This was done by one of my PhD students trying to get reconcile the different taxonomies Very much what we did yesterday with the whales. Try to get all the whales in the single name so we could compare all lists We harmonized the taxonomies, we need to use experts in this case. We used experts #### (in vertebrates) which were our ####(target) group We compiled the lists and tabulated. Finally, we compute intersections between the lists. #### We put all lists together and we see where are the coincidences. Which species coincide between lists And between 2 or 3 lists.And from then on, we calculate gaps. We calculate where is the lack of knowledge. So, those are some examples of the data sources we were using, like the red list of IUCN, GBIF, Bird Life, etc. We compiled data from several data sources. And finally, we got those three lists. A biosphere reserve, in order to compare the data, first we have to extract the data from the biosphere reserves. So, from the map of the biosphere reserve we added a buffer area, because the coordinates of the #### of the records could be uncertain So, we rather arbitrary added a buffer area that was much more precise that the area as defined by 4 dots in the management plans. Basically, what we did with GIS was to extract the information that was #### (coincided) with a particular reserve in all resources. Alright. So, once we had list, we compared them and we used probability theory So, we had a list in which for the entire Mexican biosphere reserves we had 1900 species. This was done for vertebrates only. Only vertebrates, not plants, not invertebrates. Next in the literature, we found 1700 species. There were more species found in the management plans as expected Because the management plans are specific studies of those reserves than in literature. And then we #### (mind / mined) the datasets And we found also 1700 species. The size of these plots here is proportional to the number of species richness. So we put all three together and saw what are the coincidences between all three sources. And we found that only half of the entire combined list contain species that were common to all three sources. Which left us with half of species which were missing from the sources. At least from one source. GBIF had 73% of the species, literature 72% and management plans 82% of the entire species or taxonomical space. All species combined when where 2400 different species. And they were also combined differently. Well, this section here is species that appear in the management plan and literature but do not appear in GBIF. And this area here, this area is proportional to number of species, are species appear in GBIF and literature but do not appear in management plans. OK? So, we found a lot of #### (exclusivities). 6% of the species only appear in the literature 9% of the species only appear in management plans 9% of the species only appear in GBIF. 3% both in GBIF and literature. 10% in literature and management plans. 10% in management plans and GBIF. If we split everything by taxon group, we see that fish is probably the least known group of species. Which means that if you are trying to do something with only one source of information, only management plans or only mining data from GBIF or only looking at the literature You are missing 70% of the actual knowledge. That's the lack of knowledge transfer. Because the knowledge occurred by the management plan should have gone into the literature And the knowledge in literature should have gone into the management plans. And all of them should have gone into datasets. If they aren't there, because there is no knowledge transfer. That's the gap. Well, this is even worse for amphibians. Still we are in one-third of common knowledge part. Reptiles is only one-third. And not surprisingly, birds are very well-known. with 72% commonality. And mammals is only 50%. We don't care about mammals or what? It may happen. And not interesting thing to see is that different resources behave differently too. Let's take only fish, and see what is the commonality of combined reserves. So, we take all reserves, we manufactured one single space which we call Mexico And we see what is known in Mexico from different sources. And for the entire Mexico in terms of biosphere reserves Is one-third known from all three sources, two-third known to partial sources. But if we take reserve by reserve and we do the average, it's even worse. If we take reserve by reserve and see what are the commonalities in each reserve separately and we take the average We say that one single reserve on average will have only 12% complete knowledge and will have almost 90% partial knowledge Including lots of exclusive knowledge for the average reserve which means that some of the reserves are well-known and some others may not be known If we go reserve by reserve, white means that the data come from more than one source and the colors are from exclusive sources so in the case of ###### This is the number of species which were known to at least two sources, two or three And this is the number of species that were known only to literature, only to dataset and only to management plans But it's just one case. We have many cases in which we can see that the amount of commonality can be low For instance, for this case, almost everything was known from literature and in this one everything was known only from datasets, there were no literature and no management plans And these ones also lack commonalities The #### that I put here is not arbitrary. It turns out that those reserves here are very low in diversity, those are high diversity reserves But also, the information gaps are narrow here and wider here. Those reserves here are in wide knowledge gap and those ones here are in narrow taxonomical knowledge gaps. So, this technique allows us to measure in a way how the taxonomical gaps are for special interest area which are biosphere reserves. If we had more data, we may get more information. This case analyzes how many of those species were actually listed as endangered and we discovered that even for the list that should be absolutely complete in ######### that this ##### the species that are under threats, endangered, there is a lot of knowledge lacking too. If we were in Spain, we may think OK Spain ######, no actually in Spain we found more commonalities but still a lot of commonalities especially in fish But we are only dealing with a very small fraction of the number of species that we have in Mexico. In Mexico, we have almost 3000 species, in Spain, we are dealing only with 500 species of vertebrates. Still, most of the country is covered by biosphere reserves And they don't tend to coincide with sites with high species richness Again, if we describe the data as listed or ####(unlist) in especial protection plans, we find that there are a lot of missing information about species that should have been covered by some kind of ruling and legislations or whatever. For instance, the green sectors here are species that are listed in IUCN However, ###### catalogue is missing species that actually appear in their lists That may mean two things: Either it's a gap by which the Spanish catalogue is missing important species Because ####(low) species are listed in IUCN red list or IUCN red list is including species that are not in Spain or have been badly classified or whatever Either the Spanish National Catalogue is wrong or IUCN is wrong. Which one would you bet? I probably think the Spanish National Catalogue is wrong. However, I happened to know that Spanish National Catalogue was compiled by a set of Ichthyologist who know very well The fish list So, in the case of fish actually, IUCN is wrong. So, ####(personally), #### how it works Now, we know that there are gaps. We've seen one technique to estimate taxon gaps. How do we estimate the extent of our universe? We need to estimate what is out there. And Town showed us in the morning, a number of ways to estimate the expected number of species Which ####(boiled) down to looking at the distribution of the species in list And this is probably the best measure that we have now. The measurement of completeness As Town explained, Completeness is based on the expectation of the number of species. Which means that for single inventories, Chao's work looks at the distributional of rare species Single-ton, double-tons is probably a ####(way) to go. We have one inventory, we look at the species that having #### only once or twice and we compile Chao's tool And then from there on, we use the observed number of species and expected number of species and get completeness as you already know very well However, there are also ultimate possibilities. As we saw before, we could use independent listings to try to derive this information and now the question is which one is precise? I don't have an answer to that yet. But for multiple inventories, we may use multinomial distribution which in fact was also used by Chao to derive her non-parametric system

Video Details

Duration: 17 minutes and 7 seconds
Country:
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.