"Becoming Friends with the Odd Man Out" by Paloma Ruiz de Diego

[Applause] hi everyone so I'm Paloma and I will be talking about my PhD but also about the project in which my PhD is included so um this is me I did my undergrad in the University autonoma de Madrid in biology and then I specialized in plant and fungal taxonomy conservation and diversity at Q and finally I went back to Madrid to do my PhD in the Royal botanic garden there so um uh we work here that's our garden it's really small comparing to yours and um my PhD it's included in this project that is called the Y Man Out project and basically we with um tropical tropical plants tropical plants diversity patterns in tropical Africa so we mainly focus on the Continental Tropical Africa and it's it's funded by the Spanish government this is our team as you can see and also all the institutions that work with us uh you can see Patricia Bara up up top and a little bit about uh the concepts of our project so um we work with the old man out PN it was described by Richards in 1973 and um it was described as a lower diversity in the tropical Continental part of Africa than in other tropical areas so if you compare with Madagascar that is an special case but also if you compare with tropical Africa or the near Tropics tropical Africa no sorry tropical Asia and you can see there uh the graph in the a you can see what we called uh the tropical region in Continental Africa and you can see that the real tropical rainforest is a really small area um in the v you can see a graph of the amount of species that you can see in u the different tropical regions and uh tropical Africa is also small part of that in C C is more interesting than it s and you can see the proportion of gener and the amount of species that they have in all of the regions and you can see that in Africa you have more smaller gener and less big or medium gener and finally you can see in the D uh the number of species versus the surface area of the tropical forest and you can see that the pattern is there so um you can observe the pattern but why is happening we want to answer that uh classically there are more hypothesis that those that you can see but those are the ones that we focus in um the first is time to spaci so uh it talks about uh that the lower diversity in Africa could be explained uh because the colonization of the species uh is more recent than in other tropical areas and there there's not enough time to accumulate that diversity the second hypothesis is that this lower diversity could be explained for a more um stable tectonical history so from the Paleo geology to now uh you will see less hogenes and less uh like movements of tectonical place less accidents that is not so correct nowadays because the the latest um geological P geological studies shown that has there has been a lot of activity in Africa also um then you have the area for speciation one that um implies that the lower diversity may be related to a reduction of the tropical Zone um from the Paleo paleo climate till now so um you will have less area so you will have less amount of terrain for diversity to occur and then you have the evolutionary data n one that is an special one so it's not focused on the climate or the area itself but on the characters that the lineages in Africa accumulate so they saw that some characters like for for example di dii um accumulate more in the lineages of Africa than in other parts and uh new studies in Anon and other families show that those characters that could be a handicap for diversity are not so important they are not causing so much extinctions or less diversity than in the langes that doesn't S the characters uh but we want to test it out with our model clades and finally you have the H5 that is the pooc climatical instability so uh this hypothesis uh associate the lower diversity that we can see with uh climate change events in the history of Africa so um you will have periods of aridification or you will have um uh dry uh like really large uh dry epochs after um cooling events and so historically and uh they can be linked to um moments of flower diversity or even Extinction in the past of Africa so uh what is our model CL or our uh group of a study they are the for vac um it's one of the most diverse family as you all know and is not only diverse in terms of species or gener it's also diverse in terms of um uh habitats morphological strategies physiological advantages and changes and you can see you can see that in that image you have from uh trees to geoides on cacti and you have everything that you want or everything that you can imagine you have it and uh we selected uh 35 gener and uh 19 clayes those Clays come from the wock uh philogyny of 20 2005 that is the one that you can see there and they are um marked in Orange so um why we choose those clades we chose them because they were diverse so you have really big ones really small ones gener that are uh Mega diverse and others that are even monot typical uh you have gener that are really diverse in terms of morphology or habitats and ones that are really stable and regarding morphology we were focusing on sexual system but also in habitat and also in like uh strategies different strategies that you could you could use so which ones are the the project goals uh the first goal that we have is to increase the taxonomical and the taxonomic knowledge about afrotropical Africa the second is to infer The evolutionary relationships of these claves and the Divergence times of all of them and to see which patterns SCH emerge of each of the cles of or gener that we work in then we want to ex examine the explanatory hypothesis of the o man out pattern so all the ones that I have show you we have to we want to examine them with all with all of the model clades with every of the gener and finally we want to synthesize all of this information to answer micro micro evolutive questions but uh how are we going to do this on the one hand uh we are going to do a lot of field workor because we want um uh to complete a really big sampling um Patricia went to Gabon and we went together to Congo but also and mainly we are getting all allot of our samples from edar and private collections so we have been through all the world all around and one of the aaria in which we work more is here in Missouri but also Patricia went to Tanzania and we have been to all the main collections in the in the in Europe like for example Brussels or um L Lisbon yeah Lisbon Edinburgh um also we are doing morphological comparative studies uh this was like a pilot study that started here at Missouri with the Madagascar specimens and now we want to uh update that and include all the African species too and we also want to do taxonomical reviews um but this takes time as all of you know and um it's linked with the second gold um we also are doing uh ucn assessments to Target the conservation part of the of the project uh then we get into the uh genomic uh part of the of the of the work and we are using Target sequencing or hypc uh that it's a different kind of whole sequence whole GM sequencing that you maybe know and and we are using theoria bait kit that includes includes 431 nuclear genes that are um that it's they are different from the ones that are included in the inper one they don't they only um um they only uh match in like a really small uh percentage and what is different about the hypc so um on the one hand you will start normally like with the with the libraries and with the DNA struction prevor to the libraries but then you will do an enrichment that enrichments uh includes um a lot of uh hybridization bait that are like Mega primers it's not the same technology it's not the same chemistry but what it does is that it will enrich the regions the your gen your libraries in the regions that you want that that are your Target and when you sequence that and then you process the data you will for sure have uh more sequences of that region that it's interesting for you and your group of a study and then when we have all the data we will process it and our our goal is to make phoenis and then to date those phenis to study the Divergence of the groups uh we do that with a really complex um pipeline that includes a lot of uh convolutional processes and uh quality checks finally you will have um an species tree and also all of the genene trees that allows you to do uh uh concatenated Matrix so like a super Gene tree and then when we have all the foges and we have dated them um we will use micro evolutive models to try to um test all of the hypotheses that I have talk about and those models are really complex and depending of what data you are comining your phenomic data with you will use one or another those are only an example and uh for example the one that you have here on the left is more focus on focused on um using characters with your phenomic data and we will be using that to test for the uh dead end hypothesis with sexual system for example and what we aim to do is that we want to have our phog genomical data and all of the other layers of data you can put on top so like um pooc climatical data sexual system data characters artical factors all that you can think of and we want to do complex analysis that allow us to test the hypothesis that we have set and finally I want to do a macro synthesis of everything and this is the point in which our uh project differs from the classicals approaches that you will find in biogeography um in biogeography there's this fight that in which you have either model CL studies with species that you really know that are small data sets that you can answer micro evolutive questions and those other really big data sets or of different gener even different families uh of which you maybe not are not an expert but that allow you to ask uh macro evolutive questions and we want to try to bridge them so uh we want to do a lot of model Cates that are related to one another so within theb family and um when we have all the information about all of them and we have a studied their taxonomy and we are uh working with experts in those clades when you have all the information you can analyze each of the clades and when you have all of the micro evolutive questions answered as far as you can then you combine all of them with different methods that I will talk a little bit now um to try to answer your micro evolutive questions and which ones are those methods on the one hand you have ban inference and this is a Vib model that usually is used for U Iceland um bio geography and you will have um different data sets per one of the clades for for each clay you will have a data set and you will do a me analysis of all of them comparing their uh rates so diversification rate um H speciation rate Extinction rate whatever you have and all of the uh information that is necessary so if you are um putting all the data sets by Iceland you will use um capacity of the Iceland distance and so we will be using for example number of species or the the sampling percentage that we have it depends on what is your data and the other approach that is is still um a little bit something really really preliminary is using machine learning models so um we want to train CNN so convolutional neural networks to learn of the data sets that we have worked on to predict larger data sets or data set that you have not worked on previously and that's that's a lot of things to do um finally so preliminary results that we have um is our sampling we are close to 70 to 80% of the targeted species in sampling um it has been an an immense uh work um we also have more than 1,200 uh DNA samples and we have sequence it at the moment uh 520 uh samples so that's a lot of data but we will be um having more and more because uh there's another uh PhD student in our project and she's trying to complete our sampling so um as I said we we did some field work um we went to the Republic of the Congo and we worked with the university Marin and guabi um uh Personnel um and we went to the field we we return with uh 200 246 47 numbers and um a duplicate of of all of them will be coming here when we get our work done and uh Patricia was the main uh Pi both regarding the Royal botanic garden and the Missouri Botanical Garden so we are also working in conservation um we are collaborating with the gabot tree um group of the iucn assessments so um this group is also working with the garden and we enter via Patricia and we uh usually um review each other assessments and do meetings in which we talk about our like concerns problems uh criteria and so we have published uh these four uh assessments and until now and we are working in a really large data set that includes mainly Edge um Gena so monot typical or really small gener species and these are some preliminar results they are different than the from than the one that you will see if you go to the IBC because I'm working at that in the moment uh but um we have a preliminary philogyny for the Continental African species it includes uh 160 samples and uh regarding the sampling is more or less are 70% of the species of Africa Continental Africa and it has a representation of Madagascar Asia and also the neotropics because uh we thought that for Croton it would be very important to see um how the species come go from the continent to Madagascar and and back because we know that it's not a AR uh history and what we thought or what we saw in the sanker sequencing that we had uh it's reflected also with the hypc target um so you can see that uh the Continental African species don't doesn't form a monoptic clate they are quite disperse and Madagascar the colonizations from the from Madagascar and back are constant and it will be really interesting to do the the biogeography studies of of this species but it's not done at the moment it will be a futter thing for me and now I'm including more more samples also um on the other hand we are we have preliminary results about the the the macro evolutive section of the project so this is a preliminary F including 49 uh representative of all the lineages that we wanted to Target and you can see that we have represented the three of the the the three so families of the 4B and um what was really interesting to see is that um most of the really small to medium gener so ten species or less have their their Crown before the the cooling of the pen paleogene and you will see why that is really interesting regarding the pattern that I talk before uh we we say that it's uh that that you can see a lower diversity in Continental Africa but we wanted to know how you achieve that how you come from a tropical area to the pattern that we have now and these are the preliminary results of the diversification analysis and um all the the upper ones are doing are done in Comet so compound poison process in test uh that is implemented in R and it's really quick to run if you want to give it a try and the lower ones are done in red base and they are quite similar but the ones in red base um include another dimension they uh also have uh into account the percentage of the percentage of um sampling that you have for each of the clayes that you are targeting so they they um are they are including the the the the bias that you have in your sampling and also the Computing process the Computing algorithm and how it computes uh if you compare both of them it's really different because um both models um do episodic uh dead see birth death um models that that means that um you have a time frame and you um you cut small windows in that time frame and you allow the rates to change in the windows but uh the the comet one and the red based want uh treat those rates and those windows differently so um for the for the compound one you will have exponential um changes in the rates if in all of the windows and um in the in the red red base one they will be linked so they will be free to change but they will correlate to the rates in the window before so uh it's more cohesive it's also longer to run um you can see in the first graph there that the net diversification rate so the difference between spaci and uh Extinction um it's a stable and then around 20 to 10 million years it goes really really low and then it will emerge again and that small window uh will mean a mass extinction but you will you would not would not know because maybe it's all only low divers low expiation so you have the second graph that tries to measure only extension uh if you have work in BIO geography you know that measuring extension is really hard but this this is like an appro this is an approach you you try to do that and you can see in the same in the same frame time frame you can see uh that the extension goes a little bit up but then you will have the third graph that measures the probability of a mass exension at that point and in all of the windows but you can see that the only one that that is significant so up uh 0.4 is the one that it's between uh 15 and five or 15 and 10 that is the the green uh line and the lower the lower graphs that are from RE Bas so a similar picture so you you can see in the first one the pool of the present and the the diversification the speciation goes really up but that's because um that's an artifact cost that we know the the the diversity now but we don't know the diversity that go that went extinct from later to now so it shows that we have a lot of diversity now but that's Al only an artifact but you can see the it's not so great with this color palette but you can see that around the same time the the speciation um is not going up so fast and it's the same in the orange ones with the diversification that it goes really really deep uh then the the blood the is the the confidence interval and the line is like the middle value and finally you have the extinction and in the same point both of the both of the analysis agree at the same time you have a really big on extinction rates um what does that mean um if you put that into context with with the p climate history that we know uh it shows that around uh 25 to 5 million years um you have a mass extinction happening or something happening and related to climate you can see that uh cooling went on in the paleogene uh and from the myosin you you can see the the the cooling effect and that cooling went uh uh was a company that an nidification in Africa and that aridification reduced the area of tropical forest and that gives a like a hope of testing our our hypothesis when we have a full data set because without a full data set this is only preliminary and maybe we don't see the same when we finish but now um we are more uh closer to accept the hypothesis of the paleoclimatic um consequences and also the aridification one working in in one uh both at once uh it's not included in the present but we are also doing um uh Niche um ecological niche reconstructions of all the gener and all the clades per species also and we are seeing also like movements within Africa that match the patterns of climate paleoclimate that we know and if you want to know more and to see further results because they will they will be quite different because now I have almost the full data set so it will be a great update from that uh I will be talking about that in detail in the IBC and finally um I'm here doing a research stay in my PhD I'm on my third year of fourth and here I came to work with Patricia and also with Alex and with ir and um I'm working mainly in the taxonomical review of the African Continental African species of croon uh so I had a preliminary philogyny and now I have an an extended one and I'm reviewing all of the materials that went into the philogyny and that were selected for the phogy even if we don't have a sequences and we are comparing them with other materials that we have and also working with all the floras all the descriptions all the information that we know to try to re identify them to update our philogen and to in a future um do a revision of all the Continental African species and that will take a lot of time and um Patricia Bara and Ricarda Rina are involved in that and I don't know what it will where when it will be published but um it's a really nice work to do and I'm loving it and uh also I'm here um um making uh digitalization of pictures more like of all the specimens that we are interested in and that we have samples of uh because in the future we want to include the characters the morphological aspect of the of the species um in our studies not only for the taxonomy but also to compare with the phog genomical data with the models that I have shown you and uh that will be an update of on on the study that they started in the Madagascar species of Croton and we want to include all the African ones and that's everything I don't know if it was too too quick or not because I didn't put an alarm um I'm sorry if my English was was a little bit rusted and this is like a pre pre rehearsal pre IC so if you have any questions I will try to answer them