From e55c55b72875bd835eb46be455d7705356c467eb Mon Sep 17 00:00:00 2001 From: yokobird <52844352+yokobird@users.noreply.github.com> Date: Sun, 31 Oct 2021 16:56:00 -0400 Subject: [PATCH 1/4] A Few changes to the SDM tutorial --- .../species-distribution-modeling/tutorial.md | 64 ++++++++++--------- 1 file changed, 33 insertions(+), 31 deletions(-) diff --git a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md index e1fdc52e5f666e..c0ba74cdd10245 100644 --- a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md +++ b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md @@ -2,11 +2,11 @@ layout: tutorial_hands_on title: "Species distribution modeling" questions: - - "How to build & evaluate a niche model, visualize and project it ?" + - "How to build & evaluate a niche model, visualize and project it?" objectives: - "Find and download occurrences data from GBIF" - "Find and download environmental data" - - "Process both occurrences and environmental" + - "Process both occurrences and environmental data" - "Partition occurrence data" - "Model a theoretical ecological niche and predict species distribution in a future climate scenario by using SDM" requirements: @@ -25,9 +25,9 @@ contributors: # Introduction {:.no_toc} -Species Distribution Modeling (SDM) can help understand the distribution of a species depending on its environment. It can also attempt to quantify the impact of climate change on the species habitat, direct conservation efforts and predict invasive species distributions. This is done by associating data of species occurrences (observations) with a set of environmental data (such as temperature and precipitation). +Species Distribution Modeling (SDM) can help understand a species distribution in relation to its environment. It can also attempt to quantify the impact of climate change on the species' habitat, direct conservation efforts and predict invasive species distributions. This is acheived by associating species occurrence (observations) data with a set of environmental data (such as temperature and precipitation). -The goal of this tutorial is to model a theoretical ecological niche and predict species distribution in a future climate scenario by using SDM with the Wallace interactive environment on Galaxy. We'll use the data occurrences of US *Chrysemys picta* ([Painted turtle](https://en.wikipedia.org/wiki/Painted_turtle)) from the North America region. +The goal of this tutorial is to model a theoretical ecological niche and predict species distribution in a future climate scenario by using an SDM with the Wallace interactive environment on Galaxy. We'll use the data occurrences of *Chrysemys picta* ([Painted turtle](https://en.wikipedia.org/wiki/Painted_turtle)) in the US mainland North America region. > ### Agenda > @@ -41,7 +41,7 @@ The goal of this tutorial is to model a theoretical ecological niche and predict # Step 1: Import data from public databank -In this study the datasets are all imported from the [GBIF](https://www.gbif.org/) databanks. It is also possible to import from [bison](https://www.gbif.org/), [iNaturalist](https://www.inaturalist.org/) and others. +In this study the datasets are all imported from the [GBIF](https://www.gbif.org/) databanks. It is also possible to import from [BISON](https://bison.usgs.gov/), [iNaturalist](https://www.inaturalist.org/) and others. > ### {% icon hands_on %} Hands-on: Import the data @@ -51,30 +51,30 @@ In this study the datasets are all imported from the [GBIF](https://www.gbif.org > {% snippet faqs/galaxy/histories_rename.md %} > > 2. **Get species occurrences data** {% icon tool %} with the following parameters -> - *"Scientific name"*: `chrysemys picta` +> - *"Scientific name"*: `Chrysemys picta` > - *"Data source"*: `gbif` > - *"Number of records to return"*: `10000` > > 3. Inspect the generated file with the Scratchbook {: .hands_on} -We have now a tabular file with about 10,001 lines and many columns. +We now have a tabular file with 10,001 lines and many columns. > ### {% icon question %} Questions -> 1. What do the lines in the files represent? -> 2. What are the columns 1, 2, 3 and 51 of the generated file? +> 1. What do the rows in the files represent? +> 2. What are columns 1, 2, 3 and 54 of the generated file? > > > ### {% icon solution %} Solution -> > 1. Each line represent an observation of Chrysemys picta, with its location, and some other metadata -> > 2. Column 1 is the name, column 2 the longitude of the observation of Chrysemys picta, column 3 the latitude and column 50 the country code of the location +> > 1. Each row represents a *Chrysemys picta* observation, the location, and other metadata +> > 2. Column 1 is the name, column 2 and 3 the observation's longitude and latitude respectively, and column 54 the observation's country code > {: .solution } {: .question } -We would to extraction the 4 columns previously cited and keep only occurrence records from US. +We will extract the 4 columns mentioned above and keep only the occurrence records from the US. > ### {% icon hands_on %} Hands-on: Import the data > 1. **Cut columns** {% icon tool %} with the following parameters -> - *"Cut columns"*: `c1,c2,c3,c51` +> - *"Cut columns"*: `c1,c2,c3,c54` > - *"Delimited by"*: `Tab` > - {% icon param-file %} *"From"*: output of **Get species occurrences data** > @@ -87,7 +87,7 @@ We would to extraction the 4 columns previously cited and keep only occurrence r > > > ### {% icon question %} Question > > -> > How many occurrences have been conserved? +> > How many occurrences were kept? > > > > > ### {% icon solution %} Solution > > > 64.55% of the occurrences (6,456) are conserved @@ -95,26 +95,27 @@ We would to extraction the 4 columns previously cited and keep only occurrence r > {: .question} > > 4. **Tabular to CSV** {% icon tool %} +> - You can find the option to convert > - {% icon param-file %} *"tabular file"*: output of **Filter** > - *"output csv Separator"*: `,` > - *"Header in file"*: `Yes` -> 5. Due to an issue related to the actual tabular to csv converter application to GBIF format, each `"` sign is triplicated on the resulting file. You need to replace `"""` by `"` using for example the **Replace** {% icon tool %} parts of text tool. +> 5. Due to an issue related to the GBIF format, each `"` sign is triplicated on the resulting file. You need to replace `"""` for `"`, using for example the **Replace** {% icon tool %} parts of text tool. {: .hands_on} -# Step 2: Development and evaluation of SDM using Wallace +# Step 2: Development and evaluation of the SDM using Wallace -[Wallace](https://wallaceecomod.github.io/) is a R Shiny app integrated into Galaxy. It provides an interactive environment for the rapid and effective development and evaluation of SDM, including data download, cleaning, partitioning, modeling, visualisation and predictions. +[Wallace](https://wallaceecomod.github.io/) is an R Shiny app integrated into Galaxy. It provides an interactive environment for the rapid and effective development and evaluation of an SDM, including data download, cleaning, partitioning, modeling, visualisation and prediction. ## Transmit occurrence data from Galaxy to Wallace -With this you can either upload file you've loaded earlier from Galaxy data or you can download data directly from Wallace. Let's use the data from your Galaxy history: +Here, you can either upload your Galaxy data or you can download data directly in Wallace. Let's use the data from your Galaxy history: > ### {% icon hands_on %} Hands-on: Launch Wallace -> 1. Verify your current history is the history containing previous filtered occurences data. -> 2. Open {% tool [Wallace](interactive_tool_wallace) %} in Galaxy or click here to [open the tool](https://ecology.usegalaxy.eu/root?tool_id=interactive_tool_wallace) and click on execute +> 1. Verify your current history has the filtered occurences data. +> 2. Open {% tool [Wallace](interactive_tool_wallace) %} in Galaxy or click here to [open the tool](https://ecology.usegalaxy.eu/root?tool_id=interactive_tool_wallace) and click on execute. > 3. Go to User > Active InteractiveTools, then click on the active InteractiveTool named `Wallace visualisation` > 4. In Wallace -> 0. Due to an issue related to last wallace Docker update, a manual fix is proposed waiting for a better solution. You first have to go to the **5 Partition Occs**, selecting *"Spatial Partition"* then for *"Options available"* selecting `Checkerboard 2 (k=4)` then click on the **Partition** button. This will fix the issue so now you have the *"Aggregation Factor"* displayed. Sorry for the inconvenience. +> 0. Due to an issue related to last the Wallace Docker update, a manual fix is required. First go to the **5 Partition Occs**, select *"Spatial Partition"* then for *"Options available"* select `Checkerboard 2 (k=4)` then click on the **Partition** button. This will fix the issue so now you have the *"Aggregation Factor"* displayed. Sorry for the inconvenience. > 1. Go to **1 Occ Data** > 2. In **Obtain Occurrence Data** > - *"Modules Available"*: `Galaxy History User` @@ -122,7 +123,7 @@ With this you can either upload file you've loaded earlier from Galaxy data or y > - Click on **Load Occurrences** {: .hands_on} -In the main panel of Wallace, the data will load. They can now be inspected using the map, the table, etc +In the main panel of Wallace, the data will load. It can now be inspected using the map, the table, etc. You now have your occurrence records on Wallace! @@ -139,13 +140,13 @@ You now have your occurrence records on Wallace! ## Process occurrence data -For the modelling, we need to the choose the occurrences we want to use. There is 4 ways to do it in Wallace in the **2 Process Occ** panel: +For the modeling, we need to the choose the occurrences we want to use. There is 4 ways to do this in Wallace in the **2 Process Occ** panel: 1. Selecting occurrences on the map by delimiting an interesting geographic area 2. Removing occurrences by ID -3. Delimiting a spacial thin +3. Delimiting a spatial thin - This allow to select occurrences by setting a minimum distance (in km) between the different occurrences. For example, by typing 30km, we will get all the occurrences on the map which are at minimum 30km from each other. + This allows to select occurrences by setting a minimum distance (in km) between the different occurrences. For example, by typing 30km, we will get all the occurrences on the map which are at minimum 30km from each other. 4. Considering all the occurrences @@ -179,6 +180,7 @@ This module provides a raster with environmental variables from online sources. > > 2. In **Obtain Environmental Data** > > - *"Modules Available"*: `Galaxy History User` > > - *"Select from your Galaxy History User csv file"*: imported GeoTIFF file +> > Note that if you imput your own raster, you don't have the option for time prediction later on. > {: .comment} > > - *"Select WorldClim bioclimatic variable resolution"*: `10 arcmin` @@ -236,7 +238,7 @@ There is 2 possible way to partition data: ![Checkerboard 2](../../images/species-distribution-modeling/Checkerboard.png "Example of checkerboard (k=4) with a factor 4. Source: doi:10.0.4.87/2041-210X.12261") -For both of these techniques the number of occurrences into each bin may vary. +For both of these techniques the number of occurrences in each bin may vary. > ### {% icon hands_on %} Hands-on: Partition Occurrence Data > 1. Go to **5 Partition Occs** @@ -304,7 +306,7 @@ To visualize the suitability on the map, we can use the map prediction module. ## Project Model -Wallace can use the trained model to predict possible species distributions in a different area, outside of the sampled one. Here, we will try it for the Canada. +Wallace can use the trained model to predict possible species distributions in a different area, outside of the sampled one. Here, we will try it for Canada. > ### {% icon hands_on %} Hands-on: Visualize Model Results > 1. Go to **8 Project** @@ -318,9 +320,9 @@ Wallace can use the trained model to predict possible species distributions in a > 4. Click on **Project** {: .hands_on} -We can also predict not only for a different area, but also different timing and climate. The [Global Circulation Model (GCM)](https://en.wikipedia.org/wiki/General_circulation_model#Atmospheric_and_oceanic_models) are used here to predict atmospheric fluctuation and then study climate change. Each model is different and use parameters like ocean atmosphere and others. The prediction need also a RCP scenario, a scenarios about the amount of greenhouse gases emitted in the near future. This scenario may have different predicted presence models. +We can predict not only for a different area, but also for a different time and climate. We use the [Global Circulation Models (GCM)](https://en.wikipedia.org/wiki/General_circulation_model#Atmospheric_and_oceanic_models) to predict atmospheric fluctuation and then study climate change. Each model is different and uses several parameters such as ocean atmosphere. The prediction also uses RCP scenarios, Representative Concentration Pathways take average enegy requirements across the planet (2.6, 4.5, 6.0 and 8.5 watts per meter squared) to model the concentration of greenhouse gases emitted in the near future and their effect on climate. These scenarios may produce varying predicted presence models. -We would like to predict the presence of Chrysemys Picta in Canada in 2050, given the model [CCSM4](http://www.cesm.ucar.edu/models/ccsm4.0/), a US model based on earth circulation, and a RCP value of 2.6 +We would like to predict the presence of *Chrysemys Picta* in Canada in 2050, given the model [CCSM4](http://www.cesm.ucar.edu/models/ccsm4.0/), a US model based on earth circulation, and an RCP value of 2.6 > ### {% icon hands_on %} Hands-on: Visualize Model Results with a 2.6 RCP > 2. In **Project Model** @@ -334,7 +336,7 @@ We would like to predict the presence of Chrysemys Picta in Canada in 2050, give ![CCM4 with a 2.6 RCP](../../images/species-distribution-modeling/GCM_CCSM4_RCP_2.6.png) -We would like now to compare the prediction for RCP value of 8.5 . +Let's compare the prediction for the RCP value of 8.5. RCP 8.5 describes a pathway for the current carbon emission without additional constraints, it projects a 4.3˚C temperature increase by 2100. > ### {% icon hands_on %} Hands-on: Visualize Model Results with a 8.5 RCP > 2. In **Project Model** @@ -355,7 +357,7 @@ We would like now to compare the prediction for RCP value of 8.5 @@ -165,7 +166,7 @@ We want to work on the data from the US, we will select the occurrences with the ## Obtain Environmental Data -We would like now to provide environmental variables, that are considered relevant for biological purposes. The [Bioclimatic variables](http://www.worldclim.org/bioclim) consist of such original and derived variables. We will associate them with the occurrence data, using the WorldClim Bioclims module. +Now, we would like to obtain biologially relevant environmental data for our model. The [Bioclimatic variables](http://www.worldclim.org/bioclim) consist of such original and derived variables. We will associate them with the occurrence data, using the WorldClim Bioclims module. This module provides a raster with environmental variables from online sources. The raster is composed of environmental information. Each layer of the raster contains a climatic variable: starting from BIO1 = Annual mean temperature, to BIO19 = Precipitation of Coldest Quarter. @@ -198,12 +199,12 @@ Wallace will now associate environmental data and occurrences data to train a mo 1. It first creates a buffer zone around the occurrences - We can choose the size of the buffer zone to control the area we will be working with and on which a map of suitability will be made. + We can choose the size of the buffer zone to control the area we will work with and on which a map of suitability will be made. - Several background extent can be used: - - Bounding box will define an area where the occurrences centered - - Minimum convex polygon will make an area considering the repartition of the occurrences - - Point buffers will use occurrences localities to build a buffer zone around each occurrence + Several background extents can be used: + - Bounding box defines an area centered on the occurrences + - Minimum convex polygon creates a convex area around the occurrences placement + - Point buffers will use occurrence localities to build a buffer zone around each occurrence 2. It samples Background Points @@ -224,17 +225,17 @@ Wallace will now associate environmental data and occurrences data to train a mo ## Partition Occurrence Data -By partitioning data, a dataset is divided into subsets (i.e. bins). A model is then built on each of subsets but one and test it on the last one (assuming that all the groups are independent). +By partitioning data, a dataset is divided into subsets (i.e. bins). A model is built on each of the subsets but one, it is used to test the model (assuming that all the groups are independent). -There is 2 possible way to partition data: +There is 2 possible ways to partition data: 1. **Non-spatial Partition**: a partition used when there is no bias due to space, time or sampling method - 1. Jakknife (k=n): each occurrence in the dataset is considered as equal to a bin. This is usually used when you have a small dataset with no known bias + 1. Jacknife (k=n): each occurrence in the dataset is considered as equal to a bin. This is usually used when you have a small dataset with no known bias 2. Random k-fold: partition the data randomly in a number of bins set by the user with the option `Number of Folds` 2. **Spatial Partition**: a partition used when there could be bias due to time, space or sampling method 1. Block (k=4): the area is divided in four and the different occurrences are put equally into four bins - 2. Checkerboard 1 (k=2): two bins are used according to the position of the occurrence on the grid - 3. Checkerboard 2 (k=4): four bins are used according to the position of the occurrence on the grid. This require an aggregation factor, which is the size of a second grid put on a first one. For example, with a factor 4, the grids size will be 4x4 + 2. Checkerboard 1 (k=2): two bins are used according to the position of the occurrences on the grid + 3. Checkerboard 2 (k=4): four bins are used according to the position of the occurrences on the grid. This requires an aggregation factor, which is the size of a second grid placed on the first one. For example, with a factor 4, the grid's size will be 4x4 ![Checkerboard 2](../../images/species-distribution-modeling/Checkerboard.png "Example of checkerboard (k=4) with a factor 4. Source: doi:10.0.4.87/2041-210X.12261") @@ -258,7 +259,7 @@ Wallace can build different models using either: 1. The presence-only approach BIOCLIM (Module BIOCLIM) ({% cite booth2018 %}) 2. The presence-background (presence-pseudo absence) algorithm Maxent (Module Maxent) -To evaluate these models, Wallace computes the performance on a hold-out dataset (data not used for training) and provide evaluation metrics as the AUC (Area Under the Curve) mean. As a rule of thumb, an AUC of 0.75 and above is considered good, and closer to 1 is better. +To evaluate these models, Wallace computes the performance on a hold-out dataset (data not used for training) and provides evaluation metrics such as the AUC (Area Under the Curve) mean. As a rule of thumb, an AUC of 0.75 and above is considered good, and closer to 1 is better. > ### {% icon hands_on %} Hands-on: Build and evaluate the niche model > 1. Go to **6 Model** @@ -271,7 +272,7 @@ To evaluate these models, Wallace computes the performance on a hold-out dataset ## Visualize Model Results -We would like now to visualize the model of the theoretical niche. First we will make a chart to simulate an ecological niche: +We would now like to visualize the model of the theoretical niche. First we will make a chart to simulate an ecological niche: - X-axis: temperature - Y-axis: the annual precipitation @@ -286,7 +287,7 @@ We would like now to visualize the model of the theoretical niche. First we will ![temperature vs annual precipitation](../../images/species-distribution-modeling/bioclim_enveloppe_plots.png) -The optimum environmental parameters for this species (represented in green) is between 5°C and 15°C (on the graph, values are x10) with an annual precipitation between approximately 700mm and 1250mm. +The optimum environmental parameters for this species (represented in green) are between 5°C and 15°C (on the graph, values are x10) with an annual precipitation between approximately 700mm and 1250mm. To visualize the suitability on the map, we can use the map prediction module. @@ -306,7 +307,7 @@ To visualize the suitability on the map, we can use the map prediction module. ## Project Model -Wallace can use the trained model to predict possible species distributions in a different area, outside of the sampled one. Here, we will try it for Canada. +Wallace can use the trained model to predict possible species distributions in a different areas, outside of the sampled one. Here, we will try it for Canada. > ### {% icon hands_on %} Hands-on: Visualize Model Results > 1. Go to **8 Project** @@ -320,7 +321,7 @@ Wallace can use the trained model to predict possible species distributions in a > 4. Click on **Project** {: .hands_on} -We can predict not only for a different area, but also for a different time and climate. We use the [Global Circulation Models (GCM)](https://en.wikipedia.org/wiki/General_circulation_model#Atmospheric_and_oceanic_models) to predict atmospheric fluctuation and then study climate change. Each model is different and uses several parameters such as ocean atmosphere. The prediction also uses RCP scenarios, Representative Concentration Pathways take average enegy requirements across the planet (2.6, 4.5, 6.0 and 8.5 watts per meter squared) to model the concentration of greenhouse gases emitted in the near future and their effect on climate. These scenarios may produce varying predicted presence models. +We can predict not only for a different area, but also for a different time and climate. We use the [Global Circulation Models (GCM)](https://en.wikipedia.org/wiki/General_circulation_model#Atmospheric_and_oceanic_models) to predict atmospheric fluctuation and then evaluate climate change. Each model is different and uses several parameters such as ocean atmosphere. The prediction also uses RCP scenarios, Representative Concentration Pathways take average enegy requirements across the planet (2.6, 4.5, 6.0 and 8.5 watts per meter squared) to model the concentration of greenhouse gases emitted in the near future and their effect on climate. These scenarios may produce varying predicted presence models. We would like to predict the presence of *Chrysemys Picta* in Canada in 2050, given the model [CCSM4](http://www.cesm.ucar.edu/models/ccsm4.0/), a US model based on earth circulation, and an RCP value of 2.6 @@ -357,7 +358,7 @@ Let's compare the prediction for the RCP value of 8.5. RCP 8.5 describes a pathw # Conclusion -Following this tutorial, we have loaded a dataset of occurrences to the shiny app Wallace and model the distribution of *Chrysemys picta* (Painted turtle) with the Species Distribution Modeling (SDM) method. This allowed us to visualize it’s ecological niche and how climate change can influence it’s future distribution in North America. The saved project can be used for similar future studies. +Following this tutorial, we have loaded a dataset of occurrences to the shiny app Wallace and modeled the distribution of *Chrysemys picta* (Painted turtle) with the Species Distribution Modeling (SDM) method. This allowed us to visualize it’s ecological niche and how climate change can influence it’s future distribution in North America. The saved project can be used for similar future studies. > ### {% icon details %} Some useful references > From c6e5d569fa5272f42a6cda32544e74c04ef159f7 Mon Sep 17 00:00:00 2001 From: yokobird <52844352+yokobird@users.noreply.github.com> Date: Sun, 31 Oct 2021 19:21:05 -0400 Subject: [PATCH 3/4] Some changes to the tutorial --- .../species-distribution-modeling/tutorial.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md index d1963ab63f567c..0948c0f8888f33 100644 --- a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md +++ b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md @@ -112,7 +112,7 @@ We will extract the 4 columns mentioned above and keep only the occurrence recor Here, you can either upload your Galaxy data or you can download data directly in Wallace. Let's use the data from your Galaxy history: > ### {% icon hands_on %} Hands-on: Launch Wallace -> 1. Verify your current history has the filtered occurences data. +> 1. Verify your current history has the filtered occurrences data. > 2. Open {% tool [Wallace](interactive_tool_wallace) %} in Galaxy or click here to [open the tool](https://ecology.usegalaxy.eu/root?tool_id=interactive_tool_wallace) and click on execute. > 3. Go to User > Active InteractiveTools, then click on the active InteractiveTool named `Wallace visualisation`. Wallace will open up in your browser. > 4. In Wallace @@ -141,13 +141,13 @@ You now have your occurrence records on Wallace! ## Process occurrence data -For the modeling, we need to the choose the occurrences we want to use. There is 4 ways to do this in Wallace in the **2 Process Occ** panel: +For the modeling, we need to choose the occurrences we want to use. There is 4 ways to do this in Wallace in the **2 Process Occ** panel: 1. Selecting occurrences on the map by delimiting an interesting geographic area 2. Removing occurrences by ID 3. Delimiting a spatial thin - This allows to select occurrences by setting a minimum distance (in km) between the different occurrences. For example, by typing 30km, we will get all the occurrences on the map which are at minimum 30km from each other. + This allows selecting occurrences by setting a minimum distance (in km) between the different occurrences. For example, by typing 30km, we will get all the occurrences on the map which are at a minimum of 30km from each other. 4. Considering all the occurrences @@ -166,7 +166,7 @@ We want to work on the data from the US, we will select the occurrences with the ## Obtain Environmental Data -Now, we would like to obtain biologially relevant environmental data for our model. The [Bioclimatic variables](http://www.worldclim.org/bioclim) consist of such original and derived variables. We will associate them with the occurrence data, using the WorldClim Bioclims module. +Now, we would like to obtain biologically relevant environmental data for our model. The [Bioclimatic variables](http://www.worldclim.org/bioclim) consist of such original and derived variables. We will associate them with the occurrence data, using the WorldClim Bioclims module. This module provides a raster with environmental variables from online sources. The raster is composed of environmental information. Each layer of the raster contains a climatic variable: starting from BIO1 = Annual mean temperature, to BIO19 = Precipitation of Coldest Quarter. @@ -181,7 +181,7 @@ This module provides a raster with environmental variables from online sources. > > 2. In **Obtain Environmental Data** > > - *"Modules Available"*: `Galaxy History User` > > - *"Select from your Galaxy History User csv file"*: imported GeoTIFF file -> > Note that if you imput your own raster, you don't have the option for time prediction later on. +> > Note that if you input your own raster, you don't have the option for time prediction later on. > {: .comment} > > - *"Select WorldClim bioclimatic variable resolution"*: `10 arcmin` @@ -366,6 +366,6 @@ Following this tutorial, we have loaded a dataset of occurrences to the shiny ap > - BIOCLIM: the first species distribution modelling package, its early applications and relevance to most current MAXENT studies. Diversity and Distributions ({% cite booth2018 %}). > - ENMeval: An R package for conducting spatially independent evaluations and estimating optimal model complexity for MAXENT ecological niche models. ({% cite muscarella2014 %}). > -> [Here](http://www.ipcc-data.org/guidelines/pages/gcm_guide.html) for informations on Global Circulation Model(GCM), how it's done what is taken in consideration and more. +> [Here](http://www.ipcc-data.org/guidelines/pages/gcm_guide.html) for informations on Global Circulation Model(GCM), how it's done, what is taken in consideration and more. {: .details} From 6111c3215c559e1d9d2b92b3cebf922bf5ff8838 Mon Sep 17 00:00:00 2001 From: Saskia Hiltemann Date: Thu, 10 Nov 2022 17:40:05 +0100 Subject: [PATCH 4/4] fix icon tag --- .../ecology/tutorials/species-distribution-modeling/tutorial.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md index dca28f3c6f6aaa..7fe2269645122d 100644 --- a/topics/ecology/tutorials/species-distribution-modeling/tutorial.md +++ b/topics/ecology/tutorials/species-distribution-modeling/tutorial.md @@ -99,7 +99,7 @@ We will extract the 4 columns mentioned above and keep only the occurrence recor > > 4. **Tabular to CSV** {% icon tool %} > - Click on the **Edit Attributes** {% icon galaxy-pencil %} option for the output of **Filter**. -> - On the central panel click on the {% galaxy-gear %} **Convert** tab. +> - On the central panel click on the {% icon galaxy-gear %} **Convert** tab. > - Check the **Target data** is csv. > - Click on **Create Dataset** >