Home

Training manual on spatial analysis of plant diversity and distribution

1. Auto complete Edit values v oi Select color scheme 10 15 Red Green Blue 15 20 20 25 gt 24 Add or A R Nodata No Data welellindso iat CI NoData Transparent If you have followed the procedure correctly the map displayed should be similar to the one below Chapter 3 T2 The Properties button also includes additional tabs for further information Under the History tab you will find the information on how the layer was generated which can be useful in the case of an unexpected outcome or in order to find details on how a previous analysis was carried out while the nfo tab offers relevant information on the raster including resolution and maximum minimum values a Properties Label mean temperature latin america 10 min Label mean temperature latin america 10 min Filename amp gis tutorial 3 1 basic elements mean temperature latin america 10 Filename amp gis tutorial 3 1 basic elements mean temperature latin america 10 e Legend Info History Legend gfo History EA Version Date Creator DIVA GIS oe 20090716 Columns 528 Min Value 6 4625 Procedure Climate data to map ee s Rows 533 Max Value 29 170834 Climate frmClim2Grid er F Variable TMEAN e Data Type FLT4S Units degrees C Period 0 aS x Y Min 117 333339 56 16667 Max 29 333335 32 666668 Cell size 0 166667 0 166667 Projection Map units Datum 13 14
2. Chapter 5 The concept of circular neighbourhood The following diagrams represent the concept of circular neighbourhood where the circles of different colours indicate different species richness green indicates a richness value of one yellow of two and red of three species When using this method the results around the margins of the point distribution can be distorted by assigning lower diversity at borders in the case of incomplete sampling e g at country borders Richness in a one 1 degree cell Higher diversity is clearly observed in the lower left cell however the raster has low resolution The circular neighbourhood consists of using a fixed diameter circle centred on each raster cell to assign a value to the cell in the case of the top left hand corner cell the value is one 1 since there is only one species inside the circle shown As shown each observation contributes to the value of a number of different cells Richness of 10 minute cells In this case the raster has more detail high resolution but the pattern of diversity has been lost RRARKAAN ff YY HT AN YW VAAN ZY NADP BSD AA NSSR eBlog en The circular neighbourhood repeats itself for each cell In this example the circle has a diameter of one 1 degree the same size as the cell in the first example Richness with 10 minute cell and circular neighbourhood of one 1 degr
3. Austrian le Training Manual on Spatial Analysis Bioversity International J INIA Instituto Nacional de Investigaci n y Tecnolog a Agraria y Alimentaria Development Cooperation so as atten ilies annaran a 9 20 21 22 23 24 25 26 27 28 29 30 185 156 163 185 147 147 156 160 164 162 156 164 LMCH106 of Plant Diversity and Distribution Xavier Scheldeman and Maarten van Zonneveld LMCH91 LMCH102 _ _ l a _ n ie 181 ag e237 237 494 237 ww 194 194 194 LMCH139 LMCH144 i 312 305 299 299 299 304 176 Bioversity International is part of the Consultative Group on International Agricultural Research which works to reduce hunger poverty and environmental degradation in developing countries by generating and sharing relevant agricultural Knowledge technologies and policies This research focused on development is conducted by a Consortium of 15 CGIAR centres working with hundreds of partners worldwide and supported by a multi donor Fund The international status of Bioversity is conferred under an Establishment Agreement which by December 2009 had been signed by the Governments of Algeria Australia Belgium Benin Bolivia Brazil Burkina Faso Burundi Cameroon Chile China Congo Costa Rica C te d lvoire Cyprus Cuba Czech Republic Denmark Ecuador Egypt Ethiopia Ghana Greece Guinea Hungary India Indonesia Iran Israel Italy J
4. V _parviflora bio_17 Continuous V _pulchra bio_18 Continuous V _quercifolia bio_19 Continuous Select all Deselect all Select all Deselect all V Linear features Create response curves _ Make pictures of predictions Quadratic features Do jackknife to measure variable importance Product features Hinge features Threshold features Output format Logistic Output file type Output directory E GIS Tutoriali6 4 Gap analysis Browse Auto features Projection layers directory file Browse To generate binary rasters of potential distribution areas using Maxent a similar procedure as was conducted in Step 2 of Analysis 6 3 1 is followed in the Advanced settings window under the Apply threshold rule box select the 10 percentile training presence option Maximum Entropy Parameters Advanced Experimental Add samples to background Add all samples to background Write plot data Extrapolate Do clamping Write output grids Write plots Append summary results to maxentResults csv file NONN N NODON Cache ascii files Maximum iterations 500 Convergence threshold 0 00001 Adjust sample radius 0 Log file maxent log Default prevalence 0 5 Apply threshold rule 10 percentile training pres
5. Define Grid Use parameters from another grid Options 21 Output Yarable Richness l al 22 Number of different classes Richness Pa Paint to Grid Procedure Circular Neighborhood Options a Neighborhood Options Diameter 1 Map Units Output J eee AA Apply hl Close SOK X Cancel Chapter 5 Using the Circular Neighborhood option gives you more precise patterns for the existing diversity along with a relatively high resolution Individual Task Observe what happens if you apply a larger circular neighbourhood e g 5 Map Units Spatial analysis of diversity for conservation planning 5 2 Intra specific diversity analysis based on phenotypic data Genetic diversity studies including the analysis of spatial patterns in genetic diversity are frequently based on molecular marker data see Section 5 3 However phenotypic data and particularly morphological data can be another indirect source of genetic diversity information Phenotypic data from a single individual varies as a function either of the genotype G the environment E or a combination of both the GxE effect Some traits like flower colour are not influenced by the environment When using data based on in situ characterization conducted at environmentally heterogeneous locations it is recommended to focus on these traits When working on other traits in order to minimize variation determin
6. 5 c After entering the first value the cells as well as their labels in the From column should change automatically for the next row you only need to enter values in the To column Basic elements of spatial analysis in DIVA GIS Label mean temperature latin america 10 min Label mean temperature latin america 10 min Filename amp gis tutorial 3 1 basic elements mean temperature latin america 10 Filename amp gis tutorial3 1 basic elements mean temperature latin america 10 Legend Info Histor Auto complete eo Color From D Auto complete Edit values Os fe 10 5 Edit values Select rows Pr pet Select rows Classify i t eee Classify eae i a 7 x Select color scheme m Select color scheme 10 15 ham E eas Ram D 15 0 22 0 20 20 25 Read From File Read From File 22 0 30 0 25 25 30 Nodata No Data Add or Remove Row Nodata No Data Add or Remove Row De CJ 8 0 15 0 NoData Transparent NoData Transparent ers It is also possible to change the values in the From and To columns manually To do this select the Manual option in the first box cf 6a This will allow you to change the values in these columns By choosing this option there is the risk that some values might be forgotten and will therefore not be displayed on the map Note 7 The colours of each temperature range can be changed by using the Select Color Scheme command and se
7. DIVA GIS allows you to import and export rasters using files compatible with different programmes In order to use the generated climate layers for species distribution modelling in Maxent you must convert these to ASCII files asc by using the Data Export Gridfiles Multiple Files option Les DIVA GIS 7 3 0 Project MBEem Layer Map Analysis Modeling Grid Stack Tools Help Import Points to Shapefile gt loal OF re ii To Pg kJ z c ae Import Text to Line Polygon Draw Shape Polygon to Grid w oom 3 5 Selection to New Shapefile Climate _ s ma BEHORE Ga Assign Coordinates Check Coordinates i Export Gridfile Single File E Importto Gridfile Multiple Files Write YRT file worl Export Shapefile fi File Manager Download Select the raster files grd you wish to convert to ASCII asc using the Add File button To export grd raster files as ASCII files asc select the option ESRI ASCII When exporting grd raster files as ASCII files asc you can choose to format the file names by selecting the Make valid ESRI grid names option For this analysis however do not select this option Select Output Folder and save the newly formatted rasters in the same folder as rasters in the original file type Same as Input or in another folder Select Chapter 3 12 Click Apply to start the proces
8. Discussion Group There is a google discussion group for users of this software at http2 groups google com group Maxent Please sign up for this group for discussions and questions about the software and also to get 4 Maxent programme files can be downloaded separately or together in a compressed file format maxent zip Maxent download page Windows Internet Explorer EJ yy jen princeton edu File Edit View Favorites Tools Help jy Favorites Maxent download page oY gt mp v Pager Safety Toos y Thank you for registering To complete the download process you simply need to copy the following three files to a convenient location on your computer and follow the instructions in the readme file Alternatively you can download and unzip this one zip file which includes all three of the files below Important note Some browsers especially Internet Explorer may change the name of maxent jar to maxent zip during downloading To avoid this you can specify the full name maxent jar when prompted for a name or you can manually change the name back to maxent jar after downloading has completed Here is the jar file maxent jar Here is the bat file maxent bat Here is the readme file readme txt 5 Save the files in a convenient place on your hard disk To start Maxent double click on the MS DOS batch file bat If you have Winrar data compression software installed on your computer it m
9. are available for identifying such points see Chapman 2005a This section outlines two methods included in DIVA GIS to detect outliers Reverse jackknife Chapman 2005b This method is recommended for datasets with a normal distribution of values such as those with many observations for each taxon 1 5 x interquartile range 1 5 IQR DIVA GIS 2005 This method is recommended for datasets with a limited number of observations per taxon e g n lt 20 Although these methods are convenient for detecting outliers they do not guarantee that all detected outliers are in fact errors some outliers may also be valid points One important detail to consider before deciding to remove the point s is the purpose of the study If the purpose is to model species distribution in order to identify suitable production areas then using only core records may be preferable and there is no need to include the outliers in the analysis However if the purpose of the study is to identify ecotypes that may be adapted to more extreme conditions atypical points are of interest to the study and it may be useful to keep outliers included in the analysis As usual a record of all modifications made to the original file or dataset should be maintained Chapman 2005a Chapter 4 PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 4 2 Quality Control Atypical points e Excel e Vcundinamarcensis_outliers
10. gis tutorial 6 2 potential distribution pinus_kesiya grd Legend Info History Legend nfo History Color From Auto complete Color From To Auto complete 0 0000 0 1935 Edit values 0 1 Edit values t t 0 1935 0 387 0 1935 0 3870 aiii i l 0 1 0 2 Aoii Classify Classify 0 387 0 5805 0 3870 0 5805 al 0 2 0 3 0 5805 0 774 0 5805 0 7740 l i 0 3 0 4 Select color scheme Select color scheme 0 774 1 0 7740 1 0000 0 4 0 5 Nodata No Data Ramp l O 0 5 0 6 Ka Ramp l 2 _ _ l 0 6 0 7 Read From File a Read From File 0 7 0 8 Add or Remove Row 08 09 Add or Remove Row 09 1 ool Nodata No Data x C NoData Transparent NoData Transparent Visualizing the threshold limiting potential distribution areas 20 The 10 percentile training presence threshold found in the table of thresholds generated by Maxent will be used in this analysis as explained in Step 11 The threshold value for P kesiya is 0 154 as per the table of thresholds Go to the legend of the raster and create a new class ranging from zero 0 to the threshold value select a neutral colour for this new class Properties Label Pinus_kesiya Filename gis tutorial 6 2 potential distribution pinus_kesiya grd Legend Info History Color From To Label Auto complete v e 0 0 154 Edit values 0
11. E SSR5 156 gri SSR5 160 grd E SSR5 160 gri F SSR5 163 9rd E SSR5 163 gri F SSR5 164 grd A SSR5 164 gri SSR5 166 ard E SSR5 166 gri F SSR5 192 9rd E SSR5 182 gri F SSR5 194 9rd fs SSR5 184 gri F SSR6 134 grd E SSRE 134 gri 2 SSR6 142 grd E SSRE 142 gri SSR6 146 grd E SSRE 146 gri SSR6 148 grd E SSRE 148 gri SSR6 150 grd E SSRE 150 gri SSR 177 ard E SSRT 177 ori 2 SSR7 183 9rd fe SSRT 183 gr 2 SSR7 189 grd E SSRT 189 ori SSR7 191 9rd E SSR7 191 ori F SSR7 193 9rd E SSR7 193 9ri SSR7 206 ard A SSR7 206 gri SSR7 208 ard E SSRT 208 gri SSR8 299 grd a SSR8 299 ori 2 SSR8 301 grd a SSR8 301 ori SSR8 3065 grd fs SSR8 305 gri 2 SSR8 316 9rd a SSRO 31 B gri 55R8 318 grd 3 s8R8 318 gri A rare allele SSR3 300 Chapter 5 2 Before starting the cluster analysis make a stack of presence absence rasters for the 51 alleles generated in the previous step See Analysis 5 2 1 Step 8 This file corresponds to a stack group of rasters with the same properties and will be the basis for the following analysis 3 Goto Stack Cluster and select the stack that you just created Project Data Layer Map Analysis Modeling Grid acm Tools Help Oee MQ AAS AIMKES mse leoz2h500 SHo4ss s 7 Plot Latin America Countries Cacuia alculate a Regression SSR cherimoya rand column Cluster 7 SSR3 300 Export to Text File Clo Overlap ie C
12. Provides quick access to most commonly used tasks 3 Legend Lists all layers of the current map The selected map is highlighted Layers can be visualized or hidden by checking or un checking the boxes in front of each layer name 4 Map Visualization of the current map 5 Status bar Indicates the coordinates where the cursor is located the map s scale and the raster s position and value Les DIVA GIS 7 3 0 x 105 0294 y 32 5098 Scale 1 42881984 The following table shows key navigation commands that allow you to visualize and explore various aspects of the maps Each command is activated by clicking on the respective button Depending on the information attributed to each map certain buttons may be inactive Chapter 3 Commands Zoom in zooms in on a point by left clicking the mouse or zooms in on an area by drawing a rectangle while holding down the left mouse button Zoom out zooms out from a point by a left clicking the mouse Pan moves the visible zone of the map by holding down the left mouse button and moving the mouse Zoom to active layer zooms out to the extent of the currently active layer Zoom to full extent zooms out to the maximum extent of all layers Information provides information on the element identified by the mouse in the layer selected Remove Layer eliminates any layer that has been selected Table button see further explanation below Distance but
13. and Altitude Save the datasets on your computer You will notice that the 19 Bioclim variables can also be downloaded This is useful when performing species distribution modelling analyses with Maxent see Sections 6 2 6 3 and 6 4 For this analysis however only the monthly climate layers of Minimum temperature Maximum temperature Precipitation and the raster file A titude are used They are also used in Analysis 3 1 7 to create your own CLM file clm Note Download Windows Internet Explorer g Shd le worldclim org File Edit View Favorites Tools Help gi Favorites Download WORLDCLIM e 30 arc seconds resolution worldclim data can be downloaded by 30 x 30 degrees tiles generic data format only Click on the tile you want and then select a variable Gp lees ee WBZ amai lll CL oe RIE e 20 21 a 24 gt ge iv i Sale 211 Save 410 2 Zone 33 Mean Temperature Minimum temperature Maximum temperature Precipitation Altitude Bioclim Worldctim Extract the layers from the zipped files In DIVA GIS go to Data Import to Gridfile Multiple Files DIVA GIS 7 3 0 Project WEEE Layer Map Analysis Modeling Grid Stack Tools Help ae Import Points to Shapefile gt g a gt Te Import Text to Line Polygon Polygon to Grid as Climate Assign Coordinates W Check Coordinates gt Export Gridfile Import to Gridfile Single File Write VRT file Multiple Files
14. and to Jesus Salcedo for his most valuable support in the nitty gritty work of shaping the manual We are especially grateful to the DIVA GIS pioneers Robert Hijmans University of California Davis USA Luigi Guarino Global Crop Diversity Trust Italy and Andy Jarvis CIAT Colombia for their constructive feedback Their support and encouragement strongly motivated the authors to move forward with the manual s preparation Many other persons also provided feedback on the various drafts of the manual We would like to express our gratitude to our Bioversity colleagues Karen Amaya Margarita Baena Michele Bozzano Gea Galluzzi Prem Mathur Victoria Rengifo Evert Thomas Imke Thormann Veerle Van Damme and Barbara Vinceti as well as Sixto Iman INIA Peru and Diana Lara CONIF Colombia We would also like to thank Nicole Hoagland who with her invaluable editorial assistance considerably improved the manual s readability The examples used in this manual have been tested during the delivery of various training courses We are grateful to the organizers and donors of those courses organized between March 2006 and November 2010 in Argentina Bariloche and Pergamino Chile Pucon Colombia Cali and Cartagena Costa Rica Turrialba Ethiopia Addis Abeba Italy Rome Mali Bamako and Mexico M rida We would like to particularly thank the course participants as their feedback has contributed significantly to the impro
15. of this species DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid BEjemae Tools Help Dle fa Slee fp S Make stack Calculate Regression Cluster Export to Text File Overlap 8 Select the stack of rasters for which you wish to make the calculations In this case we select the stack that was made in step 6 9 Select the Sum option to add the binary rasters of the Vasconcellea species potential distribution areas 10 Under the Output tab indicate the name for the raster file which will be generated 11 Leave all other options as default 12 Click on Apply to start the calculating process v Stack Calculate Input Ae egis tutoriale 4 gap analysis yvasconcellea thresholds grs e Sum lAs Present Abzent Present gt 25 NULL as zero Calculate Area Select Mask Cut off gt 80 Output E 4GIS Tutorial6 4 Gap analysis vase pot richness grd e Species distribution modelling and analysis Number of Vasoncellea spp O amp Ww hw ot Ooh fw Be O t m OO wD a After having edited the map as explained in Chapter 3 the potential diversity map of the Vasconcellea species in Latin America should be similar to the one above The greatest area for expected diversity is in the northern Andean zone particularly between Ecuador and Peru and between Ecuador and Colombia Gap analysis using DIVA GIS To identify gaps in species distr
16. select the field Country 9a and click on the Reset Legend button Finally click on Apply The legend for the selected layer will be displayed Colours can also be changed in this menu by double clicking on the rectangle indicating the colour for each class 9b In order to assign different colours to each class it is important to have a fill style selected under the Single tab e g Solid Fill Chapter 3 _ Properties Label Latin America Countries Source amp ais tutorial 3 1 basic elements latin america countri e Properties Label Latin America Countries Source amp gis tutorial 3 1 basic elements latin america countri Type Polygon 117 30 32 72 29 30 56 11 Type Polygon 117 30 32 72 29 30 56 11 Field COLINTRY Reset Legend F Use color of single tab v E e ARGENTINA BELICE S K BOLIVIA g BRASIL K CHILE AZ tosh cs 10 A similar process is used for numeric values under the Classes tab This is not a commonly used process and is therefore not included in this manual If you have followed the steps correctly the map below should be displayed colours may be different Le DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Ose RQSo8S8 Hi taeeR 80 soos Latin America Countries Gs ARGENTINA BELICE BOLMA BRASIL CHILE COLOMBIA COSTA RICA CUBA ECUADOR EL SALVADOR GUATEMALA GUY
17. such as predictions of species distribution under future climates j Maximum Entropy Parameters Advanced Experimental _ Random seed v Give visual warnings Show tooltips Ask before overwriting _ Skip if output exists Remove duplicate presence records Write clamp grid when projecting Do MESS analysis when projecting Random test percentage Regularization multiplier Max number of background points Replicates Replicated run type Crossvalidate NA Test sample file Browse Species distribution modelling and analysis Visualizing the results of Maxent in DIVA GIS The raster of the potential distribution of Pinus kesiya generated by Maxent is in ASCII format asc It can be found in the output folder the same folder where the HTML file is stored In order to visualize and modify these files they will be imported to DIVA GIS 13 Open the Import to Gridfile Single File option in the Data menu to import the rasters in ASCII format asc to DIVA GIS see Chapter 3 wes DIVA GIS 7 3 0 Project BEEM Layer Map Analysis Modeling Grid Stack Tools Help D E Import Points to Shapefile import Text to Line Polygon hape Polygon to Grid Point Ce exf s erty H Climate Assign Coordinates W Check Coordinates Export Gridfile Single File E amp E Importo Gridfile Multiple Files Write VRT file Export Sha
18. tmini 0 grd train O23 grd na tmin10_33 qrd Altitude mask alt Make Index HOBESs Selected Files tmint grd tmas1 grd precl grd tmin grd tras grd precz grd tmina grd tras grd precs grd tron grd tras grd precd ord tran grd trax grd precs grd tmin grd tmax6 grd prec grd tming grd tmas grd precy grd tenn grd tmasS grd precd grd tming grd tras grd precd grd tran 0 grd trax 0 grd prec10 grd tmin 1 grd trax 1 grd precd 1 grd tran 2 grd tras 2 grd prec ord Progress Chapter 3 The process of making CLM files clm can be demanding if you have a computer with relatively low processing capacity If this is the case it is better to use climate layers with a lower resolution 5 or 10 minutes Note The CLM files clm are saved as the following climate cli alt clm index clm tmin clm tmax cim prec clm The cli file contains information about the characteristics of the CLM files clm It is recommended to change the name of the climate cli file to a more specific name in order to be able to distinguish the CLM files clm from other CLM climate databases In this analysis for example the name of the climate cli file can be changed in wclim_t33_30sec cli Do not change the names of the CLM files clm otherwise they will not be recognized in DIVA GIS Assuring identical raster properties before combining rasters Sometime
19. 06 the Prune button to cut the tree at the selected distance The resulting grouping of similar cells will be displayed as a tree with fewer branches the complete dendrogram can be displayed again using the Draw button When pruning at a value of 0 1 five groups will remain 9 To visualize the result assign a name to the raster in the Grid File window and then click on Make Map GeoCluster Distance Matrix Cluster Cluster File gis tutorial 5 3 diversity molecular marker data ssrs cher presabs dtr_ Reset Copy Show rn Distance width Fl Depth ie A Draw A Prune Prune 01 Set Prune Dist Labels V Aligned On Click Distance 0 08 0 06 mya Grid File e gis tutorial 3 diversity molecular marker data ssrs 5 lusters gr lt S Make Map Spatial analysis of diversity for conservation planning As a result of grouping similar cells in five different groups three large zones can be observed each having a different allelic composition one zone covers northern Peru and southern Ecuador the second covers the central part of Peru and the third is located in Bolivia The genetic structure of cherimoya distribution can be further explored using different pruning points However be aware that differences may no longer be apparent when the number of groups is reduced For example when cutting at a distance of 0 12 only three groups are left Chapter 5 The separation be
20. 1 species diversityvasconcellea species shp _ i Define Grid qe Use parameters from another grid Output Yarable on Richness l Number of observations Paint to Grid Procedure Simple Output E SQl5 Tutorial5 1 Species diversity vasconcellea obs eS__ _ S_ SS s AC tos Select parameters fram another grid E G lS Tutorial 5 1 Species diversityyasconcellea diversity 15 Even though this analysis focuses on the number of different observations per cell which is independent of the observed unit of diversity being analysed a field must still be indicated in the Parameters window This analysis will use the Species field as was done in Step 8 16 Indicate an appropriate name for the file click on the button left to the Output box and click Apply Spatial analysis of diversity for conservation planning The result after editing the raster legend as explained in Chapter 3 of this analysis reveals that most observations originate from southern Ecuador This is a typical situation in which areas known to have high levels of diversity are often preferred sites for botanists and researchers resulting in more intense sampling efforts and a higher number of observations Such uneven sampling often takes place in national parks near major cities or in zones with high endemism This is a common problem and is known as sampling bias With the rarefaction method explained in Section 5 3 this p
21. 17 08 10 2010 18 17 08 10 2010 18 17 23 04 2010 18 30 26 04 2010 10 47 08 10 2010 18 17 08 10 2010 18 17 26 04 2010 10 47 26 04 2010 10 47 08 10 2010 18 17 08 10 2010 18 17 08 10 2010 18 17 08 10 2010 18 17 01 12 2004 16 30 01 12 2004 16 30 01 12 2004 16 30 30 06 2010 13 12 Omission on training samples blue line shows the fractions of the presence points located outside the potential area as modelled by Maxent from low to high threshold values limiting the predicted area of occurrence Cumulative threshold Training samples is synonymous to presence points Fraction of background predicted red line shows the fractions of background points from the study area included in the modelled distribution area under varying Cumulative thresholds Predicted omission black line is a reference line If the blue line Omission on training samples is well below the black line Predicted omission this might indicate some over fitting because of dependence between presence points Omission and Predicted Area for Pinus_kesiya Fraction of background predicted Omission on training samples Predictedomission Fractional value o 0 A n O D W 20 30 40 50 60 70 80 20 Cumulative threshold Species distribution modelling and analysis 11 One of the parameters used for evaluating the predictive ability of the models generated by Maxent is the Area Under Curve AUC of the Receiver Operating Charac
22. 31667 6 3333 75 25 Colombia Antioquia 75 25 6 3333 17 31667 9 566667 90 25157 34 06767 cundinamarcensi 16 98333 6 9 75 9666 Colombia Antioquia 75 9666 6 9 16 98333 10 2 85 41 79677 cundinamarcensi 18 0625 6 1666 75 5666 Colombia Antioquia 75 5666 6 1666 18 0625 9 625 89 95327 34 78015 cundinamarcensi 17 44167 6 4861 75 3952 Colombia Antioquia 75 3952 6 4861 17 44167 10 11667 86 46724 51 51493 cundinamarcensi 17 025 6 2833 75 4333 Colombia Antioquia 75 4333 6 2833 17 025 9 733333 89 29664 34 21191 cundinamarcensi 16 39167 6 4627 75 5563 Colombia Antioquia 75 5563 6 4627 16 39167 10 1 84 67395 50 71459 cundinamarcensi 20 36667 6 335 75 5527 Colombia Antioquia 75 5527 6 335 20 36667 10 98333 868 57527 38 808 cundinamarcensi 15 29583 6 6477 75 4608 Colombia Antioquia 75 4608 6 6477 15 29563 10 40833 86 73611 42 983 cundinamarcensi 22 97917 5 8 75 5666 Colombia Antioquia 75 5666 6 8 22 97917 10 675 87 5 45 79889 _cundinamarcensi 17 1875 6 155 75 3736 Colombia Antioquia 75 3736 6 155 17 1875 9 208333 91 17162 29 08647 cundinamarcensi 17 00833 6 5166 75 5 Colombia Antioquia 75 5 6 5166 17 00833 10 2 85 54 01319 cundinamarcensi 14 90417 5 9833 75 35 Colombia Antioquia 75 35 5 9833 14 90417 8 325 91 48352 22 30556 cundinamarcensi 13 45 5 7256 73 7472 Colombia Boyaca 73 7472 5 7256 13 45 9 75 81 93277 35 291 cundinamarcensi 13 45 5 7408 73 7375 Colombia Boyaca 73 7375 5 7408 13 45 9 75 81 93277 35 291 OoN WN oon on
23. 760548 690683 442616 884110 148842 89782 850317 396282 177102 931986 989685 959186 361269 287799 512869 617039 274378 311401455394 74017 215459 0 000000 648238 190875 0 000000 893069 513645 0 000000 827545 456845 f onnnnn 92167 AeA BAL OODS 216 When you open the dBase IV file dbf or text file txt in Excel a summary of the number of times each allele occurs N and the maximum distance between alleles MAXD will be shown This distance provides you with an idea of the geographic area covered by each allele The remaining parameters minimum distance MIND and average distance AVGD are less important in this instance To visualize the information presented in Excel in a more organized format change the distances from meters to kilometres eliminate decimals and arrange the data in ascending order of MaxD Different allele classes will then be revealed MaxD km 112 Chapter 5 swa s l e The first group includes those alleles that have been observed only once highlighted in yellow These alleles are defined as unique although their occurrence depends greatly on the number of samples taken in an area Because of the strong influence of sampling intensity on identifying unique alleles this type of allele is considered to be of less interest than those in the following group The alleles of most interest for identifying zones of high or unique diversity are those repeatedly observed in
24. Africa Biodiversity and Conservation 12 1537 1552 This book contains 179 pages
25. Export Shapefile fi File Manager Download Chapter 3 5 Select the BIL BIP BSQ box 6 Keep default options for the input file 7 Click Add file and select the monthly climate layers of Minimum temperature Maximum temperature Precipitation and the raster Altitude file that you would like to convert to a grd file 8 Save the generated raster files grd in the same Output Folder as the raster files in the original format Same as nput You may also choose to save them in a different folder Select 9 Click Apply to start the process Import Multiple Files to Gridfiles Type Keep input file te Add file Remove file Remove all File name 3 4G iE Tu torial 3 1 B a z ic element sh alt 33 bil EGIS Tutoriala 1 Basic elements precd 33 bil EGIS Tutornial3 1 Basic elementssprece_33 bil EGIS Tutorialy3 1 Basic elements sprecs_33 bil EAGIS Tutoriala 1 Basic elements precd 33 bil E 4 l5 Tutoriah3 1 Basic elements preco 33 bil lal 37 files Porn Wl Close 10 Now you can open the climate layers in DIVA GIS by using the Add layer option The same procedure must be followed when raster data in the following formats is being imported to DIVA GIS IDRISI Arc BINARY and Arc ASCII For example the future climate data at the Downscaled GCM Data Portal http gisweb ciat cgiar org GCMPage are available in ASCII format Note Basic elements of spatial analy
26. In this analysis clicking on the nfo tab allows you to see that the maximum temperature is 29 17 C but the exact place on the map corresponding to this extreme value is difficult to locate Try to find the sites with the highest temperature 29 17 C in Latin America To do this click on the Legend tab and create a unique class that contains temperature data superior to 29 C Select red as the colour for this range The resulting map should show a few cells in Mexico and Colombia Now you must check each cell to identify which has the highest value The status bar at the bottom of the screen gives the exact value for each cell The hottest place in Latin America is located in southern Mexico Label mean temperature latin america 10 min Filename gis tutorials 3 1 basic elements mean temperature latin america 10 Legend Info History Color From To Auto complete Edit values Select rows Classify Select color scheme ra Ramp 2 Read From File Add or Remove Row 8 8 NoData Transparent ETS Basic elements of spatial analysis in DIVA GIS 3 1 3 How to combine rasters Similar to making changes in the legends for vector layers changing the legend in rasters only affects the display not the file s original information Sometimes it is relevant to combine selected zones of different rasters for example the hot areas from a temperature raster wit
27. Learning module IPGRI and Cornell University Frankel OH Brown AHD Burdon JJ 1995 The conservation of plant biodiversity Cambridge University Press Cambridge UK Chapter 5 Grum M Atieno F 2007 Statistical analysis for plant genetic resources clustering and indices in R made simple Handbooks for Genebanks No 9 Bioversity International Rome Italy Hajeer A Worthington J John S editors 2000 SNP and microsatellite genotyping Markers for genetic analysis In Biotechniques Molecular laboratory methods series Eaton Publishing Manchester UK Hijmans RJ Garrett KA Huaman Z Zhang DP Schreuder M Bonierbale M 2000 Assessing the geographic representativeness of genebank collections the case of Bolivian wild potatoes Conservation Biology 14 6 1755 1765 Kindt R Coe R 2005 Tree diversity analysis A manual and software for common statistical methods for ecological and biodiversity studies World Agroforestry Centre ICRAF Nairobi Leberg PL 2002 Estimating allelic richness effects of sample size and bottlenecks Molecular Ecology 11 2445 2449 Mathur PN Muralidharan K Parthasarathy VA Batugal P Bonnot F 2008 Data Analysis Manual for Coconut Researchers Bioversity Technical Bulletin No 14 Bioversity International Rome Italy Petit RJ El Mousadik A Pons O 1998 Identifying populations for conservation on the basis of genetic markers Conservation Biology 12 844 855 Pritchard JK St
28. Low impact on pine Potential new area Select color scheme No Data 2 Read From File Add or Remove Aow NoData Transparent Chapter 6 After editing the map as explained in Chapter 3 the result should be similar to the map above The results reveal that climate change will particularly affect P kesiya populations in the Chinese province of Yunnan due to an expected increase in seasonality and will impact areas at lower altitudes more generally as these are predicted to become too hot As such these areas may be prioritized when developing conservation strategies including those designed to collect germplasm to ensure the ex situ conservation of genetic resources before existing stands disappear under the prevailing changes in climate n situ conservation strategies to protect populations predicted to be highly affected might focus on improving the connectivity between fragmented populations to ensure gene flow of adaptive genes An alternative option would be to assist the migration of these species to newly suitable areas shaded in green on the map above In the case of P kesiya areas of the Indonesian island of Sulawesi and the Philippine island of Mindoro are expected to be suitable for Pinus kesiya occurrence by 2050 as a result of climate change An additional in situ conservation strategy might be to increase efforts to conserve populations in low impact areas where models predict the species wil
29. Pinus kesiya has diminished in recent decades Many remaining stands are continuously threatened by unsustainable resin extraction and wood harvesting practices and are also likely to be threatened by the effects of climate change Species distribution modelling along with climate models can help to determine the most at risk populations and to identify where the conservation of genetic resources requires urgent measures van Zonneveld et al 2009a In this analysis the potential impact of climate change on the distribution of P kesiya will be explored In this analysis you will learn how to use Maxent to predict the potential distribution of a species under current and future climatic conditions and to examine the impact of climate change on a species using DIVA GIS This analysis will use the climate projections for the year 2050 under the A2 emission scenario from three different GCMs CCCMA HADCM3 and CSIRO Each model has a slightly different projection of the future climate Therefore predictions of potential Chapter 6 distribution areas for a species will vary depending on the GCM used in Maxent Here the average of the three selected GCMs will be used Many files will be generated when performing this analysis Carefully name and save these files for easy access as they will be used frequently in this section Note Steps 1 To run the Maxent model follow Steps 1 to 8 outlined in Analysis 6 2 1 Next in the Projectio
30. SANTAFE DE BOGOTA ADMa o 1 30165 Vasconcell Rec 1 of 1 Layer Yasconcellea final errors ID 1669 SPECIES Vo cundinamarcensis LATITUDE 1 6667 LONGITUDE 78 6333 COUNTRY Ecuador ADMI Chimborazo A4DM2 ADM 3 While these errors are easily detected through the visualization of the map the whole database must also be checked for less evident errors To do this go to the Data menu and select Check Coordinates If the layer with the points has been selected this layer will become the Input otherwise click on Input File and select the layer with the observations sa DIVA GIS 7 3 0 Project Mee Layer Map Analysis Modeling Grid Stack Tools Help D E Import Points to Shapefile _ mport Text to Line Polygon Draw Shape Polygon to Grid Climate Assign Coordinates vf Check Coordinates gt Export Gridfile Importto Gridfile Write YRT file Export Shapefile fi File Manager K Download Quality control 4 Select the field with the data for longitude and latitude 5 Select the polygon file with the administrative unit data Latin America countries shp starting with the higher level data the layer with countries 6 Indicate the relationship between both files Country Country and click Apply DIVA GIS will immediately check the inconsistencies in the information in both layers Check Coordinates Shape af Poirt Input File amp qie tutorial 4 1 quali
31. Temperature trax_2 5rn Precipitation prec _2 5mi Projection GEOGRAPHIC Map units DEGREES Datum wed ff OK hd Apply 2 2 2 How to import climate data in Maxent In order to ensure a smooth process when modelling data in Maxent it is recommended to use environmental raster data in the ASCII format asc ASCII files asc can be created in DIVA GIS under the Data Export Gridfiles option Analysis 3 1 5 illustrates how to prepare ASCII climate data for a specific study area Section 6 2 explains how to use Maxent with the 19 bioclimatic variables for the prediction of potential species distribution For the example outlined below the 19 rasters in ASCII format asc located in the wclim_eth_2 5min_ascii folder are used Preparing and importing data to DIVA GIS and Maxent Steps 1 2 3 4 Open Maxent and indicate in Environmental layers the raster files to be included Use the Browse button to locate the folder wclim_eth_2 5min_ascii 4 Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples Environmental layers Browse DirectoryiFile Linear features fs N Quadratic features K Product features Threshold features Hinge features N N N Auto features Create response curves Make pictures of predictions Do jackknife to measure variable importance Output format Output file type
32. a small area for example alleles SSR1 289 and SSR7 208 highlighted in green These are referred to as locally common alleles The probability that this phenomenon occurs repeatedly and within a limited area is much lower than the probability of observing single alleles only once given the differences in the sampling strategy Populations that contain a significant number of locally common alleles may be prioritized for conservation because they contain alleles and most likely useful genes as well not found elsewhere while their observation is less dependent on differences in sampling intensity Another group of alleles includes those found in a relatively large area but with very low frequencies e g SSR7 191 and SSR8 318 highlighted in orange These may indicate incomplete sampling Finally the majority of alleles are common throughout the entire study area These are called common alleles and do not contribute much information to a diversity analysis 4 Look closer at the distribution of the alleles of interest SSR1 289 SSR7 191 SSR7 208 and SSR8 318 The first step is to select the layer of the SSRs under Analysis Point to Grid Richness This time select Presence Absence as the Output Variable Use rasters with one 1 degree cells Spatial analysis of diversity for conservation planning 5 Onthe Point to Grid tab select only the SSRs of interest see Step 4 Point to grid Point to grid Input Shapefil
33. combine results from different studies Due to the availability and use of various analytical options confusion as to the accuracy of results often arises among the multitude of studies results can be difficult to compare negatively affecting their validity This chapter aims to enhance the reader s understanding of the options available when undertaking spatial analysis of diversity and the associated implications for conservation the chapter provides guidance for selecting appropriate methodologies to conduct analyses and interpret results The first challenge when conducting any type of diversity analysis is determining the appropriate level at which to work Plant biodiversity is studied at the community level ecosystem the species level and the genetic level This manual focuses solely on the study of diversity at the species and genetic levels At the species level the observed unit of diversity is the species measured as present or absent in a certain location In terms of genetic diversity the observed unit of diversity may either be a phenotypic trait the product of gene genes expression or a DNA base pair composition analyzing neutral markers or known functional DNA based on sequences or molecular weights The diversity of species varieties alleles in distinct subunits within a study area known as alpha diversity is the principal subject of the spatial analysis of diversity Subunits of a study area may refer to previousl
34. consistent among different maps To use the legend information from an existing raster click the Read From File tab and indicate file path Properties Label mean temperature latin america 10 min Label mean temperature latin america 10 min Filename amp gis tutorial 3 1 basic elements mean temperature latin america 10 Filename gis tutorial 3 1 basic elements mean temperature latin america 10 Legend Info History Legend Info History Colod From To Auto complete Color From To Auto complete v Edit values 10 Edit values Select rows Select rows 5 0 0 5 5 10 Select color scheme Select color scheme 10 15 Red Green Blue m 7 Red Green Blue 15 20 mE p E Rame E 2 20 25 f on Read From File Pez m Read From File 25 30 tes Classify Classify hd 5 hd Add or Remove Row S Add or Remove Row oo coir Nodata No Data NoData Transparent NoData Transparent Close so E chose Basic elements of spatial analysis in DIVA GIS 11 The label text describing each class can also be manually modified For this analysis change the label for the first class From 10 5 To lt 5 and for the last class From 25 20 To gt 25 Click OK to illustrate these changes on the map Properties Label mean temperature latin america 10 min Filename sgis tutorial 3 1 basic elements mean temperature latin america 10
35. diversity analysis see Chapter 5 or species distribution modelling see Chapter 6 In this analysis the species distribution modelling of an Asian pine tree Pinus kesiya is used as an example 8 Open the file Distribution Pinus kesiya grd in DIVA GIS 9 Inthe menu go to Data Export Gridfile Single File es DIVA GIS 7 3 0 Project Layer Map Analysis Modeling Grid Stack Tools Help Import Text to Line Polygon Draw Shape Polygon to Grid f 5 Selection to New Shapefile Extract Values by Point Climate Assign Coordinates WY Check Coordinates Export Gridfile Single File f Importto Gridfile Multiple Files Write YRT file Export Shapefile fi File Manager Download 10 The Export Gridfile window will display automatically select Google Earth KMZ 11 Use the Output button to define the location and name of the kmz file Export Gridtile File type BIL D ESRI ASCII O Shapetile polygons 6 IDRIS 16bit O ESAI Binary FLT O Shapetile points 5 IDRISI 32bit O GRASS C Text Google Earth KMZ e es gis tutoriala 2 export to google earth distribution pinus kesiya grd fe Output egis tutorials 2 export to google earth distribution pinus kesiya Basic elements of spatial analysis in DIVA GIS 12 Finally go to the folder where the kmz file is saved and open it gt Google Earth Be Eat View Blogs a veo Search Fl
36. diversity analysis using molecular marker data are very similar to those of an analysis at the species level In this analysis we use microsatellite SSR marker data SSRs are short tandem repeats of base pairs highly variable and evenly distributed throughout the genome Hajeer et al 2000 De Vicente et al 2004a SRR analysis looks at differences in length of microsatellite regions usually not associated to functional genes i e neutral These differences in length further referred to as different alleles are the observed units of diversity in this analysis To carry out a spatial analysis the molecular marker data must be formatted in such a way that each allele includes georeferenced information In the following analysis each allele is formatted according to the following microsatellite code weight of base pairs e g SSR1 293 Review the example SSRs cherimoya rand column dbf in Excel to become familiar with the table format Microsoft Excel SSR cherimoya rand column dbf amp File Edit View Insert Format Tools Data Window Ge Al l SSRS 65 2001 17 1218 SSR1 293 65 2001 17 1218 SSR1 293 65 2001 17 1218 SSR2 122 65 2001 17 1218 SSR2 128 65 2001 17 1218 SSR3 216 65 2001 17 1218 SSR3 218 65 2001 17 1218 SSR4 154 65 2001 17 1218 SSR4 154 65 2001 17 1218 SSR5 156 65 2001 17 1218 SSR5 156 65 2001 17 1218 SSR6 148 65 2001 17 1218 SSR6 148 65 2001 17 1218 SSR7 183 65 2001 17 1218 SSR7 183 65 2001 17
37. done by opening the raster files grd using Notepad Look at the example below and note the information in the raster grd for Mean Annual Temperature BIO1 under the current climate and future climatic conditions in Southeast Asia P bio_1 Notepad File Edit Format View Help General a ee Notepad File Edit Format View Help General Creator DIVA GIS Creator DIVA GIS Created 20071204 Tit le bio_1 Georeference Projection Datum Mapunits Columns 8640 Rows 3600 Minx 180 MaxxX L80 Miny 60 MaxyY 90 Resolutionx 0 04166666667 q Resolutiony 0 04166666667 GeorReference Projection Datum Mapunits Co lumns 8640 Resolutionx 0 04166666665 lt q Resolutiony 0 04166666665 4 Even though the number of columns and rows and the coordinates for the vertices Minx Maxx MinYy MaxY of both rasters are the same there is a difference in the eleventh decimal of the resolution Resolution X and Resolution Y The difference is extremely small and will not lead to variations in the visualization but will generate an error preventing the comparison of potential distribution areas for current and future climates when converting rasters to ASCII format appropriate for Maxent To solve this problem the different values can be manually changed in the raster documentation files grd to make them equivalent D bio_1 Notepad File Edit Format View Help General Creator DI
38. file xls 2 Open DIVA GIS and go to Data Import Points From Excel XLS Les DIVA GIS 7 3 0 Project EEEE Layer Map Analysis Modeling Grid Stack Tools Help mee Import Points to Shapefile a From text file TXT Import Text to Line Polygon From dBase file DBF Polygon to Grid W Climate Assign Coordinates WY Check Coordinates Export Gridfile Importto Gridfile Write YRT file Export Shapefile A File Manager Download Preparing and importing data to DIVA GIS and Maxent 3 In Excel spreadsheet select the Excel file that has the presence points you would like to import Vasconcellea xls to DIVA GIS Always close the Excel file before importing data to a vector file in DIVA GIS Note 4 Under Worksheet select the Excel sheet with the data you wish to import into DIVA GIS 5 Select the column with the longitude coordinates 6 Select the column with the latitude coordinates 7 Click on Save to Shapefile to generate a vector file shp Excel to Shapetile points Excel spreadsheet EGIS Tutorialy2 1 Importing observation datahyasconcellea eo Ci eY Latitude LATITUDE i Save to Shapefile o TUD LONGITU COUNTA ADM candic 4 3201 79 7585 Ecuador candic 4 4283 79 7918 Ecuador candic 4 0816 79 933 Ecuador candice 4 0646 79 6411 Ecuador candic 4 0655 79 6416 Ecuador candic 4 0653 79 6411 Ecuador candic 4 423 79 7192 Ecuador candic 4 0697
39. for use with DIVA GIS and Maxent The analyses outlined in this manual are based on two components of observation data passport data and associated data Basic passport data consists of an identification code ID ataxonomic identification of an observed individual plant or group of individuals and the location of the collection or observation site The analyses presented in this manual mainly use basic passport data commonly referred to as presence points Presence points may be associated with additional data describing the collection or observation sites e g land cover soil type etc the date of collection source of coordinates name s of collector s and the institution housing conserving the specimen Information about the collection date provides a time dimension to the analyses and can be used to analyze trends in species distribution e g to monitor possible genetic erosion The opposite of presence points is absence data While absence data can also be relevant in spatial analyses e g monitoring trends in species distribution it is often challenging to understand concrete reasons for the absence of taxa in a geographic unit complicating the use of this data in ecological analyses such as those that will be explained in Chapter 6 Ataxon might be extinct due to human disturbance or its absence might be explained by dispersal limitations or changes in the local environmental conditions Further the idea that plant c
40. into account the respective proportions of each species in the study area which is also referred to as the measurement of evenness In the case of diversity analysis using molecular markers indices of allelic frequencies as a measure of evenness are used in order to ensure relative proportions of different alleles on each locus in the study area are considered However a challenge when utilizing indices which take proportions into account is that these are not appropriate for cells with a limited number of observations which is generally the case with analyses based on high resolution rasters Thus when working with high resolution rasters richness may still be the most appropriate method to measure diversity In addition to richness and evenness other measurements to assess alpha diversity specific to certain types of data may be applied When measuring morphological traits statistical parameters such as variance and the coefficient of variation can be calculated to determine levels of phenotypic diversity at a given site See Section 5 2 Further the application of a multivariate analysis results in distances between individuals in multivariate space i e Euclidian distances This information can then be used to group similar individuals and undertake subsequent richness analyses In the case of molecular marker data in addition to the previously described diversity analysis based on allelic richness specific genetic parameters e g he
41. ki Options Output Yarable Reserve Selection owt ee Complementarity Rebelo o Paint to Grid Procedure Simple Output mes EGIS Tutorial 5 4 Conservation Strategies reserve se E k D 7D D_msr eZ 2 Each allele should receive the same weight default option yet you can also give more weight to rare alleles or species if necessary Point to Grid Parameters Scoring 4pproach Equal weight O Weighted by rarity Minimum Records per Cell 1 Masinun number of iterations Al classes v 55A1 287 55A1 289 W 55A1 290 ai S5R1 291 J 55A1 293 W S55A1 2300 J 551 303 OF ccp Hn Selection A Spatial analysis of diversity for conservation planning 3 The first output visualized is a report on the number of cells selected eight necessary to conserve all alleles and the number of unique observations the 51 alleles generated by the analysis Complementarity Process Report umber of iterations cells Number of unique observations captured 51 Complementarity Algorithm Finished All values were selected 4 Next click OK on the Report window Three rasters titled Sequence Classes and Additional Classes will be displayed Before visualizing each result improve the legend using the NoData transparent option to broaden the number of classes Sequence This raster corresponds to the most import
42. latin america 10min 7 precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min 20 precipitation latin america 10min Chapter 3 4 Finally if you would like to extract all the climate data you can repeat this process but instead of selecting the Extract Values by Points From Grid or Stack option you will need to select the Extract Values by Points From Climate data option This can only be done if CLM files clm with climate data are connected to DIVA GIS see steps outlined in Analysis 2 2 1 The resulting file will be a text file which includes all selected climate data as well as the original data if selected aa Extract Environmental Data Points egis tutonals 7 basic elements asoncellaa specie Output EAGIS Tutonal3 1 Basic elements Yasconcellea bio Climate database worldelim_2 5m Select variables J Altitude Yl Precipitation yi Minimum Temperature x Bioclim yi Maximum Temperature Include fields ID x SPECIES J LATITUDE SelectAll J LONGITUDE Select Al COUNTRY Clear All AA Apply Rl Clos
43. menu Select Records or Select Features these options are also available for quick access in the toolbar buttons and 14 Click on Select Records to select groups according to a specific variable Select by values or by a combination of variables Select by query For this analysis all occurrences of Vasconcellea in Ecuador have been selected Yellow points indicate the group selected sa DIVA GIS 7 3 0 Project Data MEGHE Map Analysis Modeling Grid Stack Tools Help D ae a Add Layer X Remove Layer 7 Latin America co Properties 7 abc Add Labels SS Identity Feature Table Filter Select Records O Copy amp Hide Show Legend Basic elements of spatial analysis in DIVA GIS Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Oe a eaaiaia lela x Pm lelz Aio ts S O E Latin America Countries Vasoncellea species o Select by values Select by query Variable COUNTRY New selection Values 15 A Argentina F Bolivia Brasil Colombia D Costa Rica TETEH O El Salvador E Guatemala D Honduras D Mexico Nicaragua E Panama a Paraguay E Venezuela t Select All Clear All AA Apply QL Close y 32 5098 Scale 1 42881984 Data A Design d 15 The next step is to save the selection as a ne
44. nodes DBN and days to harvest DH Three additional CV layers must be created 8 The layers generated must now be combined to form one layer by calculating the average CV of the five layers Go to Stack Make Stack and add the five rasters with the calculated CVs using the Add Grid button Assign a name to the stack and click on Apply Combining different rasters into a stack is only possible if all rasters have the same properties as outlined in Chapter 3 a Make Stack Make Stack Check Stack l Add Grid l Remove Grid l Remove All File name Resolutior s Resolution Minx e gis tutorial 5 2 diversity phenotypic 77 9972 e gis tutorials 5 2 diversity phenotypic 42 77 9972 e gis tutorials 5 2 diversity phenotypic 42 77 9972 e gis tutorial 5 2 diversity phenotypic 42 77 9972 e gis tutorials 5 2 diversity phenotypic 42 77 9972 E GIS Tutorial 5 2 Diversity Phenotypic data cvs ars AA Apply Rl Close 9 Goto Stack Calculate and calculate the mean of the five rasters in the stack Stack Calculate Input Stack egis tutoria 5 2 diversity phenotypic dataycys grs E Sum AsPresent dbsent Present gt 25 O NULL as zero Calculate Area Select Mask Cut off gt 80 EGIS Tutonal5 2 Diversity Phenotypic dataycvs sum grd Spatial analysis of diversity for conservation planning The combination of the layers for the five morp
45. not have access to Excel you may use Calc instead a free software programme available through the OpenOffice package which can be downloaded at http www openoffice org This manual refers to the most recent versions of the above mentioned programmes according to the date of the manual s publication It is expected that the examples and subsequent step by step instructions included in this text will remain valid in the near future however it is acknowledged that these may change as new software versions are developed or if these programmes become obsolete Therefore it is recommended that the software and programme websites be consulted periodically for any possible updates which may be required The following paragraphs provide detailed instructions for downloading and installing DIVA GIS and Maxent on your computer 1 1 Installation of DIVA GIS DIVA GIS is the main programme used to conduct and display the results of the spatial analyses presented in this manual The installation of this programme is rather straightforward All that is required is Version 7 3 0 diva730 zip of the software installer available at http www diva gis org download DIVA GIS was developed for the Windows operating system but it can also run on Mac OSX see the DIVA GIS website for additional information Note The DIVA GIS website also offers climate data see Section 2 3 thematic layers a user manual additional training materials and li
46. of use data or profits or business interruption however caused and on any theory of liability whether in contract strict liability or tort including negligence or otherwise arising in any way out of the use of this software even if advised of the possibility of such damage Additional license information For software included in this distribution of DIVA X I accept the agreement I do not accept the agreement 6 Indicate the file path where you would like to save the programme on your hard disk The best option may be to use the default location automatically offered by the installer If you prefer to install the programme in another location select the file path using the Browse button jB Setup DIVA GIS Select Destination Location Where should DIVA GIS be installed CJ Setup will install DIVA GIS into the following folder To continue click Next If you would like to select a different folder click Browse C Program Files DIV4 GIS 7 3 Atleast 14 1 MB of free disk space is required Chapter 1 7 Indicate the folder from which you will have direct access to the programme in the Startup Menu We recommend using the default option jB Setup DIVA GIS Select Start Menu Folder Where should Setup place the program s shortcuts Setup will create the program s shortcuts in the following Start Menu folder To continue click Next If you would like to select a different folder cl
47. sample size equivalent to 10 homozygous individuals or 5 heterozygous individuals Steps 1 Select the layer with the molecular marker data and go to Analysis Point to Grid Richness Under Output Variable select Rarefaction Use the same raster file as in the richness analysis for one 1 degree cells See Individual Task 2 Under the Parameters tab mark the SSRs using a Standardized Sample Size of 80 ea Point to Grid Pointto Grid Parameters Input Shapetile eis tutorial 5 4 conservation strategies ssr cherimoya ran Define Grid Use parameters fram another grid Options Field 55A5 Output Variable Al classes Richness st WI SSR1 287 We ff SSRA1 289 Riain oa 55A1 290 aretaction ial wf 55A1 291 7 55A1 293 Foint to Grid Procedure WI SR1 300 Co W 55A1 303 imple W 55A1 310 J SSR2 112 Output S5R2 118 F eap a E 4GIS Tutorials 3 Diversity Molecular marker datas Selection al Invert AA Apply M Close Ml Close As can be observed the difference in diversity between northern Peru and Bolivia where much sampling was carried out becomes even more evident in this analysis Chapter 5 Geographical distribution of individual alleles The previous analysis focused on the total number of alleles Now we will look at the distribution of individual alleles Based on their frequencies and geographic distributions different types of alleles can be id
48. sites rather than across the entire study area Williams et al 2002 Several methods exist to reduce the sample bias of a dataset but these can only resolve the problem to a limited extent Chapters 5 and 6 discuss some of these methods in further detail e g circular neighbourhood rarefaction and species distribution modelling 4 1 Quality control based on administrative unit information One way to evaluate the accuracy of a presence point is to compare the administrative unit data included in its passport data with administrative unit information extracted from thematic layers based on the geographic coordinates of the point In order to do this the passport data needs to include data for the country and preferably for lower administrative units The Data menu of DIVA GIS provides the Check Coordinates option which allows one to check the quality of coordinates based on the administrative unit data PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 4 1 Quality control Administrative units e Excel e Vasconcellea final errors shp shx dbf e Latin America countries shp shx dbf e Latin America Adm 01 shp shx dbf 4 1 1 How to verify data quality based on passport administrative unit data In the following analysis you will use a portion of the dataset corresponding to information on the geographic distribution of highland papayas this information was collected dur
49. species distribution modelling in Chapter 6 will be based on these bioclimatic variables they will also be used in this analysis Be sure to check All which will generate layers for all 19 bioclimatic variables based on an identical raster 5 Define the output folder where the climate layers will be generated under the File tab 6 Inthe selected folder 19 Bioclim raster files will be created as follows pe Date Modified Fao B101 grd GRD File 11 12 2009 12 00 Forestry E B101 gri GRI File 11 12 2009 12 00 Gates 2 8102 0rd GRD File 11 12 2009 12 00 m ccot E B102 gri GRI File 11 12 2009 12 00 GEF On Farm 2 B103 grd GRD File 11 12 2009 12 00 GEF REMERFI Si B103 44i GRI File 11 12 2009 12 00 7 Genetic erosion 2 s104 9rd GRD File 11 12 2009 12 00 cis 6104 9 GRI File 11 12 2009 12 00 7 GIS CHERLA 2 B105 grd GRD File 11 12 2009 12 00 E cis Peru I BI05 ari GRI File 11 12 2009 12 00 GIs tutorial B106 grd GRD File 11 12 2009 12 00 Alexandra fi 8106 ari GRI File 11 12 2009 12 00 Americas 2 B107 grd GRD File 11 12 2009 12 00 andy Si B107 94i GRI File 11 12 2009 12 00 arcas 2 8108 0rd GRD File 11 12 2009 12 00 SI B108 gri GRI File 11 12 2009 12 00 2B B109 grd GRD File 11 12 2009 12 00 E B109 gri GRI File 11 12 2009 12 00 L B1010 grd GRD File 11 12 2009 12 00 E B1010 gri GRI File 11 12 2009 12 00 2 B1011 grd GRD File 11 12 2009 12 00 E B1011 gri GRI File 11 12 2009 12 0
50. the rasters of environmental variables covering the study area First a niche is defined based on the environmental values that correspond to the presence points used in the analysis Then the similarities between the environmental values at a specific cell and those of the niche of the modelled species are calculated for each raster cell in the study area With this information the model calculates the probability of a species occurrence in each raster cell Even though the next analysis is based on climate data Maxent and other niche programmes also allow one to include other types of variables in the model e g soil variables As mentioned the Modeling menu of DIVA GIS includes the Predict option see Section 6 1 which provides for two integrated species distribution modelling programmes Bioclim and Domain these are different from Maxent Each programme uses different statistical methods to estimate the realized niche and to calculate the probability of species occurrence in each raster cell Therefore results are likely to differ Maxent calculates the species realized niche and probability of occurrence using an algorithm for maximum entropy Philips et al 2006 As Maxent has fared well in evaluations in comparison to other programmes Elith et al 2006 Hernandez et al 2006 it is the programme of choice Chapter 6 in this manual to undertake species distribution modelling analyses It is important to realize that when
51. the same location You can recognize this in the Attributes box when more than one presence point is indicated e g Rec 1 of 2 If this is the case you can view in the Attributes box the passport data of the different outliers present at the same location by clicking on the arrows on the right hand of the window where the passport data of each point is presented Note Normally a point would be classified as atypical after using a combination of various bioclimatic variables in the analysis Using the Outliers tab you can define the conditions and the set of bioclimatic variables important for the occurrence of a specific species you can also indicate the minimum environmental variables required in order to consider a presence point as atypical with extreme values in different bioclimatic variables 12 Go the Outliers tab and select the climatic variables you would like to include in the atypical points analysis For this analysis all variables have been selected You can also select a set of climatic variables you consider key for the occurrence of the species under study 13 Select the minimum number of variables for which a presence point should have atypical values to be considered an outlier In this analysis three variables are selected 14 The lines on the graph represent presence points Red lines represent outliers under the conditions specified Double click on one of the lines making sure that the options a Attribut
52. to download climate layers with 30 second resolution in tiles of 30 x 30 degrees from the Worldclim website http www worldclim org tiles php This section illustrates how to generate climate raster files grd from a BIL raster file bil of 30 seconds for a specific region of 30 x 30 degrees available from Worldclim Future climate layers available for download from the Worldclim webpage http www worldclim org futdown htm and from the Downscaled GCM Data Portal http gisweb ciat cgiar org GCMPage can be imported to DIVA GIS in a similar manner In Section 6 3 the use of future climate data in climate change impact studies on plant distributions and diversity is described in further detail Basic elements of spatial analysis in DIVA GIS Steps Go to the website http www worldclim org tiles php Select tile 33 by clicking on it alternatively these four files are also included in the Basic Elements folder Importing rasters to ArcGIS from DIVA GIS Raster files created in DIVA GIS can also be imported to ArcGIS To do this you must start by converting the raster files grd from DIVA GIS to ASCII files asc by using the option ESRI ASCII or by converting them to FLOAT flt files using the option ESRI binary FLT These file can then be imported in ArcGIS 3 4 Download the zipped datasets for Minimum temperature Maximum temperature Precipitation these datasets consist of monthly climate data
53. 0 2 B1012 grd GRD File 11 12 2009 12 00 E B1012 gri GRI File 11 12 2009 12 00 Articulos cientificos Bioclim South Africa 5 BiodiversityR Bolivia Bioclim Cartagena course Chapter 7 input files para Jesus O CHERLA Colm comments Data 03112009 3 Data 14102009 B1013 grd GRD File 11 12 2009 12 00 DATOS EJERCICIOS Manual SIG E B1013 gri GRI File 11 12 2009 12 00 Diva cIs B1014 grd GRD File 11 12 2009 12 00 E B1014 gri GRI File 11 12 2009 12 00 Elementos basicos d i D Files Robert BIO15 gr GRD File 11 12 2009 12 01 D Final files E BIO15 gri GRI File 11 12 2009 12 01 Sjeroie gri GRI File 11 12 2009 12 01 For Colm 2 B1016 grd GRD File 11 12 2009 12 01 Google Earth 2 B1017 grd GRD File 11 12 2009 12 01 Installation 61017 qr GRI File 11 12 2009 12 01 Karen n GRD File 11 12 2009 12 01 T Basic elements of spatial analysis in DIVA GIS From the table in Section 2 2 derived from www worldclim org bioclim we know that BIO1 refers to the mean annual temperature and BIO12 to annual precipitation Add these two layers to the map 10 1h As noted in the introduction to Chapter 2 GIS and species distribution modelling programmes use a variety of raster file types DIVA GIS uses grd format while Maxent works best with rasters in ASCII format asc
54. 0 200 E 0 200 0 400 O 0 400 0 600 m 0 600 0 800 E 0 200 1 000 E No Data Latin America Countries SS Vasoncellea species o mean temperature latin america 10 O 10 29 d E 29 30 e gis tutorial3 1 basic elements precipitation latin america 10min grd e gis tutorial 3 1 basic elements precipitation latin america 10min_rec grd Grid Stack From To New value x Oo 0 1000 1 1000 10000 0 Data Type Integer Minimum Maximum 9918 Save RCL Read RCL A OK X Close y 32 0958 Scale 1 42881984 precipitation latin america 10min 737 Row 3 Col 11 Data A Design Basic elements of spatial analysis in DIVA GIS 4 DIVA GIS allows you to combine rasters using the Grid Overlay tool We now need to identify zones with average temperatures greater than or equal to 20 C and annual precipitation below 1000 mm Under nput files select the two layers generated in the previous steps and then click Multiply Please see the following table to understand how the calculation works Grid Overlay Input files First e gis tutonalys 7 basic elementssprecipitation latin america 10min_rec grd Operation C Add O Minimum Substract Divide O Maximum Output file Result E Gl5 Tutoria3 1 Basic elementsoverlay temp prec grd Add to map AA hon M Cose Average Average Average Average temperature temperature temperature temper
55. 0 963 If the predicted area is low in comparison to the study area high AUC values doesn t necessarily reflect good model performance and simply could be an artifact of the AUC statistic Philips 2009 For information about AUC and the ROC curve please refer to Fawcett 2006 Sensitivity vs 1 Specificity for Pinus_kesiya Training data AUC 0 968 Random Prediction AUC 0 5 0 0 0 1 0 2 0 3 0 4 0 5 0 6 0 7 0 8 1 Specificity Fractional Predicted Area The table in the next step Step 12 illustrates how thresholds can be used to limit the potential area of a species in this case Pinus kesiya These thresholds represent the minimum probability of a species potentially occurring in the environment of a specific cell This concept assumes that sites with a probability above the threshold have climatic conditions appropriate for the occurrence of the species while the species would not occur in sites below the threshold The higher the threshold the more restricted the potential distribution areas There is no standard threshold value and the user must define this parameter for additional information on thresholds for potential distributions and niche limits see Lui et al 2005 Chapter 6 12 In this analysis the 10 percentile training presence threshold will be used the probability value at which 90 of the presence points fall within the potential area The remaining 10 which fall outside the potential ar
56. 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 E AVGD 112315 529031 51763 223028 0 000000 214634096226 900183 573709 202933 815713 0 000000 893501 718703 392918 292471 324706 614673 816187 139674 929194 775737 279173 418610 926758 162019 702397685715 834741 331538 295162 040476 892268 399590 850633 885704 0 000000 243369 403381 291750 047439 0 000000 943831 352155 894483 741541 334092 460561 371729 455237 990902 231982 895319 112602 208000 229928 363813 233640 256772 118712 272437 476582 35 SSR5 184 36 SSR6 134 37 SSRB 142 38 SSR6 146 186 39 SSR6 148 546 40 SSR6 150 2 41 55F7 177 53 42 SSR7 183 936 43 SSR7 189 280 44 SSR7 191 7 45 55F7 193 89 46 SSR7 206 43 47 SSR7 208 20 48 SSRB 299 416 49 SSR8 301 167 50 55R8 305 363 25660058 602972 9eN2917 077070 rn ABT 4 gt gt I Distances SSRs 158 284 410 2543594 122515 265727 2 248067 265727 2 248067 2349079 386966 2641113 593885 89782 850317 1221947 037797 2657272 248067 2601975 255826 891200 946919 2157677 288490 12507 10 467133 268120 411937 2657272 248067 2363000 788622 0 000000 0 000000 0 000000 0 000000 0 000000 89782 850317 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 676115 378699 581088 314935 963164
57. 09a worked with a minimum number of 50 presence points for two pine tree species with a broad geographic distribution throughout Southeast Asia PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 6 1 Realized niche e Excel e Vasconcellea_4species shp shx dbf For this analysis you need to have the 2 5 min worldclim climate data imported in DIVA cf Section 2 2 Species distribution modelling and analysis 6 1 1 How to analyze and compare realized niches of different species Vasconcellea species are adapted to different environments To evaluate these differences the climate data of the respective presence sites for each species can be extracted using DIVA GIS see Chapter 3 Understanding of the adaptive capacity of each species to different climatic conditions will provide information on the potential for inclusion in breeding programmes e g cold tolerance in papaya as well as for the commercial growing of Vasconcellea species taking into account optimal agroecological zones Identification of the limits of the species niche also provides key information for conservation for example species that occur in narrow climate ranges are likely to be more vulnerable to climate alterations than those with a broad climatic niche The next section outlines how to determine the niche of the following Vasconcellea species V cundinamarcensis V microcarpa V quercifolia and V parvif
58. 1218 SSR8 305 65 2001 17 1218 SSR8 316 65 1752 17 1442 SSR1 293 65 1752 17 1442 SSR1 293 65 1752 17 1442 SSR2 122 65 1752 17 1442 SSR2 122 55 1752 17 1442 SSR3 222 65 1752 17 1442 SSR3 230 65 1752 17 1442 SSR4 152 65 1752 17 1442 SSR4 154 65 1752 17 1442 SSR5 156 65 1752 17 1442 SSR5 156 55 1752 17 1442 SSR6 142 65 1752 17 1442 SSR6 142 65 1752 17 1442 SSR7 183 55 1752 17 1442 SSR7 183 65 1752 17 1442 SSRB 305 65 1752 17 1442 SSRB 316 64 3412 17 4599 SSR1 293 64 3412 _ 17 4599 SSR1 293 In this section you will learn to a Use the Analysis Point to Grid and Distance option in DIVA GIS b Use the Cluster option in DIVA GIS Chapter 5 Individual Task Visualize cherimoya sampling sites included in the diversity analysis see Chapter 3 Basic Elements Individual Task Conduct an analysis of allelic richness first using one 1 degree cells and next based on 10 minute cells with a circular neighbourhood of one 1 degree see Section 5 1 Analyze where the highest level of diversity is found The rasters generated shown above indicate that northern Peru is the region with the highest level of diversity highest number of different alleles found while Bolivia maintains the lowest level Rarefaction method As seen in the first analysis Analysis 5 1 1 uneven distribution of observations can have a significant impact on the richness analysis The rarefaction method has been developed to c
59. 34 41 99567 9 2 12 3 15 6 15 46667 3037 V cundina 7 1716 75 7633 Colombia Antioquia 75 7633 7 1716 21 49583 10 25833 93 40108 40 81323 15 4 12 3 21 23333 21 71667 2816 V cundina 7 0333 75 3166 Colombia Antioquia 75 3166 7 0333 19 64583 9 775 78 83065 45 45019 12 8 12 4 19 46667 19 65 2836 V cundina 6 3333 75 25 Colombia Antioquia 76 25 6 3333 17 31667 9 566667 90 25157 34 06767 22 5 11 9 10 6 17 13333 17 11667 3030 V cundina 6 9 75 9666 Colombia Antioquia 75 9666 6 9 16 98333 10 2 85 41 79677 22 9 10 9 12 15 56667 17 1 3418 V cundina 6 1666 75 5666 Colombia Antioquia 75 5666 6 1666 18 0625 9 625 89 95327 34 78015 23 6 129 10 7 17 75 17 86667 3336 V cundina 6 4861 75 3952 Colombia Antioquia 75 3952 6 4861 17 44167 10 11667 86 46724 51 51493 23 2 11 5 11 7 17 18333 17 08333 3351 V cundina 6 2833 75 4333 Colombia Antioquia 75 4333 6 2833 17 025 9 733333 89 29664 34 21191 22 4 11 5 10 9 16 76667 16 85 3208 V cundina 6 4627 75 5563 Colombia Antioquia 75 5563 6 4627 16 39167 10 1 84 87395 50 71459 22 3 10 4 11 9 16 15 16 5 3205 V cundina 6 335 75 5527 Colombia Antioquia 75 5527 6 335 20 36667 10 98333 88 57527 38 808 26 7 14 3 12 4 20 68333 20 1 3027 V cundina 6 6477 75 4608 Colombia Antioquia 75 4608 6 6477 15 29583 10 40833 86 73611 42 983 21 1 9 1 12 15 08333 14 98333 3354 V cundina 5 8 75 5666 Colombia Antioquia 75 5666 5 8 2297917 10 675 87 5 45 79889 29 5 17 3 2 22 51667 22 85 2993 V cundina 6 155 75 3736 Colombia Antioqui
60. 5 Fj s Wf Fj F Fj s Fj 55A1 267 55A1 259 55A1 290 55A1 291 5501 293 55A1 300 55A1 303 55A1 310 5502 112 55A2 118 5502 120 55A2 122 55A2 126 55A2 125 S5A3 216 S5A3 218 cco 3 790 f Selection All Clear Invert hi Apply M Cancel 38 Microsoft Excel Distances SSRs dbf Spatial analysis of diversity for conservation planning File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF 2 SSR1 287 3 SSR1 289 4 SSR1 290 5 8SR1 291 6 SSR1 293 7 SSR1 300 8 SSR1 303 9 SSR1 310 10 SSR2 112 11 SSR2 118 12 SSR2 120 13 SSR2 122 SSR2 126 SR2 128 S5R3 216 SSR3 218 SSR3 220 SSR3 222 20 SSR3 230 SR3 300 SSR4 143 SSR4 145 24 SSR4 147 25 SSR4 152 26 SSR4 154 27 SSR4 156 28 SSR5 146 29 SSR5 156 30 SSR5 160 31 SSR5 163 32 SSR5 164 33 SSR5 166 34 SSR5 182 2 MND 112315 529031 98936 601656 0 000000 1182386 541258 2657272 248067 567875 621069 0 000000 2643314 747000 2232807 340967 1003586 748579 2379663 697653 2657272 248067 536421 608697 2657272 248067 2604522 694471 2601975 255826 1191918 073621 2656410 934834 2601179 904615 0 000000 1089554 155549 1233690 030137 0 000000 2642452 654205 2657272 248067 2226441 909032 129437 4 183262 2596570 336769 2530188 786944 846556 684103 1337294 585461 1176946 540977 962460 889177 112315 529031 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0 000000 0
61. 79 9225 Ecuador candicg 4 3181 79 7963 Ecuador candic 4 3263 79 7907 Ecuador 2 1 3 How to import georeferenced presence points in Maxent Maxent is a species distribution modelling programme used for predicting the potential distribution of one or more species As with DIVA GIS Maxent works with georeferenced presence data This section illustrates how to import georeferenced presence points to Maxent A detailed explanation of how Maxent works is presented in Section 6 2 In Maxent data must be imported as a Comma Separated Values CSV Comma delimited file csv and should include three fixed fields columns corresponding to the following categories Species Longitude and Latitude Columns should be listed in this specific order but additional columns with more information are permitted The three column file can be prepared by starting with a database originally constructed in Excel and then saved in CSV Comma delimited csv format There are slight differences in procedure when creating CSV files using Excel 2007 and Excel 1997 2003 Chapter 2 Steps Steps for preparing a CSV Comma delimited file csv for Maxent in Excel 1997 2003 Hereafter the steps for preparing a CSV file csv in the Excel 2007 are explained The Excel file Vasconcellea xls is used in the following example the file can be found in the data folder accompanying this section 1 In a new Excel sheet copy the data for the variabl
62. ANA GUYANE HAITI HONDURAS JAMAICA MEXICO NICARAGUA x AeA PANAMA E PARAGUAY Label Latin America Countries PERY Source _ amp gis tutorial 3 1 basic elements latin america countri PUERTO RICO REPUBLICA DOMINICANA Type Polygon 117 30 32 72 29 30 56 11 SURINAME TRINIDAD Y TOBAGO Fiel T TR URUGUAY ield OUNTRY Reset Legend Use color of single tab Vasoncellea species F ARGENTINA o Single Unique Classes Oe aL aL te U 6 a E BELICE BOLIVIA CHILE v v v BRASIL v S SelectAll Clear All Ny Close x 89 1905 y 32 0958 Scale 1 742881984 Data Design Basic elements of spatial analysis in DIVA GIS Individual Task Make Brazil green and eliminate the colour in the remaining countries Result Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Osi BA cla Hxdee eRAo 0 es Latin America Countries C ARGENTINA C seuce C Bouma EE irasi C cue C coom C costarica C cusa C ecuapor CO ELSANADOR C suatewaa C cuvana C cuvane C Ham C Honouras C sawaica C mexico C NICARAGUA E Label Latin America Countries Source e gis tutorial 3 1 basic elements latin america countri C Puerto rico Ss core ne Type Polygon 117 30 32 72 29 30 56 11 C SURINAvE Single Uni
63. Bonierbale M 2000 Assessing the geographic representativeness of genebank collections the case of Bolivian wild potatoes Conservation Biology 14 6 1755 1765 Scheldeman X Willemen L Coppens D eeckenbrugge G Romeijn Peeters E Restrepo MT Romero Motoche J Jimenez D Lobo M Medina Cl Reyes C Rodriguez D Ocampo JA Van Damme P Goetghebeur P 2007 Distribution diversity and environmental adaptation of highland papaya Vasconcellea spp in tropical and subtropical America Biodiversity and Conservation 16 6 1867 1884 Williams PH Margules CR Hilbert DW 2002 Data requirements and data sources for biodiversity priority area and selection Journal of Bioscience 27 4 327 338 OOO Section B Data Analysis Chapter 5 Spatial analysis of diversity for conservation planning Chapter 5 Spatial analysis of diversity for conservation planning Statements regarding the current state of biodiversity are frequently made in articles and reports focusing on conservation These include comments such as 75 of diversity has been lost or this site is rich in diversity However in order for such claims to be used in the policy and decision making process they must be credible and based on well established methodologies for assessing levels of diversity and comparable measurements Although this may sound evident and simple many questions still exist as to the best ways to measure diversity and how to compare and
64. C Select oo 0 154 0 2 ii Classify hed 0 2 0 2 03 Mee oF l AaS Select color scheme jaa 0 4 0 4 0 5 o J m 0 5 05 06 rs Ramp BB 0 6 06 07 ne Read From File A 0 7 l 07 08 0 8 0 9 Add or Remove Row G 0 9 1 oole Nodata No Data NoData Transparent 21 An administrative unit layer seacountries shp should be added to facilitate locating the potential distribution areas Species distribution modelling and analysis Final result Potential distribution P kesiya After modifying the potential distribution map of P kesiya as outlined in Chapter 3 it should appear similar to the image above Some countries such as the Laos and Myanmar have few presence points available Maxent reveals extensive areas in these countries where this pine species may potentially occur Species distribution modelling conducted with Maxent completes the observed distribution and in this instance emphasizes the existing knowledge gaps This means that the vacant areas require further study in order to determine the location of pine populations and define conservation strategies for this species this issue will be further discussed in Section 6 4 potential distribution P kesiya Most likely the species does not occupy all the potential distribution areas represented by the model because species dispersal is limited by its reproduction system and the presence of geophysical and climatic barriers In this i
65. Classes option to distinguish between the different classes within the vector file shp i e different species genotypes countries This option must be selected as this analysis focuses on four different species 5 In the Field window select the parameter that will be used to define the different classes In this case select Species The complete list of Vasconcellea species will be displayed Chapter 6 Indicate whether or not you wish to remove duplicate presence points within the same Cell in the climate raster For this analysis select the Remove duplicates From same grid cell option This will prevent cells with many observations from contributing disproportionally to defining the realized niche Distribution Modeling Input Frequency Outliers Histogram Envelope Predict Points e gis tutorial 6 1 realized niche wasconcellea_4species shp DIVA Climate data worldclim_2 5m Remove duplicates With same coordinates vV From same grid cell One Class Many Classes Classes 4 y V cundinamarcensis y V microcarpa Y parviflora x Y quercifolia t l All l Invert Clear l Above Below Inbetween ql Close The menu includes five types of analysis five different tabs two of which will be explained in this section Histogram and Envelope The Frequency and Outliers analysis options were already described in Chapter 4 The Predict op
66. ECICI KOJ ats koi i axe j R Result after 5 future potential current potential Situation ee Shee subtracting rasters distribution areas distribution areas cell value AET cell value OHighimpactareas tt Outside of realized niche o0 0 o0 iii Low impact areas iv New suitable areas To solve this problem in DIVA GIS go to the Reclass option and change the cell value for potential distribution areas from one 1 to two 2 for one of the two rasters see Analysis 3 1 3 The following table reflects how this change in cell value will result in a different cell value for all four situations Raster of Raster of i Result after A k future potential current potential i Situation be ke Be he subtracting rasters distribution areas distribution areas cell value cell value cell value New suitable areas For the next analysis change the value for the potential distribution areas from one 1 to two 2 for the rasters of the predicted future distribution of Pinus kesiya 5 Go to Grid Reclass to change the raster cell value of potential distribution areas under future climate Pinus_kesiya_gcm_sea_2 5min_thresholded gra from one 1 to two 2 DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Stack Tools Help Dae QlQ oo gy PF Describe Ne mWOBEe Oo Rmo Overlay Scalar seacountries pinus_kesiya Neighborhood Calcul
67. Find maps driving directions hotels restaurants and more PAu e Make browsing the web faster safer and easier serene e Search from the address bar E Include Google Chrome a fast new browser for Windows and Mac Learn about Google Chrome Make Google Chrome my default browser By installing you agree to Google Earth s Privacy Policy Google Maps Earth Terms of Service By downloading installing or using the Google Earth software accessing or using the Google Maps service together the Products or Services or accessing or using any of the content available within the Products you agree to be bound by the following 1 the Google Terms of Service the Universal Terms 2 the terms found on our Legal Notices page the Legal Notices and 3 the additional terms and conditions set forth below the Additional Terms Before you continue you should read each of these three documents as together they form a binding agreement between you and Google Customize your installation of Google Earth with advanced setup Agree and Download Cancel e Registered Pro users download Google Earth Pro e Want the Google Earth plug in only Download Google Earth plug in 2 Open the downloaded installation file 3 Follow the on screen instructions The installation will start by downloading additional data Chapter 1 1 4 Data for analysis In addition to installing the required software programm
68. H BF 4 The next step consists of calculating the distance between different raster cells based on the composition of alleles found in each cell Since the data being used is binary values of one 1 for presence of one allele and zero 0 for absence of alleles leave the default option Binary selected Three methods are used to calculate distances In this analysis the Dice method will be utilized to calculate the matrix of distances It is also known as the Nei Li coefficient and can be used with co dominant marker data such as microsatellite marker data see De Vicente et al 2004b It is recommended to use the same filename as was assigned to the stack file Click OK to continue a GeoCluster Distance Mati Cluster Input Stack eis tutorial 3 diversity molecular marker dataisers cher presabs grs Method 0 Jaccard 1 a la b c 5 Lance and williams Bray Curtis b c 2 a b c a Classes in common e Dice Czekanowski Sorensen 1 2 a 2 a b c b c classes not in common e Binary Present gt 1 Use nodata as 0 Continuous Scale Use mask data nodatal Output Distance Matrix E AGIS Tutorial5 3 Diversity Molecular marker data ssrs cher presabs dmt Spatial analysis of diversity for conservation planning 5 The next step consists of undertaking a cluster analysis based on grouping cells separated by the smallest distances i e with the most similar allelic compositio
69. NT DATA 3 BASIC ELEMENTS OF SPATIAL ANALYSIS IN DIVA GIS 3 1 VISUALIZATION IN DIVA GIS 3 1 1 How to perform basic visualizations using vector files 3 1 2 How to perform basic visualizations using rasters 3 1 3 How to combine rasters 3 1 4 How to extract values from rasters based on presence points data 3 1 5 How to create custom made climate layers 11 11 12 16 16 17 18 18 20 21 23 25 28 29 34 35 36 38 41 41 43 50 57 60 62 3 1 6 3 1 7 How to import generic climate data to DIVA GIS How to make CLM files in DIVA GIS 3 2 EXPORTING LAYERS TO GOOGLE EARTH 3 2 1 How to export data to Google Earth 3 3 EDITING MAPS AND FINALIZING A PROJECT 4 QUALITY CONTROL 4 1 QUALITY CONTROL BASED ON ADMINISTRATIVE UNIT INFORMATION 4 1 1 How to verify data quality based on passport administrative unit data 4 2 QUALITY CONTROL THROUGH THE IDENTIFICATION OF ATYPICAL POINTS 4 2 1 How to identify outliers based on environmental data SECTION B DATA ANALYSIS 5 SPATIAL ANALYSIS OF DIVERSITY FOR CONSERVATION PLANNING 5 1 SPECIES RICHNESS Skel How to carry out a spatial analysis of species richness 5 2 INTRA SPECIFIC DIVERSITY ANALYSIS BASED ON PHENOTYPIC DATA 5 2 1 How to carry out a spatial diversity analysis using phenotypic data 5 3 INTRA SPECIFIC DIVERSITY ANALYSIS BASED ON MOLECULAR MARKER DATA Sal How to carry out a spatial diversity analy
70. Output directory Projection layers directory file Settings Once you have located the folder select it by clicking only once do not double click or open the folder Click Open Loon allie C diva_wordclim_2 5min wclim_eth_2 5min_ascii e File Name welim_eth_2 5min_ascii Files of Type directories and csvfiles s M Chapter 2 5 All the raster files stored in the folder will be automatically shown in the Maxent interface Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples File Browse Continuous Continuous SSSR Continuous Continuous R R Continuous Continuous Continuous R R R Continuous R Continuous Continuous R Select all Deselect all Linear features Create response curves Quadratic features Make pictures of predictions v Do jackknife to measure variable importance _ Output format Product features Threshold features Output file type Hinge features Output directory Browse w Auto features Projection layers directoryfile Browse Settings The environmental layers are now ready to be used for species distribution modelling analyses This is explained in Section 6 2 2 3 Sources of spatial and other relevant
71. S An alert sign shown below will appear on the screen if you attempt to import data directly from Excel 1997 2003 to DIVA GIS oicezo0y You may need to install database drivers For Office 2007 to make this work Go to download website The alert sign indicates that you may need to install database drivers for Office 2007 Clicking on Go to download website directs you to the Microsoft Download Center to get the respective drivers People who work with another program than Excel to manage their passport database or still have problems importing presence points from Excel to DIVA GIS after having installed the drivers can import alternatively their points to DIVA GIS from text files txt CSV files csv or dBase IV files dbf Importing points from these file types to DIVA GIS is similar to introducing them from Excel following the steps explained in this example except that the presence point database has to be saved as another file type For further explanation about importing presence points from these file types please consult the DIVA GIS Operating Manual available online at http www diva gis org documentation Note Steps for importing data from Excel to DIVA GIS 1 Excel 2007 Save your presence points as an Excel file 1997 2003 xls and close Excel 2007 If you save as Excel file xlsx the presence points cannot be imported to DIVA GIS Earlier versions of Excel Save your presence points as an Excel
72. S Setup Wizard Setup has finished installing DIVA GIS on your computer The application may be launched by selecting the installed icons Click Finish to exit Setup Launch DIVA GIS Chapter 1 1 2 Installation of Maxent The Maxent programme will be used for the species distribution modelling analysis outlined in Chapter 6 of this manual To download and use Maxent correctly you will need to have Java installed in your computer Most modern computers have Java installed by default but it is also freely available at http www java com If you find that Maxent will not run on your computer it is most likely that you do not have Java installed java com Java You Windows Intemet Explorer Goris java com File Edit View Favorites Tools Help ww Favorites java com Java You 1 2 1 Steps 1 If your navigator is Internet Explorer go to http www java com es download help JAVA YOU DOWNLOAD TODAY ree Java Download Experience Java in Action l Alice 3 f M2 Ou ae Java Alice Java Sony Ericsson Java Neil Young Java Blu ray Dise Java Road Trip Java Amazon Kindle ORACLE How to install Java ie online install xml If your navigator is Firefox go to http www java com en download help firefox online install xml 2 Click Install to start the process Java Setup Welcome Welcome to Java Java provides safe and secure access to th
73. Steps 1 Mapping different alleles is the basis for a spatial analysis of allelic composition Repeat the mapping process done in Steps 4 and 5 of the previous section Geographical distribution of individual alleles Analysis Point to Grid Richness with Presence Absence do not select any specific alleles Make a raster showing the distribution of each allele using one 1 degree cells The result generated can again be found in the newly created folder which contains distribution rasters for each of the 51 alleles SSR1 287 grd E SSR1 287 ori 2 SSR1 289 grd E SSR1 289 ori 2 SR1 290 grd es SSR1 290 gri 2 55R1 291 grd SSR1 291 ori 2 55R1 293 grd A SSR1 293 gri SSR1 300 grd SSR1 300 gri SSR1 303 ard ss SSR1 303 gri SSR1 310 rd SSR1 310 gri 9SR2 112 9rd E SSR2 112 gri 2 9SR2 118 ard E SSR2 118 0ri S5R2 120 ard E SSR2 1 20 gri 2 9SR2 122 ard es SSR2 122 gri 2 SSR2 126 grd E SSR2 1 26 0ri 3SR2 128 rd A few examples fa SSR2 1 28 gi F 55R3 216 grd E S8R3 216 gri F 55R3 218 grd E SSR2 218 gri S5R3 220 grd E SSR2 220 gri S5R3 222 grd a SSR3 222 ori S5R3 230 grd SSR3 230 gri SSR3 300 grd a SSR3 300 gri F 55R4 143 grd E SSR4 143 gri SSR4 145 9rd E SSR4 145 0ri 9SR4 147 rd E SSR4 147 ori 9SR4 152 grd E SSR4 152 gri 9SR4 154 ard E SSR4 154 gri SSR4 156 grd E SSR4 156 0ri SSR5 146 9rd E SSR5 1 46 gri A common allele SSR1 310 2 55R5 156 grd
74. The individual or group of individuals occurring at the location of an atypical point may have unique characteristics and provide interesting genetic material for breeding programmes interested in adaptive resistance to environmental stresses such as drought or extreme temperatures Atypical points may also indicate a sample bias such as under sampling in the geographic areas environments where the atypical points are found Although the different explanations for atypical points mentioned above error individuals growing in an extreme climate sample bias are all relevant in spatial biodiversity analysis it is often difficult to determine the actual cause for such points Consequently it is hard to decide what to do with them i e to eliminate them from or include them in the dataset The first step to determine appropriate corrective action is to locate the original information and verify the origin of the points However if the data was sourced from third parties like the GBIF this is often not possible and the decision to eliminate or keep atypical points requires careful consideration Points are likely to be erroneous when they reflect a completely different climate than the rest of the dataset if this is the case these points should be removed It is more difficult to determine a threshold from which a point can be considered an outlier with sufficient probability that it is erroneous Various statistical methods uni and multivariate
75. VA GIS Created 20071204 Tit le bio_1 P bio_1 Notepad File Edit Format View Help General GeorReference Projection Datum units Georeference Projection Datum Mapunits Columns 8640 Maxx 180 Miny 60 Maxy 90 Resolutionx 0 04166666667 lt q Resolutiony 0 04166666667 MaxxX 180 Miny 60 MaxyY 90 Resolutionx 0 04166666667 lt q Resolutiony 0 04166666667 q After adjusting the eleventh decimal the resolutions are now identical and the rasters can be combined This solution is appropriate for adjusting any kind of minor difference which exists in the coordinates of the vertices Winx Maxx MinY MaxY Chapter 3 3 2 Exporting layers to Google Earth Google Earth uses specially formatted files kml and kmz These formats are exclusive to Google Earth therefore information layers must be converted to files using this format in order for the data to be visualized using Google Earth DIVA GIS Version 7 1 includes the option for exporting both vector and raster layers to Google Earth PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 3 2 Export to Google Earth e Google Earth i e Vasconcellea species shp shx dbf e Distribution Pinus kesiya grd gri 3 2 1 How to export data to Google Earth Steps 1 To export a vector file shp to Google Earth it must first be opened with DIVA GIS For this analysis open the Vasconcel
76. a 75 3736 6 155 17 1875 9 208333 91 17162 29 08647 22 2 12 1 A 17 01667 17 05 3339 V cundina 6 5166 75 5 Colombia Antioquia 76 5 65166 17 00833 10 2 85 54 01319 22 9 10 9 16 75 16 63333 3350 V cundina 5 9833 75 35 Colombia Antioquia 75 35 5 9833 14 90417 8 325 91 48352 22 30556 19 5 10 4 A 14 78333 14 75 3055 V cundina 5 7256 73 7472 Colombia Boyaca 73 7472 5 7256 13 45 9 75 81 93277 35 291 19 4 75 9 13 38333 13 06667 1 oon on kwh oon onkwnN ndina ANA olommbia Boyaca Ang 22 To construct graphs of the two dimensional climatic niches reorganizing the columns according to the following is recommended Column A D Column B SPECIES Column C the climatic variable for the X axis of the graph in this case BIO1 Annual mean temperature Column D the climatic variable for the Y axis of the graph in this case BIO12 Annual precipitation 38 Microsoft Excel vasconcelela 4sp xls a File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF D E F G H J K L M N 0 P biot bio12 LATITUDE LONGITUC COUNTRY ADM1 RecNo PointNo Lon_ext Lat_ext biol bio2 bio bio4 bio5 cundinamarcensi 15 8 6 9636 75 4177 Colombia Antioquia 75 4177 6 9636 15 8 10 31667 83 87534 41 99567 cundinamarcensi 21 49583 7 1716 75 7633 Colombia Antioquia 75 7633 7 1716 21 49563 10 25833 83 40108 40 81323 cundinamarcensi 19 64583 7 0333 75 3166 Colombia Antioquia 75 3166 7 0333 19 64583 9 775 70 83065 45 45019 cundinamarcensi 17
77. a geographical area shows environmental conditions favourable for a species this does not necessarily mean that the species actually occurs in this area Dispersal limitations because of the species reproduction system and geophysical barriers can prevent a species from occupying all the geographic areas showing an environment similar to that of its realized niche It is also true that a species may not be present in areas where it could possibly occur if its natural habitat has been altered by human interference The importance of presence point data quantity As mentioned in the introduction of Section 6 1 a sufficient number of presence points is crucial for obtaining sound modeling results The illustration above shows an example of potential distribution maps of Carica papaya generated in DIVA GIS using the Bioclim modeling tool Results of potential distribution are established as the number of points increases in the modeling process from 5 to 108 presence points After 50 points prediction of potential distribution stabilizes and does not change significantly even if more presence points are included PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 6 2 Potential distribution e Excel e pkesiya csv e Maxent and Java e seacountries shp shx dbf e Folder wclim_sea_2 5min asc files 6 2 1 How to model the potential natural distribution of a plant species Pinus kesiya is a
78. able from global databases The most commonly referenced database is Worldclim Hijmans et al 2005 which uses 19 derived bioclimatic variables Busby 1991 in addition to monthly climate data maximum and minimum temperature and precipitation Compared to the commonly used temperature and precipitation parameters the 19 bioclimatic variables are more directly related to the physiologic aspects of plant growth and do not consider the timing when a particular state occurs i e it does not matter whether the hottest month is July Northern hemisphere or January Southern hemisphere Some bioclimatic variables include typical basic climate parameters e g BlO1 mean annual temperature or BIO12 annual precipitation while others combine temperature and precipitation in one variable e g BIO18 precipitation during warmest quarter Others capture aspects of seasonality e g BIO4 for temperature BlO15 for precipitation which can also be important to determine a species distribution The 19 Bioclimatic Variables BIO1 Annual mean temperature BlO2 Mean diurnal range max temp min temp monthly average BIO3 Isothermality BIO1 BIO7 100 BlO4 Temperature Seasonality Coefficient of Variation BIO5 Max Temperature of Warmest Period BlO6 Min Temperature of Coldest Period BlO7 Temperature Annual Range BIO5 BlO6 BlO8 Mean Temperature of Wettest Quarter BIO9 Mean Temperature of Driest Quarter BlO10 Mean Temper
79. al Diversity Article 7 Identification and Monitoring online Available from http www cbd int convention articles shtml a cbd 07 Data accessed October 2010 FAO 2009 The International Treaty on Plant Genetic Resources for Food and Agriculture online Available from http www planttreaty org Data accessed October 2010 GBIF 2009 The Global Biodiversity Information Facility online Available from http www gbif org Data accessed October 2010 Guarino L Jarvis A Hijmans RJ Maxted N 2002 Geographic Information Systems GIS and the conservation and use of plant genetic resources In Engels JMM Ramanatha Rao V Brown AHD Jacson MT editors Managing plant genetic diversity International Plant Genetic Resources Institute IPGRI Rome Italy pp 387 404 OOOO Section A Basic elements and data preparation Installation of software and example data for analysis Chapter 1 Installation of software and example data for analysis In order to follow the instructions for each analysis presented in this manual you will require the following computer programmes DIVA GIS Maxent Google Earth and Excel The first three programmes are available online and free of charge at http www diva gis org http www cs princeton edu schapire maxent http earth google com The Excel programme however is not freely available but is part of Microsoft Office which is installed on most computers If you do
80. alysis of diversity for conservation planning 5 To complete the analysis add the layer of protected areas using the Protected Areas Latin America shp file and analyze the current status of conservation for priority cells based on the Sequence layer This final step reveals that there are protected areas in each one of the cells providing the first indication of in situ conservation of cherimoya more detailed analyses are required for real situations The results of the analysis seem to indicate that cherimoya diversity is currently partially conserved since protected areas are located in all priority cells The cell with the highest priority however has low coverage in a protected area An analysis at a smaller scale would enable more in depth conclusions by showing whether cherimoya accessions with important alleles are actually included in these protected areas References Chapman AD 2005 Principles of Data Quality version 1 0 Report for the Global Biodiversity Information Facility Copenhagen on line Available from http wwwe2 gbif org DataQuality pdf Date accessed October 2010 De Vicente MC Lopez C Fulton T 2004a Molecular marker learning module volume 1 Using Molecular Marker Technology in Studies on Plant Genetic Diversity Learning module IPGRI and Cornell University De Vicente MC Lopez C Fulton T 2004b Molecular marker learning module volume 2 Genetic diversity analysis with molecular marker data
81. ant result the sequence of cell selection In other words the raster indicates the priority cells for conservation the cell with a value of one 1 is the most important By selecting the layer and locating the pointer over each cell the values can be seen in the lower part of the map on the status bar Here the priority cells in order of importance include 1 the cell located in northern Peru 2 the cell located a bit further south 3 the cell in Bolivia and 4 the cell located the southern part of Ecuador This analysis highlights that cells were not selected based only on the highest diversity but also based on differences in allelic composition as illustrated by including a cell from Bolivia for which diversity is not very high Chapter 5 This raster corresponds to a richness analysis showing only the information richness for those cells selected in the previous analysis Using the pointer it is possible to verify that the cell selected as having the highest priority actually corresponds to a cell with the highest level of diversity in the group 39 alleles Note in this analysis the second priority cell is not the one with the second highest level of diversity Additional Classes This raster reveals the number of new alleles contributed by each additional cell taking into account that all cells selected cover the 51 alleles The first cell contributes the largest number of alleles 89 alleles Spatial an
82. ap as a png tif or omp file but does not allow for any further editing Saving projects in DIVA GIS Vector and raster files are saved automatically when generated and modified in DIVA GIS A specific combination of files can be saved as a project which is defined as a combination of different raster and vector files If you wish to return to your work simply re open the project this will automatically open all layers that constitute the map so you will not need to create the map again Use the E icon or select Project Save as to save a project To open a project saved at an earlier time use the Project Open option or click on the icon A project will be saved as a div file This file only contains the location on the hard disk path of each layer in the project Therefore it can only be opened on the same computer on which it was originally created provided there are no changes in the location of the layers To share or exchange projects go to Project Export Project Before exporting convert the div file into a dix file which is interchangeable Remember that a dix file can be very large depending on its contents resolution and number of layers which makes it difficult to exchange Individual vector and raster files can also be shared but be sure to share all files which make up a single layer Vector files shp shx and dbf Raster files grd and gri the image files bpw and bmp are gener
83. ared and combined as will be done in this analysis The predicted potential distribution areas under current conditions and future projections will be compared identifying those distribution areas that will be strongly affected by climate change as well as those areas where the impact will be less severe and new areas for the natural occurrence of the species in the future Species distribution modelling and analysis 2 To generate binary rasters in Maxent go to the Advanced settings menu and under the Applied threshold rule select a threshold that limits the potential area This threshold 10 percentile training presence is the same as that which was manually visualized using DIVA GIS in Step 21 of Analysis 6 2 1 Run the analysis in Maxent Maximum Entropy Parameters Experimental Add samples to background _ Add all samples to background _ Write plot data Extrapolate Do clamping Write output grids Write plots _ Append summary results to maxentResults csv file Cache asciifiles Maximum iterations s00 Convergence threshold 0 00001 Adjust sample radius o Log file maxent log Default prevalence Apply threshold rule 10 percentile training presence a Examine the impact of climate change using DIVA GIS In the first part of this section binary rasters under the current climate and future projection were generated using Maxent in ASCII format asc These rasters will now be imported t
84. ate Pinus_kesiya_gcm_sea_2 5min_tl E i o Aggregate Mo Disaggregate Cut pinus_kesiya_thresholded Merge E i o Eoc New Transect Area 6 Under Input select the raster for which you will change the cell values 7 To avoid confusion during the reclassification change the respective cell values in the From column from negative one 1 to zero 0 and zero 0 to one 1 Then change the cell values from one 1 to two 2 under the New value column Chapter 6 8 Indicate where the raster with the new cell values will be saved Save the raster under a different name Normally this is done by adding the suffix rec 9 Save the raster with the new cell values ome Reclass fo Input egis tutorial 6 3 climate changepinus_kesiya_ocem_sea_2 Smin_thresholded grd e Output E 4GIS Tutoria 6 3 Climate changepinus_kesiya_gcm_sea_2 5rin_thresholded_rec Grid Stack Hew value Minimum Masimum Save REL C 10 After having changed the values go to Grid Overlay and superimpose the rasters for current and future potential distribution areas Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Meme Stack Tools Help Dice Ql AIA g7 BF Describe Overlay Scalar Reclass Neighborhood Calculate pinus_kesiya_gcom_sea_2 5min_t 7 seacountries Aggregate Disaggregate 7 pinus_kesiya Cut o Merge 7 Pinus_kesiya_qem_sea_2 5min_t New E 1 0 Eo Trans
85. ated automatically when visualizing a raster file but are not an essential part of the file References DIVA GIS 2005 User Manual version 5 2 on line Available from http www diva gis org docs DIVA GIS5 manual pdf Date accessed October 2010 Ramirez J Bueno Cabrera A 2009 Working with climate data and niche modelling Creation of bioclimatic variables Tutorial on line Available from http gisweb ciat cgiar org GCMPage docs tutorial bcvars creation pdf Date accessed October 2010 Quality control Chapter 4 Quality control One of the main objectives of undertaking the spatial analysis of biodiversity data is to provide information to assist in effective policy and decision making processes for natural resource conservation and use In order to formulate appropriate management and conservation strategies it is critical that datasets are of high quality and are precise Chapman 2005a 2005b The use of incorrect or low quality information may have significant consequences on the relevance and appropriateness of subsequent recommendations decisions and even investments Genebank and herbarium data available through biodiversity networks such as the GBIF see Section 2 3 are increasingly used in biogeographical studies however such data are from third parties and the origins are often unknown making the issue of data quality even more pertinent The specific objective of this chapter is to show users how to ide
86. ature 1 X Class Sal cundinamarcensis lv cundinamarcensis v gt o p az o te u Annual Mean Temperature 1 Display Minimurn Maanu 5 30 Display Minimum Maximum Class Frequency v Q 7 6 11 6 11 6 15 5 15 5 19 5 19 5 23 5 23 5 27 5 10 15 15 20 20 25 Annual Mean Temperature 1 Annual Mean Temperature 1 AA tore AL Close A cose Envelope 13 14 15 16 17 18 The Envelope tool allows you to visualize a two dimensional niche based on two climatic variables Select the desired species to generate a two dimensional niche Look at the different Vasconcellea species particularly V cundinamarcensis Select the two climatic variables on which the climatic niche will be based For this analysis select Annual Mean Temperature and Annual Precipitation Click Apply to visualize the two dimensional niche Green points within the blue rectangle of the climatic niche represent those presence points with a climate profile within the range limits of all 19 Bioclim climatic variables Red points represent presence points with a climate profile of which one or more of the values of the 19 Bioclim climatic variables are outside the range limits Red points within the blue rectangle are presence points having a climate profile with values for the selected variables Annual Mean Temperature and Annual Precipitation within the range limits o
87. ature 220 C and 220 C and lt 20 C and lt 20 C and Preciptation Precipitation Precipitation Precipitation lt 1000 mm gt 1000 mm lt 1000 mm gt 1000 mm wae E a a raster cell value maa T a raster cell value Combination multiplication The result is a raster where the cells with a combined value of one indicate the areas that comply with the two conditions Chapter 3 There are many other options to visualize and manipulate vector and raster layers Some important options include Selecting Grid Aggregate to combine cells and reduce the resolution of a raster see Section 6 4 and Grid Disaggregate to split cells to increase the resolution note that this can give a false sense of precision The objective of these processes is to ensure two rasters have the same resolution and extent which is critical when combining them Using Stack Calculate to combine three or more rasters with the same characteristics in resolution origin and size using more complex manipulations e g calculations than those provided by Overlay Clipping part of a raster using Grid Cut This can be used to create a dataset which includes only the area known as extent you are working on This will result in a reduction of processing time when running an analysis such as species distribution modelling which is explained further in Chapter 6 This is particularly important if the dataset you are using is very large Chang
88. ature of Warmest Quarter BlO11 Mean Temperature of Coldest Quarter BlO12 Annual Precipitation BlO13 Precipitation of Wettest Period BlO14 Precipitation of Driest Period BlO15 Precipitation Seasonality Coefficient of Variation BlO16 Precipitation of Wettest Quarter BlO17 Precipitation of Driest Quarter BlO18 Precipitation of Warmest Quarter BIO19 Precipitation of Coldest Quarter 3 Original list and further reading http Awww worldclim org bioclim Preparing and importing data to DIVA GIS and Maxent Climate data are imported to DIVA GIS from CLM clm files available on the DIVA GIS website http www diva gis org climate htm or are prepared for a specific study area based on global climate data see Analysis 3 1 6 From the CLM files basic temperature parameters maximum and minimum temperature and precipitation as well as values for the 19 bioclimatic variables can be extracted in DIVA GIS as either layers or associated data with observation points CLM files in different resolution are available at http www diva gis org climate htm PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 2 2 Importing climate data e Excel e Folder diva_worldclim_2 5min zip file downloaded from e Maxent and Java http www diva gis org climate e Folder wclim_eth_2 5min_ascii asc files 2 2 1 How to import climate data to DIVA GIS This section explains ho
89. avel to cities acro he globe dive into the depths o a aaa 7 Upgrade to the ultimate mapping application for business users Learn more Products Overview Showcase Learn Explore the world in 3D from Take a tour of Google Earth and New Watch our tutorial videos to help guide anywhere see all the places you can go you through the features in Google Earth Desktop E A Pan a Pan Searching for Places Full featured access to mp T i Google Earth l earn how to find any location on our planet wy Earth in Google Maps 3D Buildings Ocean T and beyond Search for 3D maps in your browser s businesses and get driving directions r Mobile Watch video Google Earth in the palm of your hand Moon Sky Looking for specific information Want to keep up to date For Non Profits For Businesses Subscribe to our Sightseer Newsletter For Educators For Media Read the Google Lat Long Blog For Developers For Data Providers Follow us on Twitter E Google Earth Download Google Earth for PC Mac or Linux Windows Internet Explorer z Go 8B hte dwww google com earthidownloadige agree html File Edit View Favorites Tools Help yy Favorites Y Google Earth Download Google Earth for PC fy gt mp v Pagey Safetyy Toolsy y Google earth Engish US Home Explore Download Learn Connect Help Download Google Earth for PC Mac or Linux e Zoom from space to street level tour the world can N e
90. ay be associated with JAR files hindering the running of Maxent In order to allow JAVA Maxtent to run it might be necessary to remove the file association between Winrar and JAR files To do this unselect the JAR option on the Integration Tab of the Options Settings menu Note Chapter 1 File Edit View Favorites Tools Help Q Back Q B p Search gt Folders fia Address la E GIS Tutorialtmaxent Size Type Date Modified Fae and roker Taska 1KB MS DOS Batch File 20 07 2010 09 03 f Make a new folder xentj 646KB Executable Jar File 20 07 2010 09 01 13KB Text Documen t 20 07 2010 09 06 A Publish this folder to the Web 6 If you run Maxent regularly it may be useful to create a shortcut to the MS DOS batch file bat on your desktop for quick access 1 3 Installation of Google Earth Google Earth has recently become an important tool for visualizing and sharing geographic information Google Earth allows georeferenced data on the distribution of a taxon to be visualized in combination with high quality satellite images The increasing availability of these images will undoubtedly boost the use of Google Earth while similar applications are also likely to appear in the market With Google Earth layers can be visualized in different scales ranging from the global level continent country etc to levels as specific as a single tree which could correspond to a presence point in a database providing numero
91. ayer keep it selected Le DIVA GIS 7 3 0 Project Data MEME Map Analysis Modeling Grid Stack Tools Help DALF Add Layer X Remove Layer Latin America 4 Properties abc Add Labels identify Feature Table Filter 3 Select Records Vasoncellea oe Copy Z Hide Show Legend The cursor will immediately take the shape of a cross allowing you to manually select a group of points keep holding the left mouse button down and drag the cursor over the group of points you would like to select The selected points will change colour Once again the selection can be saved as a new layer Step 15 Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Oe aeaQ ee x saoe2 Roy VAOas x 75 2150 y 4 8617 Scale 1 42881984 3 1 2 How to perform basic visualizations using rasters As mentioned in Chapter 2 DIVA GIS also manages raster data raster files have the extension grd in DIVA GIS Actually most of the analyses outlined in this manual will result in rasters Steps 1 Open the Vasconcellea species and the Latin America Countries layers Add the layer with the values for annual maximum temperatures in Latin America file Mean Basic elements of spatial analysis in DIVA GIS temperature Latin America 10 min grd This layer corresponds to a raster which is added to the legend panel in the same way as previously show
92. ber of observations increases 20 Click Apply v Regression loGrid in ec git tutorial 5 1 species diversityyasconcellea observations grd feGrid in e gis tutorial 5 1 species diversity vasconcellea diversity grd Function CD Linear e Natural Logarithm In e Result blar Dot size s vasconcellea observations 130 Show line 2 Yo vasconcellea diversity g EN Fieset Copy vasconcellea diversity D M w A A MM aS oo H0 a0 100 120 140 160 150 vasconcellea observations Spatial analysis of diversity for conservation planning The graph in the Chart tab confirms that there is a relationship between the number of samples taken and the number of species observed with a higher number of sample points more likely to result in a higher richness As the number of number of observa tions Sample points increases the graph begins to level off raising questions both about the applied sampling strategy and the cell size used in the analysis using larger cells and including more observations would be more appropriate Only in the instance where the user is familiar with the data and certain it originates from an intense and relatively homogeneous sampling could the results obtained above be considered accurate and therefore useful The results must be interpreted carefully if sampling was only partially conducted or fragmented or if the origin of the data is unknown In these cases the results may reflec
93. books This step may allow you to correct the mistake immediately If the original information is not available which is often be the case the next step is to identify the possible cause of the error refer to the example above for the point with ID 1669 where the only problem was the missing negative sign If these two steps are not possible you should consider eliminating the point Before doing so however consider the consequences of such an elimination on the results ensure that you are not creating bias You may want to keep the erroneous data if you are working at a very large scale and the problem point is considered to be important e g in a study to know the number of species per country the exact location within a country is less important 8 Now look at the other errors In the Check Coordinates menu go to the Points do not match relations tab The result of comparing passport data against its effective location on the map is displayed revealing the points where passport country information does not coincide with the country location of the point on the map na Check Coordinates Options Point outside all polygons Points do not match relations Point COUN Fola COUN ID SPEC me 50 1915 3 6833 Peru ECUADOR 805 W pariflora 3 6 on B3 Brasil COLOMBIA 1143 Vo microcarpe 8 3 BY 55 7333 2 6166 Brazil BRASIL ered Woo microcarpi 2 67 Highlight Pan To Zoom Ta l Expart Quality contr
94. ces Spatial analysis can contribute significantly to the call for the improved understanding and monitoring of biodiversity Results obtained from spatial analysis allow the formulation and implementation of more targeted and hence more effective conservation strategies Outputs from spatial studies can provide critical information on the diversity present in specific geographic areas and can be used for various purposes for example to evaluate the current conservation status of plant species and to prioritise areas for conservation Spatial information combined with available characterization and or evaluation data has also proven useful for effective genebank management e g definition of core collections identification of collection gaps etc This type of analysis is conducted using Geographic Information System GIS tools Guarino et al 2002 which allow one to carry out complex analyses combining different Spatial data sources and generate clear maps facilitating the uptake of outcomes by responsible authorities and encouraging the development and implementation of conservation policies In recent years technological advances and the growing availability of computers and GPS Global Positioning System receivers have led to the increased application of GIS analysis The general accessibility and use of the internet has also created a revolution in the sharing of biodiversity geographical and environmental data The Global Biodiversi
95. cess The DIVA GIS programme itself is only available in English Select Setup Language English 0K od Cane Installation of software and example data for analysis 4 The Welcome message will be displayed Click Next to continue i5 Setup DIVA GIS Welcome to the DIVA GIS Setup Wizard This will install DIVA GIS 7 3 on your computer It is recommended that you close all other applications before continuing Click Next to continue or Cancel to exit Setup 5 Carefully read the Terms of Agreement for the DIVA GIS software license If you agree select the box accept the agreement To continue click Next If you do not agree with the Licence Agreement you will not be able to install DIVA GIS and the installation will be aborted jB Setup DIVA GIS License Agreement Please read the Following important information before continuing Please read the Following License Agreement You must accept the terms of this agreement before continuing with the installation The software is provided by as is and any express or implied warranties including but not limited to the implied warranties of merchantability and Fitness m for a particular purpose are disclaimed In no event shall the copyright holder authors or distributors be liable For any direct indirect incidental special exemplary or consequential damages including but not limited to procurement of substitute goods or services Loss
96. cise DIVA GIS can help in making decisions regarding correction or elimination of presence points displaying quality problems The software has some useful tools to identify possible errors in coordinates based on the existing administrative unit information in the passport data see Section 4 1 or to identify suspicious presence points based on atypical environmental conditions which can indicate a taxonomic misidentification or erroneous coordinates see Section 4 2 Another important aspect of data quality and one which is difficult to evaluate is bias Bias generally occurs when a sample is not wholly representative of the area being studied This can be effectively remedied with a sound data collection strategy Nevertheless many spatial biodiversity analyses are made with some or all data originating from herbaria and genebanks Such data are usually not generated for the purpose of biogeographical studies and often entail ad hoc collecting non systematic sampling and uneven sampling efforts Chapman 2005a Frequently specimens have been collected from easily accessible areas or areas where a species is known to occur thus negatively affecting the representativeness of the data Hijmans et al 2000 These Chapter 4 issues can lead to a sample population which may or may not be representative for the species in terms of a n environmental or geographical space as the data provide information on patterns found only at the sampled
97. ct Data Layer Map Analysis Modeling Grid Stack Tools Help Dice oa f WAO Es 7 manihot ex situ o ge w PER _water_lines_dew PER_roads Peru_Towns oe ER_water_areas_dcew Explore the ex situ characterization values and identify those areas where cassava roots with the heaviest weight were found The parameter for average fresh root weight FRW will be used in this exercise Select the cassava characterization layer and go to Analysis Point to Grid Statistics Go to Define Grid File Options Cell Size and define a 10 minute raster cell size 0 1666 This raster will be used for the remaining analyses in this section In the Output Variable window select the Maximum option In the Parameters window select the parameter to be analyzed in this case FRW To finish assign a name to the generated file In the map amend the legend for improved visualization and interpretation v Point to grid Parameters Input Shapetile egis tuborial5 2 diversity phenotypic datamanihot ex situ Define Grid Create a new Grid Output Variable hatches Ores it orsdeecscnedl an 3 basinum o r E E S E Foint to Grid Procedure Simple Output L EGIS Tutorial 5 2 Diversity Phenotypic datacassay A Chapter 5 The result after changing the legend shows the location of zones of individuals with the highest root weight Cas
98. ction Chapter 6 between agricultural management practices and environmental factors Under specific agricultural practices a species can be cultivated in an environment outside its realized niche and even its fundamental niche if additional resources are available including water through irrigation or soil nutrients through fertilization Finally it is important not to overlook the relationship between genetic variation and the environment Populations of wild species and crops are capable of evolving locally adapting to site specific environments This results in differences in phenotypic expression between individuals of different populations even when they are planted together at a common site Genotypes x Environment GxE experiments enable one to identify the most appropriate ecotypes and varieties for a specific agroecological zone This type of analysis is relevant for selecting promising germplasm adapted to specific areas but will not be discussed here as it is beyond the scope of this manual This chapter focuses on the use of modelling to predict species distribution and how its results can contribute to the prioritization of sites for conservation climate change impact studies and species germplasm collection 6 1 Analysis of the realized niche of a species As mentioned in the introduction of this chapter several GIS software programmes including DIVA GIS include simplified species distribution models based on climat
99. cundinamarcensis 7 10 17 N 7 1714 75 45 47 WV 2816 V cundinamarcensis 790159 N 7 0331 7501859 2836 V cundinamarcensis 6 19 59 N 6 3331 75 1500 W 3030 V cundinamarcensis 6 5400 N 6 9000 75 57 59 W 3418 V cundinamarcensis 6909 59 N 6 1664 7503359 W Open the Values spreadsheet Copy the calculated coordinates in DD Columns H and O and paste the values using the Paste special option into the respective latitude and longitude columns Columns and P This is necessary as DIVA GIS cannot import new DD values if they are still presented as formulas Open the Final spreadsheet The coordinates in DMS are converted to DD and the presence point database is ready to be imported to DIVA GIS It is important to keep the original DMS coordinates in order to track any errors that may have occurred during the calculation of coordinates 38 Microsoft Excel Vcundinamercensis_DMSdata xls B9 File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF Al v C D E F G H latitude longitude COUNTRY ADM1 LATITUDE GPSLONGITUDE GPS 3433 V cundinamarcensis 6 9633 75 4175 Colombia Antioquia 6 57 48 N 7592503 VW 3037 W cundinamarcensis 75 7631 Colombia Antioquia 7 10 17 N TEASA VY 2816 V cundinamarcensis 75 3164 Colombia Antioquia 770159 N 75018 59 W 2836 V cundinamarcensis 75 2500 Colombia Antioquia 6 19 59 N 75 15 00 W 3030 V cundinamarcensis 75 9664 Colombia Antioquia 6 5400 N 75 57 5O VY 3418 V cundinamarce
100. d Thousands separator g T Use system separators Printing Right to left Default direction Bight to left Cursor movement Logical LeFt to right D Visual view current sheet right to left Show control characters 3 Make sure no commas are used in the characters in the column for species name s or in any additional columns if present This will help to avoid errors originating from unintentional separations in the information of the CSV file csv e g in the administrative unit information 4 Save file as CSV file csv Save in 2 1 Importing observation data y B xX Gy EJ Toos Name Size Type 2 amp L Vasconcellea csv 33KB Micro My Recent Documents 4 1 My Ne gt File name Yasconcellea csv v Save Places Save as type fae area aa Cancel A 5 In Excel open the CSV file csv to verify that coordinates have all the decimals required Chapter 2 6 Open Maxent under Samples indicate the CSV file csv with the georeferenced presence points To select the file a Go to the Browse option b Open the CSV file csv L Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples Environmental layers lo Browse DirectoryFile a b Linear features Create response curves Quadratic features Make pictures of predictions Do jackknife to measure variable importance Outp
101. d edu index shtml Global Land Cover 2000 http bioval jrc ec europa eu products glc2000 glic2000 php World Database on Protected Areas WDPA http www wdpa org World Wildlife Data and Tools http www worldwildlife org science data item1872 html Miscellaneous FAO s Geonetwork http www fao org geonetwork CGIAR Consortium for Spatial Information http csi cgiar org DIVA GIS http www diva gis org Data References Busby JR 1991 BIOCLIM a bioclimatic analysis and prediction system In Margules CR Austin MP editors Nature Conservation Cost Effective Biological Surveys and Data Analysis CSIRO Canberra pp 64 68 Hijmans RJ Cameron SE Parra JL Jones PG Jarvis A 2005 Very high resolution interpolated climate surfaces for global land areas International Journal of Climatology 25 1965 1978 Chapter 2 Basic elements of spatial analysis in DIVA GIS Chapter 3 Basic elements of spatial analysis in DIVA GIS This chapter illustrates how to use basic tools in the DIVA GIS programme to carry out common spatial analyses If you would like to learn more please consult the DIVA GIS Operating Manual available online at http www diva gis org documentation 3 1 Visualization in DIVA GIS Before starting the visualization processes it is important to know the five basic sections presented on the DIVA GIS work screen 1 Menu bar Facilitates access to all DIVA GIS commands 2 Toolbar
102. data There are an increasing number of organizations making spatial data publically available often within the context of a network The internet has made it possible to easily share and download data This section lists key online resources where environmental geographic and passport data can be retrieved Species presence points The Global Biodiversity Information Facility GBIF http www gpbif org All separate data providers to the GBIF can be found at the following site http data gbif org datasets Georeferencing and country level data Biogeomancer http classic biogeomancer org Geolocate http www museum tulane edu geolocate Geonames http www geonames org Google Geocoder http code google com apis maps index html Country level data http www diva gis org gdata Preparing and importing data to DIVA GIS and Maxent Taxonomic information Tropicos Missouri Botanical Garden http www tropicos org GRIN Taxonomy for Plants http www ars grin gov cgi bin npgs html index pl Environmental data Worldclim Current future and past climate data http www worldclim org World soil information ISRIC data http www isric org UK About Soils Soil data Digital elevation data STRM 90 meters http srtm csi cgiar org Downscaled GCM Data Portal to download future climate data http gisweb ciat cgiar org GCMPage Land cover Global Land Cover Facility http glcf umiacs um
103. dition to the efforts of the CBD the Global Plan of Action for the Conservation and Sustainable Utilization of Plant Genetic Resources for Food and Agriculture GPA adopted in 1996 and the International Treaty on Plant Genetic Resources for Food and Agriculture ITPGRFA FAO 2009 entered into force in 2004 were formulated to focus on the potential of agricultural biodiversity and its importance for agricultural production These international frameworks aim to increase information and actions to enhance the conservation and use of plant diversity For example Article 7 of the CBD calls for the identification and monitoring of biodiversity paying particular attention to those species and varieties offering the greatest potential for sustainable use and requiring urgent conservation measures CBD 2009b Similarly within the GPA Priority Action 1 calls for increased surveying and inventorying of plant genetic resources for food and agriculture while Priority Action 4 aims at promoting in situ conservation of crop wild relatives and wild plants for food production Further Priority Action 7 recommends planned and targeted collecting of plant genetic resources for food and agriculture The importance of these activities is further confirmed in Article 5 of the ITPGRFA In addition to such priorities each of these international frameworks emphasizes the need to strengthen local capacities to carry out research related to diversity and genetic resour
104. dominant species of natural pine forests in Southeast Asia and is of economic importance Observations in many countries suggest a broad distribution of this species however the available observations are dispersed For some countries only one or two presence points are available Species distribution modelling enables the identification of the potential range where this species may occur naturally van Zonneveld et al 2009a This analysis will model the potential distribution of P kesiya You will learn how to utilize Maxent to model the potential natural distribution of a species and to visualize the results generated by Maxent in DIVA GIS Species distribution modelling and analysis The observed distribution of P kesiya The map below shows the available presence points for P kesiya In some countries such as Laos and Myanmar in dark gray presence points for this species are scarce one and two presence points respectively in this dataset Maxent applies a predictive model based on these distribution points to identify the areas where this species could potentially occur After completing this analysis it is apparent that despite limited data points extensive areas within these countries have a high probability of P kesiya occurrence Using Maxent to model potential natural distribution of a species Steps 1 In Maxent enter the pkesiya csv file with the presence points in the Samples window Many files will be generat
105. e 3 1 5 How to create custom made climate layers Besides extracting site specific climate data from the information included in the CLM files clm DIVA GIS also allows one to generate parameter specific climate layers like those used in Analyses 3 1 2 3 1 3 and 3 1 4 These rasters are not only useful to gain a better understanding of the climatic conditions in the study area but can also be used in species distribution modelling given that all rasters have the same properties In the following analyses you will learn how to make climate layers in the ASCII file format asc for South Africa using the 2 5 minutes CLM files clm containing climate data linked to DIVA GIS in Chapter 2 Basic elements of spatial analysis in DIVA GIS Steps 1 Open World_adm0 in DIVA GIS and zoom in on South Africa Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Dee Qo M6m x sa el2 5 x 51 2142 y 136 2741 Scale 1 138204694 Le Diva GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Delia ea Aaa Hog x AmMeORS x 221672 y 19 5720 Scale 10057161 2 Go to Data Climate Map This opens the Climate Data to Map window Les DIVA GIS 7 3 0 Layer Map Analysis Modeling Grid Stack Tools Help Import Points to Shapefile gt m SA Baz E Import Text to Line Poly
106. e Result E GIS Tutorial 4 Gap analisis yvasec gap analysis grd FI Add to map Final result Gap analysis E i 1 oo O 1 awh _ OPN OA A WH OO oe oe ODN FD A on O E The map of observed diversity of the Vasconcellea species should look like the one presented above The intense red colour indicates areas where collecting has not yet been conducted even though these localities are suitable for many Vasconcellea species In fact in addition to the one or two species observed in these cells 12 or 13 other species Chapter 6 are expected to be found These locations occur in northern Peru and in the equatorial transition zone between the Andean and the Amazon regions Other potential distribution areas for collection are located in Colombia and Venezuela Certain cells indicated in blue have a value of negative one 1 meaning that more Vasconcellea species were observed in these areas than predicted by the potential diversity model This section outlines how to conduct a gap analysis for multiple species but a gap analysis can also be performed for a single species In that case you need to amp overlay the layer of observed and potential richness of a single species taking into account that the final map will only show the collection gaps and the zones where atypical observations were made References Anderson RP Lew D Peterson AT 2003 Evaluating p
107. e WN 23 Using Excel you can now create a scatter graph representing the different species the annual mean temperature range in the X axis and the annual precipitation range in the Y axis This allows you to visualize the differences between the species in a two dimensional niche space The graph generated in this analysis shows the climate ranges for V cundinamarcensis V microcarpa V quercifolia and V parviflora Chapter 6 24 Select the temperature values X values and precipitation values Y values for each species as a separate series a Selection of series in Excel 2003 Chart Wizard Step 2 of 4 Chart Source Data Data Range Climate niches of 4 Vasconcellea spp Annual Precipitation Mean Annual Temperature C Series Y cundinamarcensis Name cundinamarcensis Y microcarpa i v quercifolia x vales vasconcellea_bioclim C 11 Y parviflora i Y Values vasconcellea_bioclim D 11 b Selection of series in Excel 2007 Select Data Source Chart data range The data range is too complex to be displayed If a new range is selected it will replace all of the series in the Series panel a Legend Entries Series Horizontal Category Axis Labels JO rex Edit Series Series name V cundinamarcensis 15 8 y cundinamarcensis EM V cundinamarc V microcarpa 21 49583333 Series X values V q
108. e data from the presence points of individuals and or groups of individuals of a species If only a limited number of presence points is available for a given species the corresponding environmental data may not adequately represent the species realized niche This can result in the significant misinterpretation of the results Therefore it is important to ensure that the presence points used in models are of good quality see Chapter 4 and of sufficient quantity in order to obtain a sound species distribution model It should be noted that there is no standard in terms of the minimum number of points required as often this will relate to the nature of the species For rare species or species with a restricted niche only a small number of presence points may exist However in these cases even a small number of points may be highly representative of the niche As such strict guidelines on the minimum number of presence points necessary to undertake credible species distribution modelling cannot be provided However a few examples of the number of presence points used for specific species are available a studies conducted by Scheldeman et al 2007 used a minimum of 10 points for rare Vasconcellea species with a restricted distribution b the MAPFORGEN project which evaluated the natural distribution of 100 species native to Latin America used a minimum number of 20 species presence points as its threshold and c Van Zonneveld et al 20
109. e to combine the values of the smaller cells when merged to form a larger cell For this analysis select the Max option so the aggregated cell utilizes the maximum value of its composing cells If at least one of the composing cells has a value of one 1 the aggregated cell also includes a value of one 1 indicating that the species has been observed in that cell 16 In the Factor box indicate the extent to which you wish to enlarge the raster cell size In this example the 5 minute raster cell resolution will be converted to one 1 degree resolution To do this to the cells must be enlarged by a Factor of 12 in both directions since 60 minutes is equal to one 1 degree 17 Indicate in the Output box the file name for the new raster file 18 You have the option of adding a suffix to the names of aggregated rasters In this analysis use the default option _a 19 Keep the Expand and Ignore NO DATA options selected 20 Click OK to start the calculation process a n Aggregate ekind Max Ignore NO DATA o s Expand _ oFactor 12 O Truncate Chapter 6 21 After increasing the cell size of all rasters within the stack repeat the process using the Calculate option under the Stack menu Steps 7 12 above to determine how many Vasconcellea species are potentially present in each cell wa Stack Calculate M Input Stack amp gis tutorial 6 4 gap analysisvasconcellea thresholds_agg grs
110. e values used for this analysis are Min X 100 Max X 45 and Min Y 34 Max Y 21 These are the default values when the Vasconcellea layer is selected at the start of the analysis Another grid origin will generate a slightly different result 5 The cell size used in this analysis will be one 1 degree default value under Cell Size which is equivalent to 111 km at the equator 6 Click OK to accept the values Spatial analysis of diversity for conservation planning 7 Return to the Point to Grid window to select the type of analysis to be carried out Here we will undertake a species richness analysis a First select the Output Variable Number of different classes Richness b Click on the Parameters tab to select the units Point to Grid Input Shapefile l e gis tutorial 5 1 species diversity vasconcellea species sh ll Define Grid Je Create anew Grid v Option Output Variable F Grid Options E Richness hd Number of different classes Richness Minimum au 6 Point to Grid Procedure Maximum 45 Cell size 1 l gt Fix Comer Lower Right Output ma Columns 55 Adijuist With Rows Columns v Draw Rectangle Default Values Za 8 In the Parameters tab under the Field option indicate the parameter you wish to analyze For this analysis the parameter will be Species in order to analyze the species richness You are given the option to exclude specific s
111. e world of amazing Java content From business solutions to helpful utilities and entertainment Java makes your internet experience come to life Note No personal information is gathered as part of our install process Click here for more information on what we do collect Click Install to accept the license agreement and install Java now Installation of software and example data for analysis 1 2 2 How to install Maxent Maxent is an open source programme which can be downloaded at http www cs princeton edu schapire maxent The programme s executable file is downloaded directly so there is no need for separate installation Steps 1 The documents explaining the concepts used by Maxent can be found on the website listed above 2 Before downloading the programme provide the requested contact information 3 Finally accept the license agreement and download the programme Maxent software and datasets Windows Internet Explorer eJ Dhd httpwww cs princeton edut schapireimaxenti IJN File Edit View Favorites Tools Help J Favorites Maxent software and datasets yo v Pagey Safetly Toolsy Maxent software for species habitat modeling Most current version 3 3 3e see new features below Use this site to download software based on the maximum entropy approach for species habitat modeling This software takes as input a set of layers or environmental variables such as elevation prec
112. ea are those with an atypical environment not included within the limits of the realized niche Depending on the Output format selected see Step 5 select either the Cumulative or the Logistic threshold default output format is Logistic threshold For this analysis select the Logistic threshold see red arrow Cumulative threshold Logistic threshold Description Fractional predicted area Training omission rate 1 000 oos Fixed cumulative valued osz Cd 5 000 oo Fixed cumulative vaueS ozz tC i000 0o01 Fad camdaive vane 10 o0 ow 7 483 0 071 Minimum training presence 0 190 0 000 0 154 10 percentile training presence 0 102 0 083 0154 Equal training sensitivity and specificity oroz Std 0040 Balance training omission predicted area and threshold value 0 254 ooo 0 152 Equate entropy of thresholded and original distributions 0 104 0 083 Validation of the model s robustness The robustness of the model developed by Maxent can be validated using one of the methods available under the option Replicate run type in the Basic tab Validation of models is not dealt with in depth in this manual for more information consult Ara jo et al 2005 and Philips et al 2006 The robustness or transferability of the model is relevant when predicting potential distribution areas outside the observed distribution and when using different climate scenarios
113. easier to work with than cell sizes in metric distances Chapter 2 To start preparing a database with presence points it is necessary to take into account a Basic information Presence points must include basic passport data of an individual plant or of a group of individuals in a specific geographic unit They must include at least four elements an identification code ID the taxonomic name of the individual plant or the group of individuals and longitude and latitude coordinates These types of points are commonly used in spatial analysis of diversity and geographical distribution b Storing coordinates When geographic coordinates are used in DIVA GIS Maxent and other GIS and species distribution modelling programmes it is preferable that they be reported using a latitude longitude lat long coordinate system and presented in Decimal Degrees DD format DD DDDD When coordinates are available in Degrees Minutes and Seconds DMS format DD MM SS or Degrees Minutes DM format DD MM MM information should be converted to DD for use in a GIS or species distribution modelling programme Therefore the following formula is used for data conversion Decimal degrees Degrees Minutes 60 Seconds 3600 H H 1 when the coordinate is in the Eastern E or Northern N Hemisphere H 1 when the coordinate is in the Western W or Southern S Hemisphere Degrees Degrees Decima
114. ect Area pinus_kesiya_thresholded 11 Select the binary raster for future potential distribution areas under the First tab 12 Select the binary raster for current potential distribution areas under the Second tab 13 Under Operation select the Subtract option to subtract the values of the two rasters 14 Under the Result tab indicate the raster containing the results from subtracting the overlaid rasters 15 Select the Add to map option The new raster will then be automatically opened in DIVA GIS after clicking Apply 16 17 Species distribution modelling and analysis Click Apply to start calculations Grid Overlay Input files Te First EAGIS Tutorials6 3 Climate change pinus_kesiva_ gem _sea_ Smin_thresholded E GIS Tutorial 6 3 Climate change Pinus_kesipa_thresholded grd Operation C Add O Multiply D Minimum D Cover eS Substract Divide Maximum f Output file ite SResult_ E GIS Tutorials 6 3 Climate changespinus_kesiva_co_overlay grd After the raster has been opened in DIVA GIS the legend and labels can be modified To easily visualize the four situations it is recommended to modify the legend as illustrated in the following image ra Properties Label pinus kesiya _cc_overlay Filename amp gis tutorial 6 3 climate change pinus_kesiya_cc_overlap grid Manual High impact on pirn Edit values D Select rows Stays outside niche Classify
115. ed by environmental influences one option is to conduct the characterization outside the original collection site and under controlled uniform environmental conditions ex situ characterization be it in the same geographic location e g experimental fields or in a controlled environment e g greenhouse Although environmental effects may still play a role in ex situ characterization especially in experimental fields for example when accessions carry traits which make them better adapted to the chosen site s conditions characterization in experimental fields will be more relevant for comparison than those obtained from in situ characterization PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 5 2 Diversity Phenotypic data e Statistical program optional e Manihot ex situ shp shx dbf Cassava ex situ characterization data in Ucuyali e PER_Adm0 PER_Adm1 PER_Adm2 shp shx dbf Administrative data for Peru e Peru_towns shp shx dbf Municipalities in Peru e PER_roads shp shx dbf Data on roads in Peru e PER _water_areas_dcw PER_water_lines_dcw shp shx dbf Data on rivers and bodies of water Chapter 5 5 2 1 How to carry out a spatial diversity analysis using phenotypic data The following analysis outlines how to carry out a spatial analysis based on phenotypic data resulting from ex situ in the same experimental field morphological characteriza
116. ed when performing this analysis Carefully name and save these files so they can be easily located as they will be used frequently in this section Note 2 In Maxent select under Environmental layers the folder wclim_sea_2 5min to import the climate layers in ASCII format asc In this analysis the rasters have a 2 0 minute resolution which represents the value of the 19 Bioclim variables in the study area Southeast Asia See section 2 2 2 for further details about importing climate data in Maxent 3 Maxent generates a raster of potential distribution Different output file types can be selected under Output file type but the use of the ASCII file type asc is recommended 4 Inthe Output directory window select the location file path where the results of the modelling will be saved 5 We recommend unchecking the option Make pictures of predictions and maintain all other default options to assure that Maxent runs well For further information on these options please refer to the Maxent manual Philips 2009 6 Go to Settings tab to modify the parameters Chapter 6 Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples Environmental layers Browse Directory File 2 Potential distributiontwelim_sea_2 5min fk N N N R R Pinus_kesiya R R R N R Select all Deselect all Linear features Create response cur
117. ee This methodology allows one to maintain a high degree of resolution without losing the pattern of diversity Spatial analysis of diversity for conservation planning It can be challenging to define the optimal radius of the circle the best radius is often obtained by trial and error The example below illustrates that when a circle becomes too large on the right the results of the analysis are of little use having lost the level of detail needed to interpret the output Point to Grid Richness Analysis with Point to Grid Richness Analysis with a 10 minute cell size and circular a 10 minute cell size and circular neighbourhood of one 1 degree neighbourhood of three 3 degrees To observe the effect of the Circular Neighborhood option and the changes generated in the previous result the richness analysis will now be repeated using 10 minute cells with a circular neighbourhood of one 1 degree Steps 21 In the Point to Grid Procedure window select the Circular Neighborhood option 22 Use the parameters from the raster generated during the previous analysis raster cell size 0 1666 Under Circular Neighborhood Options enter one 1 as the Map Unit to indicate the desired size in the case of maps based on latitude longitude coordinates the map unit is one 1 degree Repeat the steps from the previous analysis Point to Grid Parameters Input Shapetile Age tutorial S 1 species diversitywasconcellea species shp
118. emperature 1 Size 3 Percentile 0 02500 C Show 22a 22b a Identify Climate by point se F7 77340 iy 8 81090 x FPO Y 8 81030 worldelimn_2 5rni WOaNdicans Rec 1 of Layer voandicans_outhers ID 1 SPECIES Error LATITUDE 8 8109 LONGITUDE 77 7734 COUNTRY Peru ADM1 Ancash Chapter 4 Under the Frequency tab you can simultaneously highlight outliers identified with the Reverse jackknife Outliers method and those identified through the 7 5IQR method Note that the Reverse jackknife method is not appropriate for smaller datasets as it is too rigorous and would significantly reduce the number of observations including correct data Note Individual Task Detect the erroneous points in the presence point data of Vasconcellea candicans using different climate variables to apply the 1 51QR method References Chapman AD 2005a Principles of Data Quality version 1 0 Report for the Global Biodiversity Information Facility Copenhagen Chapman AD 2005b Principles and Methods of Data Cleaning Primary Species and Species Occurrence Data version 1 0 Report for the Global Biodiversity Information Facility Copenhagen DIVA GIS 2005 User Manual version 5 2 on line Available from http www diva gis org docs DIVA IS5 manual pdf Date accessed October 2010 Hijmans RJ Garrett KA Huaman Z Zhang DP Schreuder M
119. en not the most suitable means to visualize results Through the Properties menu the classes can be adjusted This menu is accessed by double clicking on the layer Select the NoData Transparent box at the bottom of the window which will remove the black colour for those cells with no data Proceed to assign new classes at five degree intervals starting at the minimum temperature of 10 C until you reach the maximum temperature of 30 C This classification will result in eight classes a range of 40 degrees divided by 5 Therefore you will need to add three more rows in the default table The plus and minus buttons _ allow you to add or delete classes as required se Properties Label mean temperature latin america 10 min Filename sgis tutorial 3 1 basic elements mean temperature latin america 10 Legend Info History Color From To Label Auto complete 7 0 1 0 Edit values 10 80 C Select rows Classify 8 0 15 0 D 15 0 22 0 Select color scheme 22 0 30 0 pe Nodata No Data Ramp EJ Read From File Add or Remove Row J NoData Transparent e Check the following parameters before reorganizing the ranges a Select the commands Auto complete and Edit Values shown in the left column to allow the entering of values in the From and To columns b The values of the new classes must be entered in the boxes starting with the upper section of the table From 10 To
120. ence M Bias file Browse 3 4 5 6 Species distribution modelling and analysis Import the binary rasters of potential distribution areas to DIVA GIS as explained in Steps 13 19 of Analysis 6 2 1 using Import to Gridfile Multiple Files e r Import Multiple Files to Gridfiles Output Folder Same as input Select Type 6 IDRISI IMG or RST O Arc BINARY FLT Arc ASCII Save ag Integer O BIL BIF BSG Add fle File name E 4Gl5 Tutoral6 4 Gap analyses _helborni_thresholded asc E G l5 Tutorial 6 4 Gap analysis candicans_thresholded asc EGIS Tutorial 6 4 Gap analysis cauliflora_thresholded asc E 4G 5 Tutoral lt 6 4 Gap analisis _crassipetala_thresholded asc E 4Gl5 Tutoral6 4 Gap analvsis Y_ cundinamarcensis_thresholded asc E 4GI5 Tutorials6 4 Gap analais glandulosa thresholded asc al 15 files Make a stack of the binary rasters _thresholded grd by selecting the Make Stack option in the Stack menu Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Tools Help De WG Ge 1S fs f ca a9 aC Make Stack Plot Calculate Regression Cluster Export to Text File Overlap Indicate the rasters you wish to include in the stack For this analysis use the binary raster files of the Vasconcellea species _thresholded grd G AddGrid RemoveGrid Remove All File name Resol
121. ensis 7 01 59 N 75 1859 W Colombia Antioquia 2836 W cundinamarcensis 6 19 59 N 75 1500 W Colombia Antioquia 3030 V cundinamarcensis 6 5400 N 75 57 59 Colombia Antioquia 3418 v cundinamarcensis _6 09 59 N 75 33 59 VV Colombia Antioquia 2 Goto the Columns spreadsheet Create separate columns for degrees minutes and seconds of the latitude and longitude Columns E F G and L M N The Lat column Column D has been created to indicate whether the latitude coordinate is in the northern hemisphere N 1 or in the southern hemisphere S 1 The Lon column Column K has been created to indicate whether the longitude coordinate is in the eastern hemisphere E 1 or in the western hemisphere W 1 38 Microsoft Excel Vcundinamercensis_DMSdata xls a File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF Al v f ID C D E F G H LATITUDE Lat degrees minutes seconds decimal degrees latitude LONGITUDE Lon degrees minutes 1 75 3433 V cundinamarcensis 6 57 48 N 1 6 57 48 7525 03 W 3037 V cundinamarcensis 7 10 17 N TSAS AZ 2816 Y cundinamarcensis 790159 N 7501859 2836 V cundinamarcensis 6 19 59 N 7501500 3030 V cundinamarcensis 6 5400 N 7505759 A156 ndinamarcensis BA9 5G hy 7 G A Chapter 2 3 Open the DD formula spreadsheet The formula to convert latitude and longitude from DMS to DD D2 E2 F2 60 G2 3600 has been inserted in the latitude and longitude column
122. entified Among them locally common alleles are the most important from a conservation standpoint as they can indicate adaptation to local conditions and are due to their restricted distribution more vulnerable to losses than broadly distributed alleles For more information on different types of alleles refer to Frankel et al 1995 Frequency Frequent 5 10 Rare Distribution Local Local Rare Steps 1 Select the layer with the molecular marker data Go to by Class Les DIVA GIS 7 3 0 Project Data Layer Map ole 7 Latin America Countries SSR cherimoya rand column oe ssrs cherimoya rarefaction oO 18 20 DD 20 22 E 22 24 BB 24 26 HB 26 28 HB 22 30 saved in a text file txt Broad Analysis Distance Statistics Modeling Grid Stack Tools Help E Pointto Grid x sel 5 gt Point To Polygon Pointto Point Summarize Points a SHOW Statistics by Class Matrix geographic Distance Autocorrelation Lu Histogram L Regression JL Multiple Regression Select the SSRs to calculate geographical distances Assign a name to the dBase IV file dbf that will be generated Results can also be _s Distance Statistics Shape of point gis tutorial S 4 conservation strategies sar Dutput File GIS Tutorials 3 Diversity Mol Field Shape Units Decimal Degrees SSRS Values 5 F 5 5 y
123. ephnes M Donnelly P 2000 Inference of population structure using multilocus genotype data Genetics 155 945 959 on line Available from http pritch bsd uchicago edu publications structure pdf Date accessed October 2010 Rebelo AG Sigfried WG 1992 Where should nature reserves be located in the Cape Floristic Region South Africa Models for the spatial configuration of a reserve network aimed at maximizing the protection of diversity Conservation Biology 6 2 243 252 Scheldeman X Willemen L Coppens D eeckenbrugge G Romeijn Peeters E Restrepo MT Romero Motoche J Jimenez D Lobo M Medina Cl Reyes C Rodriguez D Ocampo JA Van Damme P Goetghebeur P 2007 Distribution diversity and environmental adaptation of highland papaya Vasconcellea spp in tropical and subtropical America Biodiversity and Conservation 16 6 1867 1884 Willemen L Scheldeman X Soto Cabellos V Salazar SR Guarino L 2007 Spatial patterns of diversity and genetic erosion of traditional cassava Manihot esculenta Crantz cultivation in the Peruvian Amazon an evaluation of socio economic and environmental indicators Genetic Resources and Crop Evolution 54 7 1599 1612 Species distribution modelling and analysis Chapter 6 Species distribution modelling and analysis Ecological niche is a theoretical concept frequently used in biodiversity conservation studies The concept has been successfully applied to prioritize sites for the in s
124. es Species Longitude and Latitude in this specific order 2 Microsoft Excel Vasconcellea xls Bl File Edit View Insert Format Tools Data Window GenAlex Help hi amp SPECIES W candicans 79 7005 V candicans 73 7916 Af candicans 79 9330 W candicans 9 6411 W candicans 9 6416 W candicans 79 6411 W candicans 79 7192 2 Before proceeding make sure the decimals are effectively separated by points a Go to Tools Options b Under the nternational tab select points to separate decimals c Under the nternational tab select commas to separate thousands d Click OK Microsoft Excel Vasconcellea xls File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF fe SPECIES Spelling FT Big Research Alt Click LONGITUD a Error Checking Speech Shared Workspace Share Workbook Track Changes OY candicans Protection e 2a 1 V candicans Online Collaboration a 2 V candicans 3 V candicans Goal Seek 4 V candicans Scenarios 5 V candicans 6 V candicans 7V cauliflora Macro 6 V cauliflora Add Ins 9 V cauliflora Formula Auditing AutoCorrect Options Customize bet r TUSSI UJIJI 75 6000 5 7500 76 4166 7 2333 Preparing and importing data to DIVA GIS and Maxent TE bu Calculation Custom Lists Chart International Error Checking _Speling Number handling re Decimal separator Mk
125. es you will also need to download sample data to conduct the analyses outlined in the remaining sections of the manual For each section a separate set of sample data is available these datasets will be placed in one single folder to facilitate access in DIVA GIS The datasets are available on Bioversity s website at http www bioversityinternational org training training materials GIS manual Most analyses in this manual are based on detailed climate data at a resolution of 2 5 minutes meaning that the user must download large climate data files This can be a constraint for users with poor internet connections For those with a good internet connection we suggest downloading the datasets in one zip file Unzipping the file will create a new folder where all data will already be organized separately for use with the manual For those with a slower internet connection data can be downloaded section by section Unzipping these different zip files will also organize the data as necessary for use with this manual Should you have difficulties downloading the data please do not hesitate to contact the manual s authors at bioversity colombia cgiar org x schelde gmail com or m vzonneveld gmail com Data of lower resolution climate data with less detail can be provided or if needed the data can be sent in DVD format The datasets provided are based on existing studies but have been altered to improve their applicability
126. es and b Climate data are selected The following information should display automatically when clicking on a line a Presence data Attributes b Climate data 15 The X axis of the graph shows the bioclimatic variables 1 Mean Annual Temperature 2 Mean Monthly Temperature Range etc cf table of Bioclimatic variables in Chapter 2 16 The Y axis of the graph shows the values of each bioclimatic variable for each presence point Chapter 4 17 The Copy option allows you to copy and paste the graph to any other type of document e Distribution Modeling 12 ae ale Reset Copy DhClick E Zoom X Attributes 7 Climate data diac Al ereeeee e ko E aod 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Bioclimatic Variable Min vars 3 J Annual Mean Temperature 1 M Mean Monthly Temperature Range 2 Z lsothermality 2 7 100 3 o J Temperature Seasonality STD 100 4 J Max Temperature of Warmest Month 5 J Min Temperature of Coldest Month 6 Temperature Annual Range 5 6 7 M Mean Temperature of wettest Quarter 8 geet oe pates P85 Bia song ras i nE E eee 14a ss Identify x ff 20000 3 06020 CUNdimarc Rec 1 of 1 Layer vcundimarcensis_ outliers ID 4 ESPECIES Error LATITUDE 3 0602 LONGITUDE 77 285 COUNTRY Colombia ADMI Cauca 8049790 i 2 03970 ycundimarc Rec 1of1 Layer veundi
127. ess _agg grd Chapter 6 28 In the Parameters menu go to the Field window and select the item for which you wish to run a diversity analysis In this case select SPECIES Pointto Grid 15 classes candicans cauliflora cras ipetala cundinamarcensis glandulosa qoudotiana microcarpa mongica parviflora pulchra quercifolia sphaerocarpa stipulata weberbaueri heilbornii Selection All Clear Irvert a LELLE cee eters 29 Click on Apply to initiate the calculating process After completing this process the observed diversity map of Vasconcellea species should look similar to the unedited version of the diversity map created in Section Oy ls Species distribution modelling and analysis Gap identification comparing potential and observed distribution After generating the potential and observed diversity rasters with the indicated parameters go to the Grid Overlay option and subtract the values of the two rasters similar to Steps 10 to 17 in Analysis 6 3 1 30 In the First window select the raster of potential diversity 31 In the Second window select the raster of observed diversity Grid Overlay 30mm Input files je First E GIS Tutonal 6 4 Gap analysis vasec pot nchness_agg grd le Second E T GIS T Utorial 6 4 Gap analysisyvasc obs ichness_agg grd Ee Operation O Add O Multiply D Minimum 9 Cover Substract S Divide O Maximum Output fil
128. ess according to raster 1 Use of different raster origin on same data Because of the low number of raster cells this effect is exaggerated in this example Results of a more complete sampling will be influenced to a lesser degree by changes in raster properties Richness according to raster 2 1 For more information on sampling bias refer to Hijmans et al 2000 Chapter 5 Visualization of species accumulation curve to assess possible sampling bias Steps 17 Aregression allows one to visualize the phenomenon outlined in the previous box In the Menu go to Analysis Regression Les DIVA GIS 7 3 0 Project Data Layer Map BRNIEINETEM Modeling Grid Stack Tools Help FAFS Pointto Grid gt elise l Dice o eon x a E Vasconcellea species Z Pointto Point Summarize Points Latin America Countries i Distance Autocorrelation vasconcellea observations CJ 1 10 o 10 25 Hi istogram E 25 50 Lu g i E 5o 100 Regression E 100 200 J Multiple Regression 18 Select the recently generated layers on species richness and on the number of observations 19 You can choose between a linear and naturally logarithmic regression A linear regression is more straightforward while a logarithmic regression is mathematically complex but may better represent a sample bias as it accounts for the typical levelling off of species accumulation curves as the sampling effort num
129. eviously been extensively distributed in the Andean zone of these countries Fjeldsa 2002 PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 6 4 Gap analysis e Maxent and Java e Vasconcellea csv e Vasconcellea species shp shx dbf e Folder wclim_ams_5min asc files 6 4 1 How to identify possible gaps in collections Areas of observed diversity of Vasconcellea species in Latin America were identified in Section 5 1 By using species distribution modelling in addition to this observed diversity a map of the potential diversity of these species can be generated Areas where it is likely to encounter a diversity of Vasconcellea species but where there are currently few or no records of observations can be identified by comparing maps of observed and potential diversity These gaps are areas of particular interest for germplasm collection missions Scheldeman et al 2007 The next analysis illustrates how to use DIVA GIS to create a map of potential diversity and to compare exiting gaps between a species observed and potential diversity In the previous sections rasters of potential distribution areas for the natural occurrence of a single species were explored Rasters of potential diversity can also be generated based on the realized niches of several species by using a stack of binary rasters with potential distribution areas for each individual species Chapter 6 In order f
130. f Sum AsPresent Absent Present p NULL as zero Calculate Area o Select Mask o Cut off a0 i EGIS Tutoria E 4 lap analysis asc pot richness agg grd hl Apply M Close Potential Vasc diversity Oo 00 bb iah G YA B a H i l i i D G j G a B a H amp After some modifications and editing the potential diversity map of the Vasconcellea species with the aggregated cells should resemble the one above It may seem inefficient to first generate the binary potential distribution rasters in Maxent using detailed climate date small cell size only to later aggregate these to a larger cell Species distribution modelling and analysis size rather than immediately generating the binary potential occurrence rasters using larger cells with climate data of lower resolution However it is recommended to use the two step process described as high resolution climate data usually generates more accurate predictions of potential distribution in Maxent than data of low resolution Developing the raster of observed richness of Vasconcellea species see also Section 5 1 22 23 24 25 26 27 Select the vector type file shp with the presence points of a species or group of species to analyze In this analysis the Vasconcellea species Vasconcellea species shp should be selected In the Analysis menu go to Poi
131. f diversity for conservation planning 5 1 Species richness Many diversity analyses focus on diversity at the species level As mentioned richness is the most straightforward method to evaluate alpha diversity This section outlines how to undertake this type of analysis PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 5 1 Species Diversity e Excel e Vasconcellea species shp shx dbf e Latin America countries shp shx dbf 5 1 1 How to carry out a spatial analysis of species richness The analysis below uses data from a diversity study of the Vasconcellea genus Scheldeman et al 2007 The genus Vasconcellea has 21 species all of which are related to common papaya Carica papaya and its natural populations are distributed throughout Latin America Due to an ability to adapt to high altitudes the species are sometimes known as highland papayas While some species are grown specifically for their fruit especially in the Andean region others are used as a source of genetic material for common papaya breeding programmes e g for specific traits such as tolerance to cold or resistance to the papaya ringspot virus PRSV p Further certain species are widely distributed V cundinamarcensis can be found from Costa Rica to Bolivia while others such as V palandensis have limited distribution areas and are in danger of extinction Conservation of the genus is important g
132. f these two variables but with values of one or more of the other 17 Bioclim variables values outside the range limits Chapter 6 19 The limits of the two dimensional niche are by default the 0 025 the lowest 2 5 of the points and 0 975 the highest 2 5 of the points percentiles meaning that 95 of the presence points have been taken into account in developing the two dimensional niche 5 are considered as outliers The niche width can be adjusted by changing the value of the percentile Depending on your interest you can narrow the niche to determine the environmental ranges of the core of the species distribution or enlarge the niche to determine the extreme values under which the species can still occur it is recommended to only do this after data quality has been checked as atypical values can significantly influence the ecological niche when all points percentile value O are included Js Distribution Modeling 118 observations with 106 89 8 in this envelope 83 70 3 overall 3400 3 200 3 000 2800 a 2 600 T2400 Q 22007 w a i 2 000 2 41 800 5 Z 1600 1 400 1 200 1 000 800 600 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Annual Mean Temperature 1 Environment Yariables A Annual Mean Temperature 1 T Annual Precipitation 12 0025 s ies 20 The main DIVA GIS window will display all points in yellow correspondin
133. for conducting analyses e g errors have been introduced or have been slightly modified to protect intellectual property in the case of unpublished data The datasets used in this manual should by no means be considered as appropriate for conducting specific studies Section 2 3 provides sources of spatial data that can be used for this purpose Preparing and importing data to DIVA GIS and Maxent Chapter 2 Preparing and importing data to DIVA GIS and Maxent The basis for spatial biodiversity analysis is observation data Observation data are snapshots of species trait or allele presence in time and space To analyze observation data users can use their own data publically available data provided through international platforms such as the GBIF http www gbif org or a combination of both Much of the data in the GBIF are historical observations from herbaria and genebanks which may not reflect the current presence of taxa due to recent ecological processes or human interventions such as forest conversion to agriculture and other changes in land use Nonetheless such data are still useful to gain insights into the ecological and genetic processes behind the geographical distribution of plant diversity Observation data must be organized following the format specified by the applied GIS software This chapter explains the type of data required for spatial analyses and species distribution modelling and how to prepare and format data
134. fore molecular markers are the measurements of choice to carry out an analysis of intra specific diversity Although molecular markers are for the most part not directly related to functional genes it can be anticipated that a high diversity based on molecular markers also indicates a high abundance of useful genes The following section outlines a basic spatial analysis of molecular marker data whereby microsatellites SSRs a widely used co dominant marker are utilized It should be noted though that any type of molecular marker data e g AFLPs can be used to conduct a diversity analysis provided a unique identity can be given to each variation in the DNA composition Spatial analysis of diversity for conservation planning PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 5 3 Diversity Molecular marker data e Excel e SSR cherimoya rand column shp shx dbf e Latin America Countries shp shx dbf 5 3 1 How to carry out a spatial diversity analysis using molecular marker data The data used in the following analysis is from the CHERLA project which included a major study on the diversity of cherimoya in its centre of origin for more information see www cherla com Given that the final results of this study have not yet been published only a subset of the data will be used with coordinates randomly modified to recreate a hypothetical scenario Principles of spatial
135. g to the green points in the Envelope analysis The points remaining in the original colour blue in this case correspond to points with a climate profile of which one or more of the values of the 19 Bioclim climatic variables are outside the range limits Species distribution modelling and analysis Comparison of the realized niches of different species in Excel DIVA GIS allows climate data for the corresponding presence points to be exported using the Extract values by points option see Chapter 3 Data can then be analysed in Excel spreadsheets for further visualization Individual Task Extract the bioclimatic climate variables from the worldclim_2 5m data based on Data Extract Values by Points From Climate Data for all presence points in the Vasconcellea_4species shp file See Analysis 3 1 4 Steps 21 The text txt file with the extracted climate data can be opened using Excel Each row represents a presence point Columns BIO1 BIO2 BIO19 correspond to the 19 Bioclim variables with the values corresponding to each presence point Microsoft Excel vasconcelela 4sp tt B File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF Al bd f ID go B c D E F G amp G H J ae P Q_ ID SPECIES LATITUDE LONGITUCCOUNTRYADM1 RecNo PointNo Lon_ext Lat_ext biol bio2 bio3 bio4 bio5 bio bio bio8 biog 3433 V cundina 6 9636 75 4177 Colombia Antioquia 75 4177 6 9636 15 8 10 31667 83 875
136. gon Draw Shape Polygon to Grid Points to Conyex Polygon aS selection to New Shapefile Extract Values by Points gt Ss Q Paint M Map Make CLM files Assign Coordinates W Check Coordinates Export Gridfile E Importto Gridfile Write YRT file Export Shapefile fi File Manager Download Chapter 3 3 In this window select the raster properties needed for the climate layer This can be done by a Selecting Read from Layer to use the characteristics from a raster selected in the legend of DIVA GIS This is useful when you wish to combine rasters and ensure they maintain the same properties b Selecting Draw rectangle to define an area by drawing a rectangle in the DIVA GIS window with the mouse c Manually entering defining minimum and maximum values for longitude X and latitude Y data In this example the last option is selected entering fixed values of 16 and 33 for X and 35 and 22 for Y a Climate Data to Map Climate database worldelin_2 5m Output Variable 5 an mae W Apply x Close 4 In addition to the definition of the raster you will need to indicate the climate layers to be mapped which will be given in the Output and Period Variable dropdown menus Options include basic temperature and precipitation data available at monthly intervals or the previously outlined bioclimatic variables see Chapter 2 As the
137. h the dry areas from a precipitation raster The original rasters contain information on both the selected and non selected zones which makes them of little use for such combinations Therefore new rasters of the selected zones need to be created to combine the information of interest Steps 1 For this analysis select all areas in Latin America with an average temperature of gt 20 C and precipitation lt 1000 mm Start with the temperature layer Mean temperature Latin America 10 min gra selecting all areas with averages greater than 20 C In the menu go to Grid Reclass DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Mlegml Stack Tools Help OS SACATE rere Ue Latin America Countries ae species Neighborhood Calculate mean temperature latin america 1d Aggregate Disaggregate Cut Merge New Transect Area 2 Change the values into classes make sure to cover the whole range from minimum to maximum value assigning a value of zero 0 to those cells with values less than 20 C and a value of one 1 to those equal to or greater than 20 C Assign a name to the new raster generated using the Output button The resulting display should only include cells selected with average temperatures equal to or greater than 20 C F Reclass Input egis tutoriala 1 basic elements mean temperature latin america 10 min grd Output E GlS Tutoriala 1 Basic elementsimean temperature latin a
138. hanges made to the database including the error reports and to keep the original database untouched making new and corrected versions for further processing Note The list generated by DIVA GIS in Step 7 reports five points with potential errors three of which were already detected ID 2967 319 and 1669 on the map The other two suspicious points are ID 1367 which is located 500 km away from the coast of Ecuador The level of Zoom used in the analysis did not allow for the point to be detected at first sight This point has a zero 0 latitude value as well which probably explains this error ID 796 which is located in the ocean about 1 km away from the Ecuadorian coast Considering that the point s passport data indicate a collection site very close to the coast the error is most likely due to a slight imprecision in the GPS or in the manual process of georeferencing the site This may be corrected by assigning coordinates located nearby at the coast of an island or mainland What to do with erroneous points There are three options to deal with errors correct delete or keep The most appropriate action depends on several factors Two important factors are the scale of the study global continental national or local and the number of points available for the studied taxon In either case the first step is to locate the original information and verify the origin of the point thus the importance of maintaining field
139. he place where the point is found Zoom to this option also moves the presence point line polygon to the centre of the displayed map but increases the zoom PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 3 1 Basic elements e Excel e Vasconcellea species shp shx dbf e Latin America countries shp shx dbf e COL_ADM1 shp shx dbf e Mean temperature Latin America 10 min grd gri e Precipitation Latin America 10min grd gri e World_adm0 shp shx dbf For the following analyses you need to have the 2 5 min worldclim climate data imported in DIVA cf Section 2 2 3 1 1 How to perform basic visualizations using vector files Steps 1 Open the Vasconcellea Vasconcellea species shp layer of points by using the na icon or the menu option Layer Add Layer This simple display of presence points does not provide much information it is therefore necessary to add more layers to complement these data points Lee DIVA GIS 7 3 0 Project Data ee Map Analysis Modeling Grid Stack Tools Help Dela ea QM gt da eRS 5a EOF Fs Vasoncellea species a x 66 8297 y 31 8887 Scale 1 42881984 Data Design 2 Add the layer which includes the countries of Latin America and the Caribbean file Latin America Countries shp The layer should display immediately Layer names are Chapter 3 shown in the left col
140. hological characters results in a stronger picture of the existing diversity Obviously this is a rather simplified diversity analysis Ideally a higher number of quantitative and qualitative traits should be taken into account To work with more variables especially qualitative variables multivariate statistics are required which are based on additional statistical software As this is beyond the scope of the manual please refer to Grum et al 2007 Kindt et al 2005 and Mathur et al 2008 for further information Data on phenotypic traits and or allele presence reflecting the phenotype and genotype respectively of an individual can be used as indicators for diversity at the different subunits of the study area as well as to compare directly alleles or indirectly traits genetic similarities between individuals These similarities can be used in multivariate analyses such as a cluster analysis To illustrate this practice the manihot ex situ shp file includes the results of a cluster analysis The variable Cluster shows the cluster to which each individual belongs based on similarities in traits This information enables one to conduct a diversity analysis like the one illustrated in Section 5 1 but using the number of clusters as a unit of diversity instead of different species In this case diversity is defined by the number of different clusters found in a site Individual Task Conduct a simple richness analysis based o
141. ibution the rasters of potential and observed richness of Vasconcellea species are compared To undertake the gap analysis you will need The raster of potential natural distribution areas for the species An observed distribution raster for the species with the same properties as the raster of potential distribution Raster cell size The importance of cell size in detecting spatial patterns has already been discussed in Section 5 1 The cell size is also highly relevant when identifying gaps A small cell size will serve to detect gaps at a local scale but when using small cells for a gap analysis in a large study area e g country or regional level only a limited number of the cells will contain presence points as most cells will not have been sampled This will obviously lead to many presumed gaps being identified complicating the prioritization of sites for additional collection For studies at a national or regional level a larger cell size is more suitable yet it will be up to the analyst to decide on the best cell size Chapter 6 Choice to appropriate cell size in gap analysis The figures below show a gap analysis for Vasconcellea microcarpa in Ecuador using different cell sizes They illustrate that rasters with a 10 minute cell size can be useful for a gap analysis at local scale e g province while rasters with a large cell size 1 degree are more suitable for a gap analysis at national level 1 degree gap analy
142. icans 79 candicans 79 candicans 79 candicans 79 candicans 79 candicans 72 candicans 79 candicans 77 candicans 76 candicans 76 View Help SPECIES LONGITUDE LATITUDE candicans 79 candicans 79 candicans 79 candicans F79 candicans 79 candicans 79 candicans 79 candicans 79 candicans 79 candicans 79 candicans F 2 candicans 79 candicans Fr candicans 76 candicans 76 7918 4 9330 4 411 4 a415 4 6411 4 floes 4 92245 4 r903 24 Toor d GB86 15 2130 331 7 0553 SOOO 11 35333 aloz 12 2500 4000 11 6500 wiew Help 7885 ood rols 4 9330 4 6411 4 416 4 6411 4 flo 4 92245 4 r903 4 YOO 4 S966 Li 21350 331 7 aS 5000 11 3333 9167 12 2500 4000 11 8500 cauliflora 74 7833 6 0000 Chapter 2 2 2 Importing climate data to DIVA GIS and Maxent The Bioclim Domain option in the Modeling menu in DIVA GIS allows one to carry out multiple analyses based on climate data These analyses include the identification of atypical points known as outliers see Chapter 4 the delineation of an ecological niche the prediction of potential species distribution and the subsequent gap analysis or analysis of climate change impacts See Chapter 6 Before starting any of these analyses you must import climate data to DIVA GIS Such data is freely avail
143. ick Browse Browse DIV4 GIS 7 3 8 Select the desired configuration for the DIVA GIS quick access icons jB Setup DIVA GIS Select Additional Tasks by Which additional tasks should be performed ie Sa ar Select the additional tasks you would like Setup to perform while installing DIVA GIS then click Next Additional icons Create a desktop icon Create a Quick Launch icon 9 A window will be displayed summarizing the file path and options you selected Review the information to ensure it is correct and proceed by clicking nstall jB Setup DIVA GIS Ready to Install Setup is now ready to begin installing DIVA GIS on your computer Click Install to continue with the installation or click Back if you want to review or change any settings Destination location C Program Files DIVA GIS 7 3 Start Menu folder DIVA GIS 7 3 Additional tasks Additional icons Create a desktop icon a Installation of software and example data for analysis 10 Installation will begin and the required files will be installed on your computer i5 Setup DIVA GIS Installing Please wait while Setup installs DIVA GIS on your computer Extracting files C Program Files Common Files esri MOLT 20 ocx eee Cancel 11 If all steps are completed correctly the installation process should run smoothly To finalize click Finish 6 Setup Diva GIS Completing the DIVA GI
144. ing a highland papaya study conducted in Latin America Scheldeman et al 2007 In view of the objectives of this analysis some errors have been introduced into the dataset The clean version of this dataset will be used in the inter specific diversity analysis of Section 5 1 1 Note that the Latin America Adm 01 layer contains some imprecisions at administrative level 1 These do not influence the analyses in this exercise but might prevent this file being used in other analyses For an up to date maps at Adm 01 level please visit the DIVA GIS website http www diva gis org Data or the GADM database of Global Administrative Areas http www gadm org Quality control Steps to use the check coordinates function in the data menu of DIVA GIS 1 The first way to identify potentially erroneous points is to visualize the dataset on a map Open the Vasconcellea final errors shp file to visualize the Vasconcellea collection points in Latin America To see all points on the map use the Zoom to Theme button which displays the full extent of all open datasets 2 Now add the polygon file with the Latin American countries Latin America countries shp Data errors are immediately obvious as some points are located outside the study area Latin America To access the passport data for each point and determine the ID use the Information button in the main menu For the purpose of this analysis the data presented illustrates three
145. ing the projection of vector files using DIVA GIS By selecting Jools Projection conversions between different projections such as latitude longitude and UTM can be made 3 1 4 How to extract values from rasters based on presence points data The previous steps detail how to display and manipulate vector and raster layers It will often be useful to combine vector and raster data based on a common geographical location In the exercise outlined below the precipitation value is extracted from the climate raster for each point in the Vasconcellea database vector file based on its geographical location Hide all previously generated grid layers Steps 1 Select the layer with the Vasconcellea points Vasconcellea species shp and go to the Data menu Select Extract Value by Points and then select From Grid or Stack A stack is a group of rasters with the same characteristics in resolution origin and size Les DIVA GIS 7 3 0 Project BElem Layer Map Analysis Modeling Grid Stack Tools Help are i b If Imal E Import Points to Shapefile m ec Import Text to Line Polygon Draw Shape Polygon to Grid Points to Convex Polygon Extract Values by Points F preci Climate E 7 From Climate Data i Assign Coordinates W Check Coordinates O E Export Gridfile m amp Importo Gridfile O write VRT file m Expor Shapefile mj f File Manager mear Download Basic elements of spatial ana
146. ings correspond to what is shown in the screenshot below Maximum Entropy Parameters Basic A Experimental L Random seed Give visual warnings Show tooltips Ask before overwriting 7 C Skip if output exists Remove duplicate presence records Write clamp grid when projecting Do MESS analysis when projecting Random test percentage o Regularization multiplier Max number of background points ooon Replicates Replicated run type Crossvalidate 7 Test sample file Browse 8 After modifying the Settings option return to the main window Click Run to start calculating the species potential area Running Maxent When you press the Run button a progress monitor appears which describes the steps being taken After the environmental layers are loaded and the initialization process is complete progress towards developing the model will be displayed Pinus_kesiya Gain is 2 261692 ee ee The gain is closely related to deviance a measure of goodness of fit The higher the gain the more discriminative the predicted distribution for species occurrence is in comparison to a random distribution For example if the gain is two 2 it indicates that the average likelihood of the presence samples is exp 2 7 4 times higher than that of a random background pixel For further information refer to the Maxent background paper Philips et al 2006 and the Maxent manual Philips 2009 Results fro
147. ional information which may not be possible to observe when using only two dimensions The relevance of using DIVA GIS is that the data of all 19 climatic variables can be extracted from the locations of presence points after which such data can be further analyzed using statistical programmes and software such as Genstat http www vsni co uk software genstat or R http www r project org 6 2 Modelling the potential distribution of a species Often the available set of presence data does not cover the entire range of a species natural distribution Species distribution modelling programmes such as Maxent Philips et al 2006 enable one to approximate the full distribution range These programmes are practical tools to identify those areas where a species is likely to occur The results of the species distribution modelling analysis can be used for different combined spatial analyses e g evaluating the impact of climate change on the distribution of species which will be discussed in the next section identifying collection areas explained in Section 6 4 or identifying suitable zones for crop and tree production as mentioned in the introduction of this chapter Species distribution modelling programmes identify sites with similar environments to those where a species has already been observed as potential occurrence areas The data required to identify these potential distribution areas include species presence points as well as
148. ipitation etc as well as a set of georeferenced occurrence locations and produces a model of the range of the given species Further description of this approach can be found in a o Steven J Phillips Miroslav Dudik Robert E Schapire A maximum entropy approach to species distribution modeling In Proceedings of the Twenty First International Conference on Machine Learning pages 655 662 2004 pdf o Steven J Phillips Robert P Anderson Robert E Schapire Maximum entropy modeling of species geographic distributions Ecological Modelling 190 231 259 2006 datasets used in this paper are available below pdf Terms of use This software may be freely downloaded and used for all educational and research activities This software may not be used for any commercial or for profit purposes The software is provided as is and does not come with any warranty or guarantee of any kind The software may not be further distributed By clicking on the download button below you agree to these terms ease provide your name institution and email address prior to downloading Name Institution Email Current version recommended 3 3 3e Older archived versions 3 3 3a 3 3 3 03 3 2 03 3 1 3 3 0 beta 3 2 19 03 2 1 03 1 0 O3 0 6 beta 3 0 4 beta 3 0 3 beta 3 0 2 beta 3 0 1 beta 3 0 beta 2 3 0 2 2 0 2 1 0 2 0 0 1 8 2 1 8 1 1 6 2 1 0 beta e Acceptterms and download Reset
149. ironmental layers When additional raster files grd are stored in the Directory File folder they are automatically included in the list of environmental layers This can generate errors or undesired results This may occur if ASCII files are created in DIVA GIS based on grd files See Chapter 3 and stored in the same folder as the original raster files grd By saving the relevant ASCII files for the analysis in a separate folder or de selecting all undesired raster files in the list of environmental layers in Maxent such errors can be avoided Maxent is capable of conducting an analysis for several species at once when presence points for each species are saved in the same CSV file csv This is explained in Section 6 4 The Settings option allows you to modify the conditions under which Maxent generates a potential distribution model A relevant parameter for this analysis is the Remove duplicate presence records option found under the Basic tab For further information about the Settings please refer to the Maxent manual Phillips 2009 Species distribution modelling and analysis 7 With the Remove duplicate presence records option duplicate presence points in one raster cell are removed from the analysis to reduce sampling bias which would favour the climatic conditions of those sites where sampling was highly concentrated For this analysis make sure that this option is selected checked and that the other basic sett
150. is analysis the prefix of the T min is tmin the prefix of T max is tmax and the prefix of Prec is prec Make sure the boxes to the right of the selected parameters are checked Thereis also an option to include data of radiation rad and potential evapotranspiration PET Such data are not necessary for the 19 bioclimatic variables and this data is currently not available at Worldclim Therefore they will not be used in this analysis Indicate the altitude alt layer Keep all other options as default Click OK to start the process Basic elements of spatial analysis in DIVA GIS a Climate Files File Browser File prefies e tmaxt2 grd Tmin tmin tmasz grd SE tman grd Tmas tmas E gt GIS Tutorial trae grd a 3 1 Basic elements tmaxs grd Prec prec triax6 grd tras grd Rad trnaxtt grd tman ord PET tmint grd i tmin _23 grd tmin 33 grd Suffix tron 0 grd tmin O23 grd fe tmint 0_33 ard Altitude mask alt Make Index Selected Files ES Progress 10 After having clicked OK the layers of the three datasets should be imported automatically in the first three columns of the selected files a Climate Files File Browser File prefines E tras 2rd ee ee tmax2 grd Tmin La i tmax3 grd Tmas tmax trad grd a 3 1 Basic elements tmax5 grd Frec prec tmas ord trae ord Rad trax grd tmax9 ord PET tmini grd tmin _23 grd f tmind 33 grd auiii
151. itu conservation of wild species and in planning collection missions for crop genetic resources e g Jarvis et al 2005 Scheldeman et al 2007 While different definitions for ecological niche have been formulated the concept basically refers to the environmental space a species occupies under natural conditions Puliam 2000 The generally accepted definition as provided by Hutchinson 1957 distinguishes between a fundamental niche and a realized niche A fundamental niche is the range of environmental conditions under which a species can theoretically exist whereas the realized niche is defined by the combination of negative interactions e g competition and predation that restrict a species presence and positive interactions e g facilitation that expand the environmental ranges in which a species is able to grow Geographic information systems GIS such as DIVA GIS include the ability to model ecological niches based on available environmental data from sites where species have been observed presence points Databases which provide detailed climatic data based on interpolations of data collected by climatic stations worldwide already exist such as Worldclim Hijmans et al 2005 but the availability of data for other relevant environmental factors such as soil variables is still limited Thus many GIS tools approximate the value of the ecological niche using climatic variables Known as the climate envelope Guarino et a
152. iven its potential for both fruit production and papaya breeding and such conservation efforts will greatly benefit from information on the distribution and diversity of the genus Before starting the analysis it is important to remember that a sufficient number of observations within a raster cell is necessary in order to undertake reliable diversity analyses The credibility and accuracy of the final result will depend on the quality of the sampling strategy although the choice of the cell size used for the analysis ideally the cell size should be defined while formulating the sampling strategy will also influence the quality of the final result If an analysis is run with cells that are too small the resulting raster will generate a high resolution map with results of limited value as each cell will most likely contain too few presence points often only one to detect a spatial pattern of species diversity On the contrary if raster cells are too large they will have a sufficient number of observations but the map will be of poor resolution complicating its interpretation and use The following analysis will use the number of species as the measured unit of diversity The analysis will be conducted at the regional level Latin America and the raster will use a cell of 1 degree x 1 degree 111 km x 111 km at the equator line see table in Chapter 2 Here you will learn to use the Analysis Point to Grid menu in DIVA GIS to conduct an in
153. l Decimal Longitude Minutes amp Latitude Minutes amp Degrees Degrees Seconds Seconds Eastern ace Northern Peyer 60 20 15 E 60 3375 24 00 45 N 24 0125 Hemisphere Hemisphere Western ee Southern E l 60 20 15 W 60 3375 24 00 45 S 24 0125 Hemisphere Hemisphere It is recommended that DD points have a precision of at least four decimals Conversion of DMS or DM data into DD format in Excel can be conducted by using the text functions DMS data are stored as text RIGHT MID LEFT Be careful not to generate a false sense of precision by creating more decimals when converting data from DMS to DD format A coordinate in DMS format including information on only the degrees and minutes e g 60 05 has an actual precision of two decimals but can be presented in DD format as a coordinate with four decimals 60 0833 When presence point data includes coordinates in DMS format with less than four decimals only degrees and minutes or only degrees it is recommended to add an extra field to specify the precision of the coordinates before converting these from DMS into DD format Note In addition to the latitude longitude coordinate system another common coordinate system is Universal Transverse Mercator UTM While DIVA GIS can operate using UTM the latitude longitude system in DD format is still preferred as it is more likely to be compatible with the available thematic layers administrati
154. l 2002 Analyses of the ecological niche of wild species and their genetic resources using presence points are applied under several conditions the most important being 1 The species should be in a state of equilibrium with its environment in other words the environmental ranges are restricted by competition and predation and not by dispersion limitations 2 Theavailable environmental variables e g climate variables used in the modelling are determinant abiotic factors in shaping the natural species distribution 3 No presence points should be included of specimens grown in plantations field collections and botanical gardens which may be located in an environment outside the realized niche of the species In practice one or more of these conditions above are often not met nonetheless species distribution modelling is still a useful tool to approximate the realized niche and the natural distribution of a species As such species distribution modelling is useful for prioritizing Conservation activities The ecological niche concept can also be used to identify agroecological zones ideal for growing specific crops and trees Ecocrop FAO 2007 which is included in DIVA GIS as well as Homologue Jones et al 2005 are examples of modelling programmes that can be utilized for such analyses The identification of optimum production zones is a more complex task than identifying the natural distribution area due to the dynamic intera
155. l for capacity building on the spatial analysis of biodiversity data The authors have developed a set of step by step instructions accompanied by a series of analyses based on free and publically available software DIVA GIS a GIS programme specifically designed to undertake spatial diversity analysis and Maxent a species distribution modelling programme The manual does not aim to illustrate the use of each individual DIVA GIS and Maxent command option but focuses on using GIS tools to help answer common questions relating to the spatial analysis of biodiversity data Throughout the manual the importance of proper sampling is stressed however it is beyond the scope of the document to elaborate on sampling theories The manual also does not discuss the statistical analysis of diversity data in detail instead when statistical methods and programmes are mentioned in the text the reader is referred to alternative reference materials for further information After following the instructions and completing the analyses outlined in this manual it is anticipated that the reader will have the capacity to carry out basic spatial diversity analyses to address common questions in conservation biology and plant genetic resources research References CDB 2009a Convention on Biological Diversity Global Strategy for Plant Conservation online Available from http www cbd int gspc Data accessed October 2010 CDB 2009b Convention on Biologic
156. l survive in 2050 protecting them from threats caused by human interference These predictions only deal with potential impact The models may overestimate the impact of climate change as species may possess the capacity to adapt to a range of climatic conditions Several tree species have a high degree of genetic variability and may be able to tolerate a broad range of climates Multi site trials conducted with pine species such as P kesiya have shown that the species adapts well to a broad range of climates and is also likely to adapt to new climatic conditions even though studies conducted using species distribution modelling predict future conditions to be inadequate van Zonneveld et al 2009b It should be noted that soil conditions competition predators and other factors also influence the presence of a species and represent additional limitations to the species current distribution and possible future displacements However since climate is considered to be the main driving force affecting distribution areas in the future models predicting the effects of climate change have not yet focused on or included these other factors In spite of their limitations envelope models are considered to be a useful tool in establishing an initial appreciation of the potential impact of climate change on the distribution of species Pearson and Dawson 2003 Species distribution modelling and analysis 6 4 Identification of gaps in collection
157. lea species shp file 2 Select Data Export Shapefile 5 DIVA GIS 7 3 0 Project REE Layer Map Analysis Modeling Grid Stack Tools Help De Import Points to Shapefile gt om HATEHA F or rs Import Text to Line Polygon Draw Shape Polygon to Grid Points to Convex Polygon X j S 4 1 Extract Values by Points a Climate Eig Assign Coordinates care 2 Check Coordinates Export Gridfile Importto Gridfile Write VRT file Export Shapefile A File Manager Download 3 Inthe Export Shapefile window under the Group by option select Species This will allow the visualization of a specific species in Google Earth The D option allows you to add text to each point Since the Vasconcellea species file has multiple points for this analysis it is recommended to select the option lt none gt 4 The Output button allows you to name the newly generated kmz file Export Shapefile Google Earth KMZ e gis tutorials 2 export to google earth vasoncellea species s i Output E GIS Tutorial 3 2 Export to Google Earth vasconcellea kmz Basic elements of spatial analysis in DIVA GIS 5 Go to the folder where the kmz file was saved and open the generated file 6 In the Places window you will find the folder Temporary Places containing the newly generated file Click on the triangle sign and notice that kmz file contains subgroups Select only the V quercifolia species
158. lecting the options available such as Red Green Blue By doing this the colder zones will display in red and the hotter in blue 7a The order of the colours can be inverted by clicking on the button with two arrows highlighted below 7b Label mean temperature latin america 10 min Label mean temperature latin america 10 min Filename gis tutorial 3 1 basic elements mean temperature latin america 10 Filename gis tutorial 3 1 basic elements mean temperature latin america 10 Legend Info History Color From Suto complete v Auto complete r Edit values R F Em Edit values Select rows Select rows Classify Classify v i v Select color scheme Add or Remove Row oo NoData Transparent NoData Transparent Chapter 3 8 You can also personalize the colours if preferred Double click on the first colour and choose the desired tone you can select either Basic colors or Custom colors Repeat the process for the colours in each class Basic colors el eee re ee i i EE EE Eee B ff fee im Custom colors Hue 160 Red Sat U Greer U Colors olid Lum 0 Blue 0 Addta Custom Colors 9 The Ramp button allows you to select only extreme tones this tool will automatically suggest the intermediate scale of tones For this analysis select the extreme tones pale blue and burgundy red 10 Some circumstances require legends to be
159. llustration the Indonesian Chapter 6 archipelago includes several areas with climatic conditions similar to the realized niche for P kesiya however these areas are not included in the species natural distribution as a result of the limitations noted above If desired it is possible to limit potential distribution areas using a fixed distance buffer area around each presence point see Willis et al 2003 This approach may adequately reflect the dispersal limitations due to a species reproduction system but will not solve the problem of geophysical and climatic barriers which can be unexpectedly close to represented presence points Alternatively one might consult the literature or contact experts on the species being studied and then compare this information with that provided by the potential distribution model Occasionally the literature will refer to areas where a species occurs naturally even though the model does not predict its occurrence in these locations In such instances the environment corresponding to the database of the presence points may not be completely representative for the climatic niche and the area is therefore not captured For example the GRIN Taxonomy for Plants USDA ARS National Genetic Resources Program 2009 reports the presence of P kesiya in Bhutan and the Chinese province of Xizang however the model generated by Maxent does not predict the occurrence of the species in these areas To reso
160. lora The section explains how to utilize the Modeling menu in DIVA GIS to analyze climatic niches and to compare the niches of different species using Excel Using the modeling menu in DIVA GIS to examine the realized niche of a species Steps After importing the Bioclim climatic variables to DIVA GIS see Chapter 2 the realized niche of a species can be identified based on the species presence points and corresponding climate data It is possible to display the realized niche in different ways using DIVA GIS by visualizing the frequencies of the different climate parameter ranges as histograms or by visualizing a two dimensional climatic niche based on two climate parameters These options for analysis are available in the Modeling menu in DIVA GIS 1 Open the Vasconcellea_4species shp file in DIVA GIS and keep the file as the selected layer 2 Goto Modeling Bioclim Domain to open the Distribution modeling window Les DIVA GIS 7 3 0 Project Data Layer Map Analysis BQnRENIM Grid Stack Tools Help D AA E Bioclim Domain Osa aq ai 7 vasconcellea_4species Evaluation O EcoCrop Terrain Modeling 3 In the Distribution modeling window under the Input tab go to DIVA climate data and select the climate database to be used This analysis will use the Worldclim climate data at a 2 5 minute resolution file wordclim_2 5min This was the file imported to DIVA GIS in Chapter 2 4 Select the Many
161. lve this discrepancy a more detailed study and or contacting local experts is required Thus it should be noted that species distribution modelling scientific literature and expert data provide complementary information The combination of these information sources will help to provide a complete picture of the natural distribution of a given species Comparison between potential natural distribution of P kesiya according to Maxent and the natural distribution according to literature The natural distribution of P kesiya according GRIN Taxonomy for Plants USDA ARS National Genetic Resources Program 2009 is defined by administrative boundaries highlighted on the map in dark gray Countries included in the natural distribution are China Xizang Yunnan Bhutan India Laos Myanmar Thailand Vietnam and the Philippines Luzon Species distribution modelling and analysis 6 3 Modelling the impact of climate change on species distribution Global climate change is ever more evident IPCC 2007 Consequently geographic areas corresponding to biomes ecosystems and species ecological niches are changing which is likely to affect the natural distribution of many species Species distribution modelling can be used to provide a rapid evaluation of the potential impact of climate change on the distribution of ecosystems and the species that inhabit them The process consists of detecting changes in species distribution by com
162. lysis in DIVA GIS 2 Select the Grid option and mark the raster from which you want to extract the data this exercise uses the raster file with the precipitation data Precipitation Latin America 10min grd Leave the default values Single File Lines and insert a name for the Output File which will be a text txt file You need to use the vector file s identification D fields since these are key for combining the generated text file with the data in the observation database xls or dbf file Extract point values from gridfstack Points shapefile amp sale tutorial 3 1 basic elements wasoncellea species shp Grid Stack eis tutoriala basic elements precipitation latin america 10min grd Grid Stack Output EGIS Tutoriala 1 Basic elements vasconcellea prec data tt Single file Values seperated by Lines Tabs Include field m ID fio z 3 When you open the resulting text file in Excel you will notice that the IDs are combined with the extracted raster values It is now possible to use the information from both sources e g for ecological niche analyses This is further explained in Section 6 1 38 Microsoft Excel vasconcellea prec data tt el File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF fe grid precipitation latin america 10min T precipitation latin america 10min precipitation latin america 10min precipitation latin america 10min precipitation
163. m Maxent Results are saved in the folder selected under the Output Directory One of the saved files is an HTML document which summarizes all results The analysis of the most important parameters is briefly described below For more information on the analysis see Anderson 2003 and Phillips et al 2006 Chapter 6 9 Open the HTML document Pinus_kesiya in your internet browser e g Firefox Chrome or Internet Explorer 10 The omission rate is a statistic indicating model performance The Omission and Name plots 2welim_sea_2 5min maxent log E maxentResults csv E Pinus_kesiya asc El Pinus_kesiya csv pinus_kesiya dbf Pinus_kesiya html E Pinus_kesiya lambdas pinus_kesiya shp E pinus_kesiya shx fS 1Pinus_kesiya_explain bat El Pinus_kesiya_omission csv E Pinus_kesiya_sampleaver El Pinus_kesiya_samplePre sj seacountries dbf a seacountries shp seacountries shx welim_sea_2 5min zip 12 KB 4 KB 10 278 KB 2 KB 245 KB 3 707 KB 21 KB 8 509 KB Type File Folder File Folder Text Document Microsoft Office Ex ASC File Microsoft Office Ex DBF File Chrome HTML Do LAMBDAS File ESRI Shapefile SHX File MS DOS Batch File Microsoft Office Ex Microsoft Office Ex Microsoft Office Ex DBF File ESRI Shapefile SHX File WinRAR ZIP archi Predicted Area plot consists of three lines Date Modified 08 10 2010 18 17 08 10 2010 18 16 08 10 2010 18
164. marcensiz_outliers ID 3 ESPECIES Error LATITUDE 2 0397 LONGITUDE 80 4979 COUNTRY Ecuador ADMI Guapas Individual Task Find all erroneous points in the Vasconcellea cundinamarcensis presence point database the points indicated as errors Presence points of small datasets Presence points of small datasets can be tested to determine if they are atypical by using the 7 5 QR method available under the Frequency option 18 To identify atypical points in the V candicans dataset using the 1 5 IQR method start by following Steps 1 to 6 above using the new dataset Quality control 19 Define the climate variable based on which you will apply the 1 5 IQR outlier method In this analysis select Annual Mean Temperature 20 Select the 1 5 IQR window to establish the limits of the range 21 Atypical points are those falling outside the 7 5 QR limits defined by the two outer lines 22 When the Climate data and Attributes boxes are checked the following information should be displayed automatically when clicking on a dot a Presence data Attributes b Climate data Distribution Modeling Frequency Outliers Histogram Envelope Predict ala Reset Copy On Click Zoom Canter Wicinte ad 13 valid non duplicate observations out of 17 Ca a gt te D 5 gt D u D T 3 E 3 O 45 6 7 8 9 1011121314 1516 1718 19 20 21 22 23 24 25 26 27 Annual Mean T
165. merica 10 min_rec grd Gid Stack New value x O Data Type Real Minimum 56 4625000 Masimum 29 170833 oK X Chose Chapter 3 DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Cee R QQ lea WS La mean temperature latin america 10 E ooo o E 0 200 0 E 0 400 0 E 0600 0 E 0 200 E No Data v Latin America Countries v Vasoncellea species mean temperature latin america 10 E Input e gis tutorial3 1 basic elements mean temperature latin america 10 min grd Qutput amp gis tutorial 3 1 basic elements mean temperature latin america 10 min_rec grd Grid Stack To New value x Oo 20 0 30 1 Data Type Real Minimum 56 4625000 Maximum 29 170833 Save RCL Read RCL SA OK X Close Y 32 4063 Scale 1 42881984 mean temperature latin america 10 min 18 4500007 Data Design 3 Add the precipitation raster Precipitation Latin America 10min grd By using the Reclass option you can select the cells where precipitation is below 1000 mm see Step 1 Les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Cee 82 QQ lela LK La v precipitation latin america 10min_r Hl No Data precipitation latin america 10min E o 1984 E 1924 3967 DD 3967 5951 GB 5951 7934 HB 7934 9918 mean temperature latin america 10 E 0 000
166. n Use the UPGMA method default to group the cells Click OK to continue again it is recommended to use the same file name P GeoCluster Input Distance Matrix egis tutorial 5 3 diversity molecular marker data ssrs cher presabs dmt m UPGMA Weighted Lookup Distance Cell Cell Distance 0 0 o Lookup Output Cluster File e gis tukoa 5 3 diversity molecular marker data sers cher presabs dtr 6 Under the Draw tab click on the Draw button A dendrogram tree will be displayed showing similar cells i e those with a similar allelic composition GeoCluster Cluster File E esabs dt Show A width l Depth l l i Set Prune Dist Labels V Aligned On Click Distance 0 08 Chapter 5 7 Cells can be grouped according to the level of similarity as indicated by small distances This can be done by pruning the dendrogram Click on the dendrogram at the distance value where you wish to group the cells and then click the Prune button for example at a distance of 0 1 Alternatively you can enter the desired value for clustering the cells in the in the Distance window 8 Click on GeoCluster Distance Matrix Cluster Cluster File gis tutorial 5 3 diversity molecular marker data ssrs cher presabs dtr Reset Copy Show i Distance Width 1 Depth z f Draw f Prune Prune 01 e V Set Prune Dist T Labels V Aligned OnClick Distance 0 08 0
167. n layers directory file window import the climatic variables under future conditions gcm_sea_2 5min to the database Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples Environmental layers File otential distribution Pinus_kesiya csv Browse Directory File jial 6 3 Climate changetwelim_sea_2 5rnin bio_1 bio_10 bio_11 bio_12 bio_13 bio_14 bio_15 bio_16 bio_17 bio_18 bio_19 Select all Deselect all 7 Linear features Create response curves _ KN N N R RN Pinus_kesiya R E K E EJ R 7 Quadratic features Make pictures of predictions _ Do jackknife to measure variable importance _ Output format Vv Product features W Threshold features Output file type ee lemures Output directory E GIS Tutoriah6 3 Climate change Auto features Projection layers directoryffile 31 6 3 Climate change gem_sea_2 Smin Browse Settings In order for Maxent to process the environmental raster files of both current and future conditions in this case wclim_2 5min and gcm_sea_2 5min all files must have the same parameters in terms of raster properties resolution and raster corners or vertices Note Apply threshold rule to create binary rasters Maxent can also generate binary presence 1 and absence 0 rasters of potential distribution areas This format is useful when layers of potential species distribution are comp
168. n the results of the cluster analysis variable Cluster with the same raster used in previous analyses Chapter 5 There will be differences between the results of analyses based on the coefficient of variation and those resulting from multivariate analyses The cluster analysis uses a more complex multivariate statistical methodology and a greater number of variables Willemen et al 2007 Nonetheless the diversity tendencies revealed in the results of the two types of analyses are the same low diversity exists around the city of Pucallpa bordering the river where the road to Lima begins and a very high level of diversity is present in the central zone close to the road to Lima lower left corner of the map The cell with the highest diversity is the same in both analyses As discussed in Section 5 1 the circular neighbourhood technique may also be used for such analyses 5 3 Intra specific diversity analysis based on molecular marker data As noted above phenotypic morphological data can be used to conduct intra specific diversity analyses however the influence of the environment where the characterization was conducted will always affect the results Allelic composition or gene sequences of plant individuals are not influenced by such environmental factors a change in environmental conditions e g wet vs dry year does not alter DNA base pair composition but will alter phenotypic appearance e g leaf size or growth There
169. n with a vector file ks DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help De aao Qa G mean temperature latin america 1 E 7 0 10 E 0 20 O 80 150 E 15 0 22 0 E 22 0 30 0 BB o Data v Latin America Countries Vasoncellea species e x 91 6751 y 32 3028 Scale 1 42881984 Data 4 Design Notice the new layer covers the previous ones i e country borders and observation points The order in which the layers are listed in the legend column corresponds to the order in which they are displayed Superimposed To modify the order reorganize the layers by dragging them into the desired position To continue relocate the temperature layer in the bottom position in order to allow the other layers countries and points for species occurrence to be visualized on top Ls DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Ded aao QQ lt gt a x Lh v Latin America Countries v Vasoncellea species o v mean temperature latin america 1 E 7 0 10 E 0 80 oO 8 0 15 0 Bi 150 220 ee E 220 300 HB o Data x 105 1329 y 29 3007 Scale 1 42881984 mean temperature latin america 10 min 19 4583339 Data A Design Chapter 3 6 The different classes of temperature are expressed as ranges this information is displayed in the legend As a default DIVA GIS uses five classes with equal ranges which are oft
170. nce of different species distribution modelling methods Ecography 29 773 785 Hijmans RJ Cameron SE Parra JL Jones PG Jarvis A 2005 Very high resolution interpolated climate surfaces for global land areas International Journal of Climatology 25 1965 1978 Hutchinson GE 1957 Concluding remarks Cold Spring Harbor Symposia on Quantitative Biology 22 415 427 Species distribution modelling and analysis IPCC 2007 Climate Change 2007 Synthesis Report Cambridge University Press New York USA Jarvis A Williams K Williams BD Guarino L Caballero PJ Mottram G 2005 Use of GIS for optimizing a collecting mission for a rare wild pepper Capsicum flexuosum Sendtn in Paraguay Genetic Resources and Crop Evolution 52 671 682 Jones PG Diaz W Cock JH 2005 Homologue A computer system for identifying similar environments throughout the tropical World Version Beta a 0 Centro Internacional de Agricultura Tropical CIAT Cali Colombia Liu C Berry PM Dawson TP Pearson RG 2005 Selecting thresholds of occurrence in the prediction of species distributions Ecography 28 385 393 Pearson RG Dawson TP 2003 Predicting the impacts of climate change on the distribution of species are bioclimate envelope models useful Global Ecology and Biogeography 12 361 371 Phillips SJ Anderson RP Schapire RE 2006 Maximum entropy modeling of species geographic distributions Ecological Modeling 190 231 259 Phillip
171. nd the International Network for the Improvement of Banana and Plantain INIBAP Supported by the CGIAR Citation Scheldeman Xavier and van Zonneveld Maarten 2010 Training Manual on Spatial Analysis of Plant Diversity and Distribution Bioversity International Rome Italy ISBN 978 92 9043 880 9 Cover Credits Photographs Xavier Scheldeman Bioversity International Maps Maarten van Zonneveld Bioversity International DNA gel electrophoresis images I aki Hormaza IHSM la Mayora Consejo Superior de Investigaciones Cientificas Spain Bioversity International 2010 Contents ACKNOWLEDGEMENTS INTRODUCTION SECTION A BASIC ELEMENTS AND DATA PREPARATION 1 INSTALLATION OF SOFTWARE AND EXAMPLE DATA FOR ANALYSIS 1 1 INSTALLATION oF DIVA GIS 1 1 1 How to install DIVA GIS 1 2 INSTALLATION OF MAXENT 1 2 1 How to install Java 1 2 2 How to install Maxent 1 3 INSTALLATION OF GOOGLE EARTH 1 3 1 How to install Google Earth 1 4 DATA FOR ANALYSIS 2 PREPARING AND IMPORTING DATA TO DIVA GIS AND MAXENT 2 1 PREPARING AND IMPORTING PRESENCE POINTS 2 1 1 How to convert DMS data into DD format 2 1 2 How to import georeferenced presence points to DIVA GIS 2 1 3 How to import georeferenced presence points in Maxent 2 2 IMPORTING CLIMATE DATA TO DIVA GIS AND MAXENT 2 2 1 How to import climate data to DIVA GIS 2 2 2 How to import climate data in Maxent 2 3 SOURCES OF SPATIAL AND OTHER RELEVA
172. nd to a subset of information used in the study conducted by Scheldeman et al 2007 and are therefore slightly different from the actual results given in the paper In the complete study rare species endemic to southern Ecuador are also included thus contributing to high levels of diversity in this zone 10 11 12 13 Individual Task Use DIVA GIS tools to display a meaningful legend of species richness see Analysis 3 1 2 To undertake this task you will need to know the value of each cell which can be determined by clicking on the Information button or using the arrow together with the information shown on the status bar of the map see Chapter 3 The final result should be a map similar to the one below Check the number of observations in each cell To compare results of the previous and current analyses it is important to use the same raster definitions see Chapter 3 Go to the Analysis Point to Grid Richness see Step 1 Under the Define Grid option select Use parameters from another grid Click on the Options button to select the raster from which you wish to use the parameters the one created in the steps above Using the nput button select the richness raster file grd created in the steps above click on OK Chapter 5 14 Return to the Point to Grid window and select Number of Observations in the second window of the Output variable Point to Grid Input Shapetile Agis tutorial 5
173. nd useful when formulating conservation strategies Cluster analysis Richness analyses carried out thus far take into account the number of alleles in a raster cell alpha diversity but they disregard the variations in composition among the different cells beta diversity While two cells may have a similar richness they may also display a completely different composition of alleles How can diversity be analysed in this case The DIVA GIS programme includes a cluster analysis tool to assess differences in diversity between raster cells In the analysis outlined below this tool is applied in order to analyze differences in allelic composition thereby providing further insight in the genetic structure existing across the species geographical distribution Several software programmes have been developed to carry out the analysis of allelic composition in different populations One such programme is Structure http pritch bsd uchicago edu structure html which assigns genotypes to groups based solely on allelic frequencies independent of the subunit of the study area raster cells in which they are Spatial analysis of diversity for conservation planning located or any other a priori definition of populations For more information about Structure please refer to Pritchard et al 2000 The following analysis however only explains how to undertake a cluster analysis in DIVA GIS using alleles as the observed unit of diversity
174. nks to other relevant publications and websites Chapter 1 1 1 1 How to install DIVA GIS Steps 1 Download the compressed installer from the following URL http www diva gis org download a Download DIVA OIS Windows internet Explorer G X EJ MRD www diva gs Org downioed Fie Edt View Favorites Tools Help We Favontes 2 Download DIVA GS t an account or log in DIVA GIS Comments Are you willing or able to projections to utm Ider versions mpor and save legend Pronde good gratcule value labels in map and design As far as I can tell this is 1 have now managed to do thes Yes I am referring to the Perhaps you can first expla 2 Save the file to your hard disk File Download Do you want to open or save this file e Name diva 30 zip Type WinRAR ZIP archive 3 71MB From WWW diva gis org While files from the Internet can be useful some files can potentially harm your computer If you do not trust the source do not open or B _ x save this file what s the risk 3 To execute the DIVA GIS installer first decompress i e un zip the file saved in Step 2 above To decompress the file use free software such as 7 zip http www 7 zip org or Izarc http www izarc org After decompressing the file you will be able to view the setup exe file Click on setup exe The Select Setup Language window will then be displayed select the language for the installation pro
175. ns et al 2000 With the use of more complex analytical methods this error may be remedied for example by using rarefaction which recalculates the diversity measured at each subunit of the study area to a standardized identical number of samples Petit et al 1998 Leberg 2002 see Section 5 3 Still the richness of a specific site might be difficult to assess using this alternative especially when a limited number of observations are available This is often the case when a high resolution raster e g with cells of 1 km or 5 km is used for a spatial diversity analysis at a large scale for example the analyses described in Sections 5 1 5 3 and 5 4 In this instance it is impossible that each of the raster cells often in the thousands possess a high number of observations Chapter 6 discusses how to apply species distribution modelling to address the issue of incomplete sampling or sampling bias Another disadvantage of measuring richness is that this methodology does not consider the relative proportions of the number of observed units of diversity For example at a site where 150 observations are recorded for a total of three species there may be 50 observations for each species or 148 observations for one species and only one observation for each of the remaining two The first situation would clearly be more diverse than the second Several indices such as the Shannon and Simpson indices have been developed to assess diversity taking
176. nsis 75 5664 Colombia Antioquia 609 59 N 75 03359 W Preparing and importing data to DIVA GIS and Maxent Georeferencing presence information Sometimes presence points lack geographic coordinates and have only a description of their location in the form of administrative unit data These data can be classified as follows Country Administrative unit level 1 Adm1 state department region province of a country Administrative unit level 2 Adm2 province canton municipality Locality city town national park etc ID Taxon Country Adm1 Adm2 Locality 1 Capsicum chinense Jacq Bolivia Pando Nicolas Suarez Cobija 2 Capsicum chinense Jacq Bolivia Pando Nicolas Suarez Cobija 55 Capsicum frutescens L Bolivia Beni Vaca Diez Riberalta 72 Capsicum eximium Hunz Bolivia Tarija Mendez San Lorenzo 73 Capsicum eximium Hunz Bolivia Tarija Mendez San Lorenzo Example of administrative unit data with Capsicum accessions originating from Bolivia In such cases databases known as gazetteers can be referenced for assistance Gazetteers are lists of administrative units e g municipalities with respective geographic coordinates that can assist in assigning georeferenced information to the points of interest Gazetteers with administrative unit data for most countries are available online and are freely accessible from the DIVA GIS web page http www diva gis org gdata The files are generally in dBase IV forma
177. nt to Grid Richness ks DIVA GIS 7 3 0 Project Data Layer Map EAEVSES Modeling Grid Stack Tools Help AMAF gt Point To Polygon Estimators of Richness Pointto Point Turnover Summarize Points Diversity q Molecular Distance and Diversity Si Distance Reserve Selection Autocorrelation Statistics Lu Histogram L Regression J Multiple Regression Under Define Grid window select the Use parameters from another grid option to ensure the parameters of the raster with the observed diversity the one now being created are equivalent to those of the raster with potential diversity To do this under Options select the raster of potential richness created during the previous steps Select Richness and Number of different classes Richness as the Output Variable Select the Simple option in the Point to Grid Procedure box Select the button to the left of the Output box to indicate the name and location of the resulting raster Point to Grid j Farameters Input Shapefile Se kis tutorial 6 4 gap analysis wasoncellea shp Define Grid Use parameters from another grid Options r Output Variable Richness ar Number of different classes Richness Pa Nas Point to Grid Procedure roa f Simple a Output T J E SGlS Tutorial 6 4 Gap analysis wase obs nichness_a A coss Select parameters from another grid EGIS Tutoria 6 4 Gap analysis vase pot richn
178. ntify possible erroneous presence points using different tools and how to take corrective actions to ensure high levels of data quality Poor data quality can result from various causes such as errors in site descriptions imprecise coordinates or even mistakes or changes in taxonomic identification Errors are frequently made when recording coordinates in the field especially when a data transcription step is included or when entering data into a database Georeferencing records from an office setting or at a distance based on site descriptions can also lead to a poor data quality Two key aspects of data quality include the accuracy and precision of geographic coordinates The accuracy of coordinates determines the ability to correctly represent the site of collection observation of a presence point Precision refers to the level of detail of the coordinates necessary to represent the described site effectively Precision can be assessed by reviewing the method in which the coordinates were determined e g maps versus GPS or according to the number of decimals included in the coordinates as already discussed in Section 2 1 A lack of accuracy in the analyzed data will inevitably lead to errors in the output results of the analysis while a lack of precision will often result in conclusions of limited use as they are only representative at a very low resolution Data can be very precise but inaccurate and can also be very accurate but highly impre
179. o DIVA GIS as was done in Steps 13 to 19 of Analysis 6 2 1 3 Open DIVA GIS and import the presence absence rasters of the potential distribution areas Pinus_kesiya_gcm_sea_2 5min_thresholded asc and Pinus_kesiya_thresholded asc as integer value raster files these files can be found in the output folder Check the Save as Integer option under Import to Gridfile Multiple Files aa Import Multiple Files to Gridfiles Output Folder Type f Same asinput Select IDRISI IMG or RST Arc BINARY FLT f Arc ASCIl fw Save as Integer e BIL BIF BSO Remove file Remove all PE AGIS Tutorial 6 3 Climate change Pinus_kesiya_gem_sea_2 5min_threshalded asc EGIS Tutorial 6 3 Climate changePinus_kesia_thresholded asc ose _ Neco Chapter 6 4 After closing the Import to Gridfile Multiple Files window open the rasters using DIVA GIS The presence absence map generated for P kesiya under current climate conditions Pinus_kesiya_thresholded grd should be similar to the following illustration Overlaying rasters of current and future potential distribution areas A useful way to identify the impact of climate change on the distribution of groups of species is to overlay the rasters of current and future potential distribution areas Binary rasters are recommended for constructing maps which are easy to interpret however rasters that gradually show potential distribution areas with increasing
180. o any other type of document Chapter 4 AJA Reset Cony On Click E Zoom Attributes By Cimate data 112 valid non duplicate observations tout of 1444 100 90 a0 ro 60 30 40 Cumulative Frequency 30 20 68 9 1041 12 13 14 15 16 17 16 19 20 21 22 23 24 25 26 27 Annual Mean Temperature 1 Percentile 0 02500 C Show H11 510R 4 M ICICI JMe A a Dic Latin America Countries x lo B x vcundimarcensis_outliers af x 75 19400 Y 13 33220 z yveundimarc Rec 1 of 1 Layer ycundimarcensis_outliers gt ia ID 5 amp 3e i ESPECIES Error e LATITUDE 3 3322 LONGITUDE 75 194 COUNTRY Colombia ADM1 Huila A Distribution Modeling Input Frequency Outliers Histogram Envelope Predict On Click Zoom V Attributes 7 Climate data 112 valid non duplicate observations out of 144 100 x 75 19400 Y 3 33220 worldclim_2 5m Climate Graph Bioclim Altitude 378 Cumulative Frequency 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 Annual Mean Temperature 1 Variable Annual Mean Temperature 1 v V Outliers Size 3 i Percentile 0 02500 Show 1 510R Save Stats y 9 8190 Scale 1 10587245 Quality control Sometimes there is more than one outlier at
181. of incomplete sampling However it may be attributed to the rules of probability whereby although the likelihood is low within a random sampling scheme with sufficient observations this situation will occur comparable to tossing a coin several times but never having it land heads up Another explanation for the atypical occurrence of this allele may be the presence of local adaptation as allele SSR7 191 is located at the same locus as allele SSR7 208 See map above locally common in the centre of diversity The dominant presence of allele SSR7 208 replacing allele SSR7 191 may indicate a process of local adaptation However as the analysis is based on neutral markers most likely not accounting for gene expression such an explanation should be made with caution in order to understand the peculiar distribution of alleles at this locus further research is required The situation with allele SSR8 318 is unique The area where this allele is found covers central Peru and Bolivia thus there is a high probability that the allele is also present in the area between these two locations which was not sampled i e southern Peru Presence in this area would mean the allele is more common than what is indicated on the map As such the result suggests that the sampling was incomplete It also indicates that there may be new alleles outside the area of highest diversity northern Peru As discussed in Section 5 4 these considerations are important a
182. ol This analysis allows you to identify further errors in addition to those detected after the first analysis described in Step 7 ID 805 According to its passport data this sample was collected in Peru but according to the map the point is located in Ecuador Since the imprecision in the coordinate is less than 1 km the mistake may indicate an error generated when georeferencing or recording the coordinates The collector may have crossed the country boundary unknowingly which could be the origin of the inconsistency in the administration data This is a common problem and the decision as to what to do with the data point is difficult as it may well be that an error in the descriptive part of the passport data is causing the inconsistency while the coordinates are correct Many studies allow a margin of error around the borders of administrative unit data for example 5 km ID 1143 The point is located in Colombia while its passport information indicates it was collected in Brazil Considering there is a huge gap between the site and the Brazilian border this is clearly a significant error which needs to be resolved if the cause of the problem can be identified or the point must be deleted In this example the problem is a missing negative sign in one of the coordinates ID 2724 This point also illustrates a common problem which is caused by mistakes or discrepancies in the spelling of administrative units i e B
183. ollectors simply overlooked the presence of a specimen is always a possibility Presence points however are generally much easier to relate to more credible factors such as environmental variables Therefore the analyses in this manual are based on presence points only Specimens are often associated with data providing further details as to individual plant characteristics and data used to assess intra specific diversity and or evaluation of agronomical traits which can then be associated with spatial information e g climate data Thus associated data can include information on morphological traits e g size shape colours physiology e g days to germination days to flowering evaluation e g yield tolerance to abiotic and biotic stress or DNA base pair composition e g 1 In crop genetic resources conservation such data are referred to as passport data For more information visit Bioversity s web page on Multi crop Passport Descriptors MCPD www bioversityinternational org index pho id 19 amp user bioversitypublications pil showUVid 2192 Chapter 2 molecular marker data An analysis of diversity based on morphological characterization phenotypic diversity is outlined in Section 5 2 while the analysis presented in Section 5 3 is based on molecular marker data Representing the real world using GIS Real World Raster Real world represented by vectors left and by a raster right GIS
184. olumn shp shx dbf e Latin America Countries shp shx dbf e Protected Areas Latin America shp shx dbf 5 4 1 How to identify priority zones for in situ conservation or germplasm collection Through this analysis you will learn to use the Analysis Point to Grid and Reserve Selection option in DIVA GIS to assist in defining priority conservation areas The Reserve Selection procedure uses an optimization algorithm that was originally developed to minimize the area needed to conserve flowering plant diversity in South Africa See Rebelo and Siegfried 1992 In this analysis we will use that algorithm to define the minimum number of geographic units needed to conserve all genetic diversity measured through molecular markers and to identify in sequence of importance the geographic units that should be prioritized for conservation The data used in the previous analysis will be used again The same type of analysis can be run at the species level as well Chapter 5 Steps 1 Use the molecular marker data for cherimoya to start the analysis and select the Analysis Point to Grid option For this analysis select the Reserve Selection option as well as the Complementarity option Use a raster with one 1 degree cells preferably one of the rasters used in the previous analyses Point to Grid Parameters Input Shapefile 6 4 conservation strategies ssr chenmoaya rand column sh Define Grid Create anew Grid
185. ompare richness among cells that have a dissimilar number of observations or samples for more information refer to Petit et al 1998 and Leberg 2002 The rarefaction 3 Results may be slightly different from the map shown as they will depend on the origin of the raster In this analysis the raster used had the following characteristics X Min 82 Max 62 Y Min 24 Max 0 Spatial analysis of diversity for conservation planning method recalculates the richness measured in the different cells as if a standard number of observations were made in each cell Only cells with an equal or higher number of observations than the standardized number are included in the analysis cells with fewer observations are excluded The choice of the standardized number of observation is a trade off between the number of cells included in the analysis and the maximum richness to be calculated For example if a small number of observations is chosen almost all cells will be included in the analysis however the maximum diversity a cell will be low as it cannot be higher than the defined number of samples If a high number of observations is chosen the number of cells with at least this number of observations essential to be included in the rarefaction calculation will be low resulting in a limited number of cells with a calculated value The following analysis is based on eight microsatellites whereby 80 allele observations were used as the fixed
186. oneous coordinates are apparent when double checking this second level of data In addition to those points previously described ID 805 ID 1143 ID 2724 which showed errors at the country level and the points ID 319 ID 796 ID 1367 ID 1669 and ID 2967 located outside of the study area these error points include ID 689 This point is located on the boundary between two provinces in Ecuador Morona Santiago and Chimborazo As was the case with ID 805 the problem may be related to an imprecise coordinate but the most probable explanation is that the collector crossed the provincial boundary unknowingly resulting in an error in the descriptive data When the distance between the point and the boundary is small as is the case here generally there is no need to take corrective action ID 2729 The latitude of this point is listed as zero 0 a mistake that was previously described see point ID 319 This point clearly illustrates the need to establish controls at lower administrative levels as points with this type of error are not always located in the ocean or in another implausible site ID 2870 The collection site has different names at the Colombian department level When the point was recorded the department s site name was Cundinamarca but several zones in this locality including the collection zone were later renamed as the Distrito Capital The change of a name in the administrative unit often generates erro
187. optimal size of the cells depends on the size of the geographic area and on the objective of the study The size of the study area is referred to as the extent while the level of detail given by the cell size is called the resolution Using different cell sizes results in rasters with different resolutions smaller cells generate rasters with higher resolution Throughout this manual different raster cell sizes are used ranging from one 1 degree Preparing and importing data to DIVA GIS and Maxent approximately 111 km at the equator to 30 seconds approximately 1 km at the equator Though still rather large geographic units these sizes are appropriate for many types of spatial analyses and species distribution modelling at the national or regional level These types of analyses are presented in Chapters 5 and 6 Approximate size of geographic units at the equator rounded to km Degrees Size GIS and species distribution modelling programmes use a range of different file types when processing raster data The rasters used in the DIVA GIS programme are in grid format but the programme also allows one to import or export other types of raster file types see Analyses 3 1 5 and 3 1 6 A raster in DIVA GIS consists of two files one GRD file grd and one GRI file gri however only the GRD file is shown when the raster is opened in DIVA GIS The GRD file contains the raster s general information such as
188. or Maxent to process the environmental raster files all files must have the same parameters in terms of raster properties resolution and raster corners or vertices Note Steps 1 Use Maxent to generate potential natural distribution models for all the different Vasconcellea species following the explanation given in Steps 1 to 8 of Analysis 6 2 1 where the potential distribution for one species was modelled Maxent can also carry out an analysis for multiple species simultaneously when presence points for each species are saved in the same CSV file csv This is the case for the 15 Vasconcellea species Vasconcellea csv file For this analysis models will be generated for the Vasconcellea species using the rasters of the Bioclim variables with a 5 minute resolution for Latin America and the Caribbean these can be found in the wclim_ams_5min folder elak L Maximum Entropy Species Distribution Modeling Version 3 3 3e Samples Environmental layers File GIS Tutoriali6 4 Gap analysis Vasconcellea c Browse DirectoryFile j 4 Gap analysistwelim_ams_Smin Browse V _candicans bio_1 Continuous V _cauliflora bio_10 Continuous V _crassipetala bio_11 Continuous V _cundinamarcensis bio_12 Continuous V _glandulosa bio_13 Continuous V _goudotiana bio_14 Continuous V _microcarpa bio 15 Continuous V _monoica bio_16 Continuous
189. ordan Kenya Malaysia Mali Mauritania Mauritius Morocco Norway Oman Pakistan Panama Peru Poland Portugal Romania Russia Senegal Slovakia Sudan Switzerland Syria Tunisia Turkey Uganda and Ukraine Financial support for Bioversity s research is provided by more than 150 donors including governments private foundations and international organizations For details of donors and research activities please see Bioversity s Annual Reports which are available in printed form on request from bioversity publications cgiar org or from Bioversity s Web site www bioversityinternational org The geographical designations employed and the presentation of material in this publication do not imply the expression of any opinion whatsoever on the part of Bioversity or the CGIAR concerning the legal status of any country territory city or area or its authorities or concerning the delimitation of its frontiers or boundaries Similarly the views expressed are those of the authors and do not necessarily reflect the views of these organizations Mention of a proprietary name does not constitute endorsement of the product and is given only for information Bioversity International Via dei Tre Denari 472 a 00057 Maccarese Rome Italy Tel 39 0661181 bioversity publications cgiar org www bioversityinternational org Bioversity International is the operating name of the International Plant Genetic Resources Institute IPGRI a
190. paring the potential distribution areas in the current climate based on climatic conditions at presence points with the potential distribution areas based on a species current climate preferences under future climatic conditions Future potential distribution areas of occurrence are identified using climate layers based on the projections of General Circulation Models GCM Climate research institutions from various countries generate these models http www ipcc data org which predict future climatic conditions under different emission scenarios developed by the Intergovernmental Panel on Climate Change IPCC for further information see http www ipcc ch publications and data publications and data shtml Be careful not to develop a future potential distribution map based on the future climate conditions at the present day presence points The current and future potential distribution areas both need to be based on the species climate niche that is calculated with current climate data Note PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 6 3 Climate change e Maxent and Java e pkesiya csv e seacountries shp shx dbf e Folder wclim_sea_2 5min asc files e Folder gcm_sea_2 5min asc files 6 3 1 How to evaluate the impact of climate change on the distribution of species Despite several on going in situ pine conservation projects in Southeast Asia the area occupied by
191. pecies from the analysis oy un checking the boxes in front of each species In this analysis all species will be included Point to Grid 15 classes candicans cauliflora crazsipetala cundinamarcenzis glandulosa goudotiana microcarpa monoica parviflora pulchra quercifolia sphaerocarrpa shipulata weberbaueri x heilbornii Selection E SSeS L E E E E E L E cere Chapter 5 9 Click on the Main tab and select the button to the left of the Output box the button with the ellipsis Enter the name of the file to be saved and its file path Finally click on Apply The resulting raster will show the number of species observed in each cell a Point to Grid Parameters Input Shapetile Agi tutorial 1 species diversity vasconcellea species shp Define Grid Create a new Grid Options Output arable Richness Number of different classes Richness Paint to Grid Procedure Simple Output p3 EAGIS Tutorial 5 1 Species diversityvasconcellea div Wl Close Spatial analysis of diversity for conservation planning Results of the analysis after moving layers show that cells in southern Ecuador and central Colombia contain up to nine different Vasconcellea species indicating that overall Ecuador and Colombia possess a higher diversity of this genus as compared to other Latin American countries Data used in this analysis correspo
192. pefile File Manager Download 14 In the File type window select ESRI ASCII 15 Select the desired ASCII raster file asc from which to generate a raster file in grd format For this analysis select the ASCII file generated by Maxent Pinus_kesiya asc 16 Under Save as Integer you can select whether or not to generate a raster in which values are presented in integers in the legend of DIVA GIS For this analysis the raster will not have integral values therefore it is important not to select this box 17 Under Output File give the name for the raster file grd that will be generated by DIVA GIS 18 Click OK to initiate the process Importo Gridfile File type A IDRISI ESRI binary FLOAT at Input File e gis tutorial 6 2 potential distibution pinus_kesiya asc Output File e vgis tutoralyb 2 potential distibution pinus_Kesiya grd 18 ee ee eee ee ee ane ey eer Chapter 6 19 After the raster generated in Maxent has been imported to DIVA GIS this file can be opened visualized and modified Visualization can be improved by modifying the standard legend as described in Chapter 3 The following steps outline the type of changes which can be made the choice of colours and the use of a more gradual scale can vary according to the user s preferences Ja Properties 3 _ Properties Label Pinus_kesiya Filename gis tutorial 6 2 potential distribution pinus_kesiya grd Filename
193. probabilities can also be used for more detailed analysis In this analysis the binary rasters of current and future potential distribution areas of P kesiya will be overlaid Overlaying binary rasters results in four possible situations for each cell High impact areas areas where a species potentially occurs in the present climate but which will not be suitable anymore in the future ii Areas outside of the realized niche areas that are neither suitable under current conditions nor under future conditions as modelled iii Low impact areas areas where the species can potentially occur in both present and future climates iv New suitable areas areas where a species could potentially occur in the future but which are not suitable for natural occurrence under current conditions Species distribution modelling and analysis Each binary raster has two values presence 1 and absence 0 When these two values are added or subtracted the only possible results for the cells are negative one 1 zero 0 and one 1 However a problem exists in that a fourth value is required to represent the four possible situations outlined above The following table illustrates this problem subtracting the rasters results in cells with the same value for the second and third situations ii areas outside of the realized niche and iii low impact areas as neither experiences any change hence the value of zero 0 when combining B
194. que Classes C TRINIDAD Y TOBAGO E ay Field COUNTRY C VENEzuELA Reset Legend Use color of single tab C ARGENTINA E BELICE e BOLIVIA EE ERASI CHILE SelectAll Clear All Aew O Tee E Yasoncellea species o Si8 i8 8 8 x 101 8202 y 30 1288 Scale 1 42881984 Data Design 11 Now remove the countries legend return to the Single tab and eliminate Solid Fill and visualize the layer with the Vasconcellea points again This time assign a different colour to each species following the procedure explained in the previous steps Le DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Osi ae Cle gt E Latin America Countries 7 Yasoncellea species Slej x Label Vasoncellea species Source gis tutorial 3 1 basic elements vasoncellea species Type Point 97 10 18 85 48 44 31 33 Single Unique Classes Field SPECIES Reset Legend Use color of single tab o W candicans Y cauliflora V crassipetala W cundinamarcensis WY glandulosa Y goudotiana qi S iS ARER Vi microcarpa z A J SelectAll ClearAll Ry Close x 114 9675 y 32 6134 Scale 1 42881984 Data De
195. r shp file and the worldclim_2 5min climate data file should appear automatically in the Points and DIVA Climate data windows If they do not appear open them manually 5 Keep the From same grid cell option checked to exclude duplicates in the climatic variables cells Distribution Modeling oa Frequency Outliers Histogram Envelope Predict sat Points e gis tutorial 4 2 quality control atypical points vcundimarcensis_outliers shp 4 Pa rer e DIVA Climate data worldclim_2 5m Stack i eo Remove duplicates With same coordinates V From same grid cell OneClass Many Classes 6 Goto the Frequency tab to understand the climate distribution of the presence points of V cundinamarcensis 7 Select the climatic variable of interest For this analysis the selected variable is the Annual Mean Temperature file Click Apply to see the results on the graph 8 Using the Outliers option atypical data according the Reverse jackknife method will be highlighted in dark green 9 Check the Zoom Attributes and Climate data boxes and click on any point in the graph The corresponding presence and climate data for that point will be displayed and the point will be highlighted on the map for a few seconds 10 To adjust the size of the graph use the Zoom in or Zoom out buttons located on the upper left hand of the menu 11 The Copy option allows you to copy and paste the graph int
196. razil versus Brasil This is obviously not an erroneous point but rather a difference in spelling language and can be easily corrected though it is not always necessary to do so Data control at lower level administrative units When passport data includes information at a lower administrative level e g departments or provinces it is recommended to run a check for errors at this level as well The initial procedure is the same as explained in Steps 1 to 5 above In Step 6 however where relationships are indicated the layer with the lower administrative unit data Latin America Adm 01 shp must be added After doing so continue with Steps 7 and 8 simply adding one more level in the relationship field with the Fields of Shape of Point and Fields of Shape of Polygon tools Chapter 4 a Check Coordinates Options Points outside all polygons Points do not match relations Points do not match with r Shape of Port Input File e gis tutonals4 7 quality control administrative units yvasconcellea final erors shp Field of Longitude LONGITUDE a Field of Latitude LATITUDE Shape of Polygon Input File EGIS Tutoriahg d Quality control Administrative units Latin America Adm 01 shp Relations Fields Fields of Shape of Point Fields of Shape of Polygon Relation COUNTRY eT 1 E Relation 2 ADM z mel Belaian EO O O o hf Apply A Close New presence points with possible err
197. redictive models of species distributions criteria for selecting optimal models Ecological Modeling 162 211 232 Araujo MB Pearson RG Thuiller W Erhard M 2005 Validation of species climate impact models under climate change Global Change Biology 11 1504 1513 Elith J Graham CH Anderson RP Dudik M Ferrier S Guisan A Hijmans RJ Huettmann F Leathwick JR Lehmann A Li J Lohmann LG Loiselle BA Manion G Moritz C Nakamura M Nakazawa Y Overton J Mc Townsend C Peterson A Phillips Su Richardson K Scachetti Pereira R Schapire RE Soberon J Williams S Wisz MS Zimmermann NE 2006 Novel methods improve prediction of species distributions from occurrence data Ecography 29 129 151 FAO 2007 Ecocrop on line Available from http www ecocrop fao org Date accessed October 2010 Fawcett T 2006 An introduction to ROC analysis Pattern Recognition Letters 27 861 874 Fjeldsa J 2002 Polylepis forests Vestiges of a vanishing ecosystem in the Andes Ecotropica 8 111 123 Guarino L Jarvis A Hijmans RJ Maxted N 2002 Geographic Information Systems GIS and the Conservation and Use of Plant Genetic Resources In Engels JMM Ramanatha Rao V Brown AHD Jacson MI editors Managing plant genetic diversity International Plant Genetic Resources Institute IPGRI Rome Italy pp 387 404 Hernandez PA Graham CH Master LL Albert DL 2006 The effect of sample size and species characteristics on performa
198. resolution such as those generated in Analysis 3 1 6 and when using future climate data This section illustrates how to make CLM files in DIVA GIS As an example in this analysis CLM files will be prepared from the datasets generated in Analysis 3 1 6 Minimum temperature Maximum temperature Precipitation and Altitude Remember that all rasters must have the same extent and be of the same resolution as is the case for the previously generated datasets The preparation of CLM files in DIVA GIS is also explained by Ramirez and Bueno Cabrera 2009 Steps 1 Each dataset Minimum temperature Maximum temperature Precipitation contains 12 files corresponding to monthly values Make sure the names of these files are differentiated by the numbers 1 through 12 The numbers need to be located at the end of the file names Except for the end number the file name should be the same in each of the 12 files For each dataset this is the prefix that will be used to develop the CLM file If this is not the case the file names need to be changed accordingly For this analysis the datasets of tile 33 and the names of the grd and the gri files must be changed manually Chapter 3 2 Note In the case of the dataset Precipitation prec1_33 grd and prec1_33 gri become prec1 grd and prec1 gri prec2_33 grd and prec2_33 gri become prec2 grd and prec2 gri prec12_33 grd and prec12_33 gri become prec12 grd and prec12 gri In the case of
199. roblem can be partially addressed but using this method does result in the loss of certain observations In Section 6 4 improving diversity studies using species distribution modelling is explained The Point to Grid option in DIVA GIS also contains several estimators of richness to partially overcome this problem These tools are also useful to estimate the additional number of species that can occur in each geographic unit of measurement which are not yet observed due to under sampling See Section 6 2 2 of the DIVA GIS Manual Version 5 2 for more information on these estimators http www diva gis org docs DIVA GIS5_manual pdf The best solution though is to prevent such bias from occurring by ensuring even sampling to the greatest extent possible Effects of changing grid origin on the result of a Point to grid analysis A Point to Grid Analysis takes into account the observations found in each cell of the raster e g richness looks at the number of observed units of diversity species for instance in each cell The definition of the raster will obviously influence the result of the analysis The coming sections explain how differences in raster cell size resolution influence the final result In addition to being defined by the size of its cells a raster is also defined by its origin the minimum and maximum X and Y values entered in Grid Options Below is a simple illustration of this effect based on a raster of four cells Richn
200. rs when using historic data from herbariums or museums ID 2906 This point has a spelling mistake a common error when using characters such as the which do not exist in all languages ID 2943 Another instance of a zero 0 value input for the latitude ID 3023 A problem in the name of a department Valle abbreviated common name versus Valle del Cauca complete name ID 3068 This point has the same problem described for the point with ID 2729 above Quality control 4 2 Quality control through the identification of atypical points Atypical points or outliers are presence points located outside the limits of the species normal environmental ranges Atypical data occupy an ambivalent place in spatial biodiversity analyses Outlier identification methods can help to detect erroneous presence points in a dataset which should be removed to ensure data quality An atypical environment might be an indication of incorrect presence data resulting from different types of errors erroneous geographical coordinates erroneous taxonomic classification or introduction of individuals in places which do not correspond to their range of natural occurrence e g production systems and botanical gardens On the other hand an atypical point can also indicate an individual or a group of individuals that have adapted to environmental factors different to those of the most naturally occurring individuals and populations
201. s a Export Multiple Gridfiles File Type Output Folder BIL generic binary CO Shapefile polygon Same as Input Select 7 e 10 IDRISI 16bit shapefile point IDRISI 32bit H at s a a a IEN Ce H N a a SA os E binary FLT a a r uw uw a P ESAI asc Make valid ESAI grid names 9 merer Add file Remove file Remove all 19 files File name E G IS Tutoriala 1 Basic elements B 1019 grd E GIS Tutonalys 1 Basic elements BI0O1 grd EAGIS Tutornalys 1 Basic elements BIO Z grd E GlS Tutoralys 1 Basic elements BIO grd EAGIS Tutonalys 1 Basic elements BIO 4 grd EMGIS Tutoriala Basic elements B05 rd 1 lal Af Appl a Booe 13 The generated ASCII files asc can now be used in Maxent for species distribution modelling see Analysis 2 2 2 3 1 6 How to import generic climate data to DIVA GIS The climate layers created in the previous section based on Data Climate Map are derived from the CLM files clm using the 2 5 minute resolution data described in Chapter 2 for DIVA GIS The Worldclim website www worldclim org provides more detailed climate data up to 30 seconds or 1 km at the equator For analysis in a small area using highly precise presence points these climate data might be the most appropriate however rasters with 30 second resolution for the entire world occupy large amounts of space on the computer Therefore it is useful
202. s S 2009 A Brief Tutorial on Maxent on line Available from http www cs princeton edu schapire maxent tutorial tutorial doc Date accessed October 2010 Puliam HR 2000 On the relationship between niche and distribution Ecology Letters 3 349 361 Scheldeman X Willemen L Coppens D eeckenbrugge G Romeijn Peeters E Restrepo MT Romero Motoche J Jimenez D Lobo M Medina Cl Reyes C Rodriguez D Ocampo JA Van Damme P Goetghebeur P 2007 Distribution diversity and environmental adaptation of highland papaya Vasconcellea spp in tropical and subtropical America Biodiversity and Conservation 16 6 1867 1884 USDA ARS National Genetic Resources Program Germplasm Resources Information Network GRIN Online Database National Germplasm Resources Laboratory Beltsville Maryland Available from http ars grin gov cgi bin npgs html taxon pl 28462 accessed October 2010 van Zonneveld M Koskela J Vinceti B Jarvis A 2009a Impact of climate change on the distribution of tropical pines in Southeast Asia Unasylva 60 231 232 24 28 van Zonneveld M Jarvis A Dvorak W Lema G Leibing C 2009b Climate change impact predictions on Pinus patula and Pinus tecunumanii populations in Mexico and Central America Forest Ecology and Management 257 7 1566 1576 Willis F Moat J Paton A 2003 Defining a role for herbarium data in Red List assessments a case study of Plectranthus from eastern and southern tropical
203. s an error occurs when two or more rasters are combined during calculations in DIVA GIS e g overlay or when two stacks of rasters are used as input in Maxent e g to compare areas of potential distribution of a species under different environmental scenarios or climates see Section 6 3 This is usually due to differences in raster properties Raster properties i e resolution cell size extent number of rows and columns raster corners or vertices min and max X and Y values need to be identical in order to combine them in DIVA GIS as explained in Analysis 3 1 3 and to use them as inputs in Maxent after being converted to ASCII format asc To accomplish this in DIVA GIS rasters should be created with identical properties as mentioned in Analysis 3 1 5 Nevertheless after using these options to create rasters of identical properties very small differences in the decimals of the resolution and or in the coordinates may still remain This may happen if rasters of identical properties are created from different datasets imported to DIVA GIS e g current and future climate data sets or soil and climate datasets Basic elements of spatial analysis in DIVA GIS Therefore if an error occurs in the calculations when combining rasters or stacks in DIVA GIS or in species distribution modelling with Maxent it is recommended to verify the differences in resolution and or in coordinates for the vertices of the rasters This can be
204. s named decimal degrees Columns H and O 38 Microsoft Excel Vcundinamercensis_DMSdata xls B9 File Edit View Insert Format Tools Data Window GenAIEx Help Adobe PDF 7 fe D2 E2 F2 60 G2 3600 B C D E F G R SPECIES LATITUDE Lat degrees minutes seconds decimal deqrees latitude LONGITUDE Lon degrees minutes 1 75 3433 V cundinamarcensis 6 57 48 N 1 6 57 48 6 96331 7502503 3037 V cundinamarcensis 7 10 17 N 75 45 47 WV 2816 Y cundinamarcensis 790159 N 7501859 2836 V cundinamarcensis 6 19 59 N 7501500 3030 V cundinamarcensis 6 5400 N 7505759 415 ndinamarcensis BA9 5G io G A Go to the Formulas spreadsheet Special formulas based on the Excel text functions RIGHT MID LEFT are inserted in the columns named Lat Lon degrees minutes and seconds Columns D and K Columns E F G and L M N in order to separate the values of the hemisphere degrees minutes and seconds from the Latitude and Longitude columns Columns C and J into separate columns Go to the Copied formulas spreadsheet The formulas are copied in each row to determine the latitude and longitude in DD for every presence point 38 Microsoft Excel Vcundinamercensis_DMSdata xls B9 File Al Edit View Insert Format Tools Data Window GenAIEx Help Adobe PDF v C F G H l J K LATITUDE minutes seconds decimal degrees latitude LONGITUDE Lon degrees minutes 75 3433 V cundinamarcensis 6 57 48 N 57 48 6 9633 75 2503 W 3037 V
205. s of wild plant species As mentioned in Section 6 2 one of the uses of potential distribution maps is to detect gaps in the data on a species distribution A gap refers to a location where species distribution modelling predicts that a species could potentially occur but where specimens and or germplasm of wild species or crops have not actually been collected Gaps may indicate that accessions and specimens from these areas are missing in germplasm or herbarium collections e g Jarvis et al 2005 Scheldeman et al 2007 On the other hand local studies and observations of the species in these areas may be available but the information may not be broadly disseminated and not included in initiatives such as the Global Biodiversity Information Facility GBIF www gbif org that promote the use of data by the general public see Section 2 3 It is a real possibility however that a species simply does not exist in the area predicted due to dispersal limitations Such a case was explored in Section 6 2 where Pinus kesiya was predicted to be present in several south eastern islands of Indonesia but did not naturally occur in those zones Some species may also have disappeared from an area due to deforestation selective extraction or other anthropogenic pressures on the natural habitat For example Polylepis forests in Bolivia and Peru are currently very fragmented based on ecological niche studies these forests were demonstrated to have pr
206. sava accessions with the heaviest roots are found mainly in the south eastern part of the study area This analysis provides information on a specific characteristic in this case fresh root weight but does not offer information on the areas diversity In order to gather this information you will need to look at the range of the analyzed parameter FRW 4 Under Analysis Point to Grid Statistics select the Range parameter Use the same raster as in the previous analysis FRW Note that the previously used established legend can no longer be applied as the calculated parameter range instead of maximum has been changed and dimensions will be different Adjust the legend to the new results 2 Results may be slightly different from the map shown as they will depend on the origin of the raster This analysis used a raster with the following characteristics Min X 78 Max X 71 Min Y 12 Max Y 5 Spatial analysis of diversity for conservation planning The resulting map reveals that those cells corresponding to sites where the heaviest weight is found also correspond to the cells with the greatest ranges in weight Even though the range provides an idea of the variability within a parameter it has two disadvantages firstly the range only takes into account the extreme values not the distribution of values and secondly the value of the final result will depend on the units of measurement in this case grams Combining the
207. se results with the results of other trait analyses such as length of leaf petiole or number of leaf lobes may be difficult but is nonetheless important to understand the overall diversity which needs to include different traits By using a parameter without dimensions i e the coefficient of variation CV which instead considers the distribution of the parameters around the average such analysis combinations are more feasible Under Analysis Point to Grid Statistics select the Coefficient of Variation parameter This map provides a better picture of the diversity of the FRW parameter in the different cells Analyze the CV of another trait i e the length of leaf petiole LLP Use the same raster properties from the previous analysis in order to compare the results of the two analyses Chapter 5 Areas with a high variability for the LLP parameter are different from those for fresh root weight How do you then define where the sites with the highest cassava diversity are located The solution lies in combining the different traits As previously mentioned unlike the range the use of the Coefficient of Variation enables the combination of different parameters which is necessary to undertake such an analysis 7 Five traits will now be included for comparison the two previously analysed traits fresh root weight FRW and leaf petiole length LLP and three new parameters number of leaf lobes NLL distance between
208. shp shx dbf e Vcandicans_outliers shp shx dbf e Latin America countries shp shx dbf For this analysis you need to have the 2 5 min worldclim climate data imported in DIVA cf Section 2 2 4 2 1 How to identify outliers based on environmental data In this analysis you will learn how to identify outliers using DIVA GIS and data for V cundinamarcensis and V candicans Scheldeman et al 2007 Errors were intentionally included in the database for the purpose of the analysis Errors can be indentified where the species name has been replaced with the word Error in the dataset The V cundinamarcensis dataset is large enough 144 observations to be revised using the Reverse jackknife method while the dataset for V candicans is smaller 17 points and will need to be revised using the 1 5 QR method Steps 1 Open the vector shp files Vcundinamarcensis_outliers shp and Latin America countries in DIVA GIS 2 To detect atypical points using the Reverse jackknife method select the point file by clicking on the folder in the legend 3 Go to Bioclim Domain in the Modeling menu Les DIVA GIS 7 3 0 Project Data Layer Map Analysis IMENT Grid Stack Tools Help DAF CY Bioclim Domain M External Models 7 Latin America Countries Evaluation o 3 EcoCrop vcundimarcensis_outliers o Terrain Modeling Quality control 4 The Vcundinamarcensis_outlie
209. sign Chapter 3 12 Using the options in the Properties window assign a red circle to observations for the species V cundinamarcensis and a green triangle for the species V glandulosa The other species should remain invisible Le DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help DEt Garaa F Latin America Countries F Yasoncellea species V cundinamarcensis 4 WV glandulosa Label Vasoncellea species Source gis tutorial 3 1 basic elements vasoncellea species Type Point 97 10 18 85 48 44 31 33 Single Unique Classes Field SPECIES Reset Legend F Use color of single tab V candicans Y cauliflora Y crassipetala V cundinamarcensis 4 V glandulosa Y goudotiana V microcarpa W mannira SelectAll Clear All Er x 101 6132 y 32 7169 Scale 1 42881984 Data Design It is important to realize that a selection at the legend level will not change the content of the vector files To carry out an analysis of a subset of data you must first select the subset and save it as a different vector file Return to the Single tab in the legend and select a single symbol for all points All observations are again visualized with a single symbol 13 Now select not display a specific group of Vasconcellea species using either of the two attribute selection alternatives provided by DIVA GIS in the Layer
210. sis at national level 1 degree gap analysis at province level 10 minutes gap analysis at province level Yellow areas showing potential presence where the species was actually observed Red potential distribution areas for the species but where it has not been observed Gray areas outside of the observed and potential presence of the species In view of the limited capacity to effectively identify gaps when using high resolution rasters it is important to increase the cell size of the binary rasters from the previously generated potential Vasconcellea diversity raster Species distribution modelling and analysis Steps continued from the previous section 13 In the Grid menu select the Aggregate option Wa DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Meite Stack Tools Help De gt QQ fe DE Describe Pike Jil Overlay Scalar Fd i Reclass E Po no Neighborhood E i Calculate 2 2 Latin America Countries wo Disaggregate Cut Merge Baaeaooo omnnroaanp wo SoU aH New Transect Area Emam esas aepos 14 In the Input box select the stack of rasters of which to increase the cell size In this analysis select the stack with the binary rasters _thresho ded of the potential distribution areas for each Vasconcellea species this is the stack created in Step 4 15 Under the Kind option select the type of calculation you wish to us
211. sis in DIVA GIS Merge grids When you want to use 30 second resolution climate layers for a specific area the area may extend over all or part of the adjacent 30 x 30 degrees tiles After having generated the raster files grd the climate layers of the adjacent tiles can be merged in DIVA GIS using the Grid Merge option First stack each dataset using the Stack Make Stack option and then merge them using Grid Merge Concatenate Grids Input B Gid Stack Compare file names First E GIS Tutoriala 1 Basic elements trin_23 gre second EGIS Tutorialys 1 Basic elementsstrin_33 gre Result E 4GI5 Tutoriala 1 Basic elements trin_2s_ 33 gre In this image the stacks of the 12 minimum temperature layers from zones 23 and 33 are merged 3 1 7 How to make CLM files in DIVA GIS In the previous section datasets with layers of monthly precipitation minimum and maximum temperatures were generated When performing an ecological analysis in DIVA GIS it is often easier to work with CLM files clm with its 19 bioclimatic parameters than with the individual climate layers For example climate data from CLM files are used in the Analysis 3 1 4 and Sections 4 2 and 6 1 The DIVA GIS website provides climate databases clm files of current climatic conditions up to a resolution of 2 5 minutes 5 km at the equator see http www diva gis org climate You may however want to prepare CLM files with climate data of 30 second
212. sis using molecular marker data 5 4 IMPLICATIONS FOR THE FORMULATION OF CONSERVATION STRATEGIES 5 4 1 How to identify priority zones for in situ conservation or germplasm collection 6 SPECIES DISTRIBUTION MODELLING AND ANALYSIS 6 1 ANALYSIS OF THE REALIZED NICHE OF A SPECIES 6 1 1 How to analyze and compare realized niches of different species 6 2 MODELLING THE POTENTIAL DISTRIBUTION OF A SPECIES 6 2 1 How to model the potential natural distribution of a plant species 6 3 MODELLING THE IMPACT OF CLIMATE CHANGE ON SPECIES DISTRIBUTION 6 3 1 How to evaluate the impact of climate change on the distribution of species 6 4 IDENTIFICATION OF GAPS IN COLLECTIONS OF WILD PLANT SPECIES 6 4 1 How to identify possible gaps in collections 66 69 14 74 ff 79 80 80 87 88 97 99 99 111 112 118 119 132 133 139 140 141 147 148 159 159 167 167 Acknowledgements The authors are indebted to many colleagues and peers for their important role in the development of this manual We would like to acknowledge Nora Castaneda Bioversity CIAT who with her knowledge on using GIS methodologies to carry out spatial analysis of biodiversity data participated in the initial phase of development We also would like to give special thanks to Colm Bowe Centre for Underutilized Crops UK whose feedback on the first versions of the manual contributed to shaping its current form
213. sl a Field SSAS e gis tutorial 5 4 conservation strategies ser cherimoya ran Bi classes Detine Grid Create a new Grid T Output Variable Richness Presence 4 bsence Point to Grid Procedure Simple ka Output J E GIs Tutorial 3 Diversity Molecular marker datap YES Contrary to most analyses conducted in DIVA GIS results of the Presence Absence Analysis do not display automatically To visualize the results open the layer and go to the sub directory marked with the name assigned to the file Results will be displayed in the following way Case 1 Common alleles in a reduced area SSR1 289 SSR7 208 10 observations Max distance 99 km 20 observations Max distance 268 km Comparing these two maps with the map of diversity generated in the previous analysis reveals that locally common alleles are found in the zones with higher richness of alleles confirming what was previously observed northern Peru is a hotspot for the diversity of cherimoya Chapter 5 Case 2 Rare alleles SSR7 191 and SSR8 318 with a broad distribution SSR7 191 SSR8 318 7 observations Max distance 891 km 5 observations Max distance 1646 km Allele SSR7 191 is found in different areas in northern Peru but is not located within the area of greatest diversity where the presence of unique alleles indicates rather complete sampling suggesting that this situation is not the result
214. software can process two different types of spatial data vector and raster data Vector data is composed of a set of georeferenced points which form either groups of points lines or polygons and represent actual geographic areas In this manual the georeferenced points of plant individuals presence points or the administrative unit layers of countries polygons serve as examples of vector data The most commonly used vector data file format is the shapefile DIVA GIS uses shapefiles to represent vector data A shapefile consists of at least three files with the same file name but a different file name extension The extensions of these files are shp shx and dbf Together these three files make up a shapefile The SHP file shp is the main file that stores spatial features and is the one shown in the dialogue box when opening the vector file using DIVA GIS The non spatial attributes of the spatial features are stored in the dBase IV file dbf while the SHX file shx contains the indexing information of the vector file it is used by computer programmes to more quickly access the shp file and link that file to the records in the dBase IV file dbf Environmental data from specific geographic areas may also be georeferenced and are normally organized in a raster Rasters consist of a grid of cells of identical size with each cell containing a value for a variable of interest e g temperature soil type The
215. subgroup a W Temporary Places WS vasconcelea b v candicans b MI amp v cauliflora gt MB v crassipetaa ME v cundinamarcensis b A v glandulosa b M v goudotiana b ME v microcarpa b VO v monica Y v parviflora b MO v pulchra b MVE v quercifola b MO v sphaerocarpa b VE v stipulata b VE v weberbaueri b MO v x helbomi e v Layers Earth Gallery gt gt amp Primary Database f 00gle Eye alt 3655 mi 7 Using the Pan and Zoom tools in Google Earth locate the area in northern Argentina where there is a large presence of Vasconcellea quercifolia around the city of Salta Notice that some points are located in downtown Salta which illustrates the danger of using a very high resolution Most likely it was not possible to georeference these points to the same degree of precision or resolution as the satellite imagery that can be viewed in Google Earth ES Temporary Places a E amp vasconcellea t v candicans v cauliflora v crassipetala b v cundinamarcensis t v glandulosa v goudotiana v microcarpa v monaca v parifiora U v pukhra VG v sphaerocarpa T v stipulata v weberbaueri v x heilbomii v Layers E amp primary Database a Google Chapter 3 an Salvador de Jujuy Ralipal E San Pedro Salta Google Eve et 63 62 The next activity consists of visualizing a raster such as those resulting from a
216. t dbf which can be opened and searched in Excel using the Edit Find option However it may not always be possible to georeference a site using a gazetteer This might be the case if information on the locality is incomplete or simply not available Another difficulty may arise if several places have the same name see the table below In this situation differences at higher administrative levels such as Adm1 can help to distinguish between places and resolve the conflict In addition to gazetteers other programmes such as Biogeomancer available at http www biogeomancer org allow georeferencing of points based on site information 3 Microsoft Excel MEX dbf E File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF A13106 z fe Jalapa D E E 24 8833 7101 5500 Coahuila De Zaragoza Jalapa 13107 Jalapa 13108 Jalapa 13109 Jalapa 13110 Jalapa 13111 Jalapa 13112 Jalapa 13113 Jalapa PPL 20 8666 101 7166 Guanajuato PPL 20 1666 102 0166 Michoacan de Ocampo PPL 9 5333 96 9166 Veracruz Llave PPL 7 7166 92 8166 Tabasco PPL 7 6333 99 5666 Guerrero PPL 7 3333 99 2666 Guerrero PPL 6 5000 95 4666 Oaxaca pp ri oo 797 EGRA i m ia n Sia n Sba n Sia n Fa Fia n Sda n ia n Gazetteer for Mexico showing more than one site referred to as Jalapa Chapter 2 2 1 2 How to import georeferenced presence points to DIVA GIS As mentioned presence points can be imported from Excel to DIVA GI
217. t an invalid biased representation of reality AS mentioned in the introduction of Chapter 4 many spatial biodiversity analyses are partially or totally based on data compiled from herbaria and genebanks which often reflect non systematic and uneven sampling Hijmans et al 2000 Chapman 2005 Individual Task Carefully analyze diversity in Ecuador based on a 10 minute raster cells of approximately 18 km only in Ecuador using the option Draw Rectangle Hint What will be the desired raster cell size Your results may be slightly different from the following as they will depend on the origin of the raster Since you are only analyzing results for Ecuador disregard the warning message indicating that there are points outside the selected raster Note Using the circular neighborhood option for richness analysis Thus far the results of analyses have been heavily dependent on the definition of the raster especially on the size of the cell or resolution Small cell size generates a higher resolution detail but risks losing spatial patterns for example when cells are so small that each cell only contains one observation This situation can be improved by applying the Circular Neighborhood option which considers the diversity in adjacent areas With this option each cell receives the value of diversity found within a circle with a specified diameter centred on the cell instead of the value of diversity found within the cell alone
218. ter specific diversity analysis of the Vasconcellea genus Chapter 5 Simple richness analysis in DIVA GIS based on point to grid analysis Steps 1 Start by visualizing two layers Vasconcellea data and Latin America Countries Then select the layer for the Vasconcellea species and go to Analysis Point to Grid Richness Wes DIVA GIS 7 3 0 Project Data Layer Map UEVAGS Modeling Grid Stack Tools Help oc Som Sa gern 3 HO afl g gt Point To Polygon Estimators of Richness v Latin America Countries Pointto Point Turnover am SH C Z Summarize Points Diversity aoa 7 Yasconcellea species Molecular Distance and Diversity Reserve Selection Statistics 2 S Distance gt Autocorrelation Lit Histogram Le Regression J Multiple Regression 2 Next define the properties of the raster that will be used for the analysis The dimensions of the study area as well as the resolution cell size must be determined In the Point to Grid window go to the Define Grid option and select Create a New Grid default option 3 Click on the Options window to define the raster properties origin and extent of the study and the resolution cell size 4 The study area can be defined using one of the following options a Inthe Options window manually enter the values for the X axis and Y axis you may wish to select the default options or b Draw the extension on the map using the Draw Rectangle tool Th
219. teristic ROC curve The AUC is equal to the likelihood that a randomly selected presence point is located in a raster cell with a higher probability value for species occurrence than a randomly selected absence point Species distribution modelling in this manual includes presence points only To still be able to calculate the AUC Maxent replaces absence points with randomly generated points from the study area The AUC can then be interpreted as the likelihood that a randomly selected presence point is located in a raster cell with a higher probability value for species occurrence than a randomly generated point Phillips et al 2006 The fractional predicted area on the X axis of the AUC graph is the fraction of the total study area where the species is predicted present while the sensitivity on the Y axis is the proportion of presence points in the modeled area of occurrence on the total number of actual presence points Phillips 2009 The highest predictive power of a model generated by Maxent is reached when the AUC has a value of 1 In practice no AUC will be lower than 0 5 which is similar to Random prediction In that case Maxent has no predictive power at all Araujo et al 2005 recommend the following interpretation of AUC for the models generated Excellent if AUC gt 0 90 Good if 0 820 gt AUC lt 0 90 Acceptable if 0 70 gt AUC lt 0 80 Bad if O 60 gt AUC lt 0 70 Invalid if 0O 50 gt AUC lt 0 60 In the case of this analysis AUC is
220. terozygocity or the distribution of locally common alleles for describing differences in intra specific diversity between subunits can be used as well Complementary to alpha diversity is beta diversity which focuses on divergence in species trait or allelic composition between different subunits of the study area For example to understand how genetic diversity is spatially structured subunits of the study area can be clustered based on the genetic similarity of cell composition see Section 5 3 In Section 5 4 reserve selection an analysis is presented on how to combine measures of both alpha and beta diversity to prioritize sites for conservation In the process of selecting areas for conservation emphasis is most often placed on conserving the highest number of species or alleles see Petit et al 1998 It is however important to realize that focusing conservation only on those sites with the highest levels of diversity may lead to a failure to identify threatened species found only at sites with generally low levels of diversity e g high mountain ecosystems which reveal a low number of species but where such species are unique and not found in other ecosystems Usually biodiversity studies assess the status of species or genetic diversity at a specific point in time While dynamic changes in biodiversity can be detected when site data is collected several times this topic is beyond the scope of the manual Spatial analysis o
221. the dataset Minimum Temperature tmin1_33 grd and tmin1_33 gri become tmin1 grd and tmin1 gri tmin2_33 grd and tmin2_33 gri become tmin2 grd and tmin2 gri tmin12_33 grd and tmin12_33 gri become tmin12 grd and tmin12 gri In the case of the dataset Maximum Temperature tmax1_33 grd and tmax1_33 gri become tmax1 grd and tmax1 gri tmax2_33 grd and tmax2_33 gri become tmax2 grd and tmax2 gri tmax12_33 grd and tmax12_33 gri become tmax12 grd and tmax12 gri Renaming of the 72 files might be facilitated by using batch rename software e g Rename Master After having changed the file names go to Data Climate Make CLM files Les DIVA GIS 7 3 0 Project BEE Layer Map Analysis Modeling Grid Stack Tools Help 3 DI aa Import Points to Shapefile Import Text to Line Polygon Polygon to Grid Assign Coordinates 3 Make CLM files Y Check Coordinates Export Gridfile Importo Gridfile Write YRT file Export Shapefile A File Manager Download Under File Browser select the folder where the Minimum temperature Maximum temperature Precipitation raster files grd and Altitude raster were saved All raster files grd in the indicated folder are shown at the right hand of the File Browser Indicate the raster file prefixes in the boxes T min minimum temperature T max maximum temperature and Prec precipitation The prefixes include all the characters before the number 1 12 For th
222. the position of raster corners number of columns and rows and cell size the GRI file contains the values for each raster cell and is therefore significantly larger When importing raster data to Maxent it is recommended to use the ASC Il raster file type which can be generated in DIVA GIS from GRD files An ASCII file consists only of one file asc file ESRI ArcGIS software has its own file types but can also import ASCIl files 2 1 Preparing and importing presence points As mentioned previously presence points can be compiled from data of vegetation inventories plant collection expeditions or publicly available sources like the GBIF Section 2 3 provides links to several information sources of this kind GPS Global Positioning System equipment is now widely used to georeference presence points Presence points can be organized in an Excel file and then converted into appropriate formats for spatial analysis using GIS programmes such as DIVA GIS or species distribution modelling programmes such as Maxent This section outlines the minimum required data for each presence point how to format this data in Excel and how to then import such data to DIVA GIS and Maxent 2 Towards the poles metric distances in the longitudinal east west direction will become shorter and become zero at the North and South Pole where one can walk around the world in only a couple of meters This manual will use cell sizes with degrees which are
223. the possibility to Conduct conservation actions in several areas Combining alpha and beta diversity analyses allows for the optimization of resources to conserve the greatest amount of diversity possible An area with the highest alpha diversity is usually prioritized as the first site for in situ conservation excluding any logistical constraints However a more complex question exists as to Spatial analysis of diversity for conservation planning what area should be considered as the second priority if additional resources remain available The second priority area should not necessarily be that with the second highest degree of alpha diversity as a large portion of this diversity may already be conserved in the first priority area In this case beta diversity must be taken into account focusing on areas with a species or allelic composition different from that of the area already prioritized This concept of complementarity is considered in DIVA GIS Conservation activities generally focus on including the highest number of species but may also target the conservation of a particular species In this case alleles are used as the observed unit of diversity to define priority in situ gene conservation areas This concept is outlined in the following analysis focusing on cherimoya PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 5 4 Conservation Strategies e Excel e SSR cherimoya rand c
224. tion combined with passport data information about the site where material was originally collected The data used in this analysis comes from a study on the diversity and genetic erosion of cassava Manihot esculenta in the Peruvian Amazon Ucayali Region Willemen et al 2007 In this section you will learn to use the Analysis Point to Grid menu in DIVA GIS to undertake an intra specific diversity analysis based on morphological data Spatial analysis of statistics of phenotypic data based on point to grid analysis Steps 1 Start by reviewing the DIVA GIS options outlined in Chapter 3 Basic Elements Create a map of Peru which shows three administrative levels country line width 3 regions line width 2 and districts line width 1 Add layers with rivers and bodies of water in blue roads in red and towns and villages in gray The resulting map should look like the following les DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack Tools Help Osi ea Qje Mol x SE WeLz E 0 a 7 PER_water_lines_dew PER_roads 7 Peru_Towns o 7 PER_water_areas_dew 7 PER_adm2 x 67 2776 y 0 2448 Scale 1 8617885 Data A Design Spatial analysis of diversity for conservation planning Now add the layer with the characterization data of cassava Manihot ex situ shp and zoom in on this data using the Zoom to theme option Pes DIVA GIS 7 3 0 Proje
225. tion models potential species distribution using the Bioclim algorithm In this manual however species distribution modelling is done with another method and software Maxent and the Predict option is therefore not explained in further detail Histogram T 10 11 The Histogram tool constructs frequency histograms which show the distribution of a species along customized ranges for different climatic variables Select the desired species to generate a histogram For this analysis select V cundinamarcensis Select the desired climatic variable to visualize the climate range frequencies where this species was observed In this case select Annual Mean Temperature Click Apply to display the histogram The width and number of bars of the histogram as well as the maximum and minimum values can be modified In this analysis 5 C and 30 C are defined as the minimum and maximum values in order to generate five bars at five degree intervals Species distribution modelling and analysis 12 The histogram can be copied and pasted directly into a document using the Copy option Distribution Modeling 5 Distribution Modeling Input Frequency Outliers Histogram Envelope Predict seJ A Reset Copy Angle 0 Barwidth Bar number cy IE Reset Copy Arigle T E Show values 40 5 Show values 5 o Environment Yariable Environment Yariable annus Mean Temper
226. ton allows you to calculate distance between different features The Table button allows you to access the table of attributes of a vector file these attributes can be explored with the Statistic Highlight Pan to and Zoom to buttons The table is a read only file and cannot be changed You must access the original file used to generate the vector file see 2 1 2 in order to modify the data It is recommended to prepare a new spreadsheet file with the modified data and not to alter the original file In this way if a modification is invalid you can still return to the original data Mev candicans 4 3201 FO7885 candicans 4 4283 3 7918 candicans 4 0916 9 933 candicans 4 0646 79 6411 candicans 4 0655 9 6476 candicans 4 0653 73 6411 candicans 4 423 73 7192 candicans 4 0697 9 9225 candicans 4 3181 9 S63 candicans 4 3263 79 7907 candicans 15 213 r2 0006 p Record 1 of 1000 Statistic Highlight Pan To Zoom To Statistic in the case of numeric values the Statistic button provides a summary in an additional window of the basic statistics of numerical variables Highlight highlights the presence point line polygon selected for a few seconds Pan to the presence point line polygon you wish to view may not be in the section Basic elements of spatial analysis in DIVA GIS of the map initially displayed This tool moves the map with the same zoom to t
227. tween most populations in Peru and Ecuador has disappeared as a result of the cutting On the other hand the cherimoya accessions in Bolivia are confirmed to be different from those in Ecuador and Peru Although the previous analyses indicated Bolivia had low levels of diversity the allelic composition of the materials in the country is quite distinct from that of the other study areas 10 When there are too many groups it is also difficult to identify patterns in the composition of alleles Cutting the dendrogram at a distance of 0 065 generates 11 groups of alleles Having too many groups as illustrated in the map above complicates the interpretation of results Despite the high number of groups the accessions in Bolivia continue to represent a homogenous group separated from the others 5 4 Implications for the formulation of conservation strategies The three previous analyses focused on using spatial analyses to detect areas of high diversity alpha diversity and to a lesser extent to understand differences in the diversity between areas beta diversity An understanding of the extent and distribution of diversity is critical to designing effective and appropriate conservation strategies It is also vital in order to identify key sites for carrying out ex situ and in situ conservation activities priority areas for collection and protection particularly since resources allocated for conservation are frequently scarce limiting
228. ty Information Facility GBIF is a platform providing public access to biodiversity data from national museums herbaria and genebanks worldwide In October 2010 the GBIF contained roughly 39 million georeferenced plant observations GBIF 2009 This training manual is intended for scientists professionals and students who work with biodiversity data and are interested in developing skills to effectively use spatial analysis programmes with GIS applications It has been designed to serve as a self teaching manual but may also be used for training courses The manual explains basic diversity and ecological analyses based on GIS applications Results of these analyses offer a better understanding of spatial patterns of plant diversity helping to improve conservation efforts The training manual focuses on plants of interest for improving livelinoods e g crops or crop wild relatives and or those which are endangered Inter specific and intra specific diversity analyses using different types of data are presented species presence morphological characterization data phenotypic data and molecular marker data data of DNA base pair compositions or molecular weights Although this training manual focuses on plant diversity many of the analyses described can also be applied when studying other organisms such as animals and fungi This manual has been published as a result of the increasing number of requests received by Bioversity Internationa
229. ty control miguel Fields of Shape of Point Fields of Shape of Polygon Relation 1 COUNTRY al es Relation 2 fl Relatinn 4 Y Wl Close 7 The first group of inconsistencies includes those points outside all the polygons Points outside all polygons tab In this analysis these points are located outside the polygon of Latin America This option is very useful for identifying points with unlikely locations such as a point of a plant species located in the ocean Check Coordinates Options Points outside all polygons Points do not match relations Points do not match with r SPECIES LATITUDE LONGITUDE COL Woot heilbornii 3 3331 50 8419 1 5667 Wo parviflora 1 5667 80 856419 80 68633 0 Wo parviflora 0 80 6833 ro 63da 1 6667 1669 Wo cundiname 1 6667 fo 6333 0 0 2967 Y cundiname 0 0 Row 0 of 0 Fan Ta oom To Export In the fourth column of the information displayed check the identification ID of data with errors With this information you can return to the original data file in Excel or in a dBase IV file dbf and check the details in order to determine how to correct the error The options Pan to and Zoom to see Chapter 3 also allow you to identify other points that may include mistakes These problem points can also be exported to a text file using the Export function Chapter 4 It is strongly recommended to document in detail all c
230. typical errors all apparent due to the location of points outside the study area Case A ID 2967 The point s coordinates have values of 0 0 which locates this point in front of the coast of Africa This error can occur in the database if there is an absence of data as some programmes automatically fill empty cells with a value of zero 0 Case B ID 319 The situation is similar to the previous one except in this case incorrect information is only present for one coordinate longitude In a database it is possible for one geographic coordinate of a presence point to be missing A question to consider What would happen if the value for latitude were zero 0 instead of for longitude Would the error also be as obvious and easy to identify Case C ID 1669 This point is far away from the study area Here the error has likely occurred due to the omitting of the negative sign for points located in southern latitudes or western longitudes resulting in incorrect positions on the map Chapter 4 a Identify s 0 13149 o 0 11016 Vasconcell Rec 1 of 1 Layer Yasconcellea final errors ID 2967 SPECIES Vo cundinamarcensis LATITUDE LONGITUDE 0 COUNTRY Colombia ADMI Cundinamarca ADM 2 _ Identify lt 0 29436 yo 3 31005 Vasconcell Rec 1 of 1 Layer Yasconcellea final errors ID 319 SPECIES VW x helborni LATITUDE 3 3331 LONGITUDE 0 COUNTRY Ecuador ADM Azuay 40M2 Mabon ADM
231. uercifolia 19 64583333 vasconcellea_bioclim C 111 C Fis 15 8 21 49583 V parviflora 17 31666667 Series Y values 16 98333333 vasconcellea_bioclim D 112 D 2602 3350 25 Final result Climate niches of 4 Vasconcellea spp V microcarpa E c s cs w a Lo D fa E gt c c T Mean Annual Temperature C Species distribution modelling and analysis The graph representing the two dimensional niches for the different Vasconcellea species should be similar to the graph above Differences in annual precipitation and mean annual temperature in the realized niches are clearly observed The V cundinamarcensis and V microcarpa niches are large in comparison to the V quercifolia and V parviflora niches suggesting that these species have adapted to a wider range of environments The V quercifolia realized niche is limited to temperate zones with moderate annual rainfall this is a species typical of the Interandean valleys in Bolivia and the V parviflora niche is limited to hot and drier environments coastal areas in northern Peru and Ecuador suggesting the species is well adapted to locations with high levels of environmental stress Similarly a multivariate niche with all 19 climatic variables can be used to describe differences between the species Multivariate analyses such as the Principal Component Analysis PCA are beyond the scope of this manual but do provide addit
232. umn as they are added Check or uncheck those boxes next to the layers you are interested in visualizing or hiding 3 Colombia is used here as an example to provide further detail during this analysis Add the layer with Colombia s departments COL_ADM1 shp Les DIVA GIS 7 3 0 Erma Project Data Layer Map Analysis Modeling Grid Stack Tools Help KACCA COL_ADM1 sC Latin America Countries x 94 7807 y 32 5098 Scale 1 42881984 lel x aqe eRs 4 Information on Colombia s administrative divisions can be obtained by selecting the newly added layer click on it and clicking on the table icon This opens the table information We can now find specific features in the shp file For example find out where the Casanare Department is located using the buttons Highlight Pan To and Zoom To a COL_ADMI1 COL COL COL COL COL COL COL COL COL COL i COLOHEIA COL COLOMEBL COLOMBIA COLOMBIA COLOMBIA COLOMEBL COLOMBIA COLOMBIA COLOMBIA COLOMBIA COLOMBIA COLOMBIA Amazonas Antioquia Arauca Atlantico Bolivar Boyaca Buenaventur Caldas Caqueta Casanare Cauca Cesar Record 1 of 33 Statistic Highlight Pan To Zoom To 5 Now hide uncheck the layers of Vasconcellea species and remove the layer of the Colombian departments Expand the map so it shows all Latin American countries Basic elements of spatial analysis in DIVA GIS To change a la
233. us visualization possibilities However when using Google Earth there is the risk of making high resolution interpretations with data of low precision For example you may Know you have georeferenced species locations to a precision of 5 km while the resolution of the Google Earth images may be as high as 1 x 1m Trying to relate the species dataset to the Google image and its resolution would be meaningless in this instance 1 3 1 How to install Google Earth The installation of Google Earth is very straightforward and consists of downloading the installer opening it and following the on screen instructions In order to be able to run Google Earth you must be connected to the internet i e online Installation of software and example data for analysis Steps 1 Download the Google Earth installer from http earth google com after accepting the conditions Google Earth Windows Internet Explorer ep Je SB ntp www google comiearthyindex htm x B File Edit View Favorites Tools Help A D v Pager Safety Toos v Google earth English US vp Favorites i Google Earth Home Explore Download Learn Connect Help Get the world s geographic information at your fingertips Download Google Earth 5 e Fly to any place around the world e See 3D buildings imagery and terrain e Find cities places and local businesses Become a World Traveler Google Earth Pro Tr
234. ut file type Product features Threshold features Hinge features Output directory Auto features Projection layers directoryfile Settings ci ca a File Name Vasconcellea csv onereea 6b Flesoftype eves OOO ven e canca Steps for preparing a Comma Separated Values CSV file for Maxent in Excel 2007 1 In anew sheet copy the data for the variables Species Longitude and Latitude in this specific order 2 Check that decimals in Excel are separated by points a Go to Office Button Excel Option b Under the Advanced tab select points to separate decimals c Under the Advanced tab select commas to separate thousands 3 Save the CSV file csv and open it using Notepad Preparing and importing data to DIVA GIS and Maxent 4 Make sure the different data are separated by commas if this is not the case make the necessary changes This can be done using the Replace function in the Edit Replace menu a Delimited by semicolons b Replacement by commas c Remember to save the changes made to the file d Finally follow Step 6 in the section above for Excel 1997 2003 to open the CSV file csv file in Maxent e vasconcellea csw Motepad File E eee ee File KK SMR EK KR KEKE ESS Edit Format Edit Format candicans 7G candicans 79 candicans 79 candicans 79 candicans 79 cand
235. utions Resolution Miris 1065 0 083333332 0 083333332 121 1250001 32 3750000 56 00000001 34 58333325 a e gis tutorial 6 4 gap analysis v _caulil 1087 1065 0 083333332 0 083333332 121 1250001 32 3750000 56 00000001 34 58333325 e gis tutorial 6 4 gap analysis v _crass 1087 1065 0 083333332 0 08333333 121 1250001 32 37 50000 56 00000001 34 58333325 e gis tutorial 6 4 gap analysis v _cund 1087 1065 0 083333332 0 083333332 121 1250001 32 3750000 56 0000000 34 58333325 e gis tutorial 6 4 gap analysis v _glanc 1087 1065 0 083333332 0 083333332 121 1250001 32 3750000 56 0000000 34 58333325 e gis tutorial 6 4 gap analysis v _goud 1087 1065 0 083333332 0 08333333 121 1250001 32 3750000 56 0000000 34 5833332 e gis tutorial 6 4 gap analysis y _micrc 1087 1065 0 083333332 0 08333333z 121 1250001 32 3750000 56 0000000 34 5833332 e gis tutorial 6 4 gap analysis v _mone 1087 1065 0 083333332 0 08333333 121 1250001 32 3750000 56 00000001 34 5833332 e gis tutorial 6 4 gap analysis v _parvi 1087 1065 0 083333332 0 08333333 121 1250001 32 3750000 56 0000000 34 5833332 E GIS Tutorial 6 4 Gap analysis vasconcellea thresholds grs AA Appl A close Assign a name to the stack click Apply and then Close Chapter 6 7 To estimate the potential richness of Vasconcellea species in Latin America go to Stack Calculate this sums the stack of binary rasters representing the potential area
236. ve unit information climate land cover UTM coordinates can be converted into the latitude longitude system in decimal degrees by using Excel calculation sheets these are available on the internet from sites such as http www uwgb edu dutchs UsefulData UTMFormulas htm Preparing and importing data to DIVA GIS and Maxent The Projection option in the Tools menu of DIVA GIS also allows you to convert lat long formats into UTM or vice versa PROGRAMMES AND DATA FILES TO USE IN THIS SECTION Programmes Data Files e DIVA GIS Folder 2 1 Importing observation data e Excel e Vcundinamarcensis_DMSdata xls e Maxent and Java e Vasconcellea x s 2 1 1 How to convert DMS data into DD format This exercise uses the Vcundinamarcensis_DMSdata xls Excel file as an example the file can be found in the data folder accompanying this chapter and consists of seven spreadsheets in which DMS coordinates are calculated stepwise into DD Steps 1 Open the Vcundinamarcensis_DMSdata xls Excel file and select the Start data spreadsheet Note that the latitude and longitude coordinates are presented in DMS 38 Microsoft Excel Vcundinamercensis_DMSdata xls a File Edit View Insert Format Tools Data Window GenAlEx Help Adobe PDF B G D E D LATITUDE LONGITUDE COUNTRY ADMI 3433 V cundinamarcensis 6 57 48 N 7502503 W Colombia Antioquia 3037 V cundinamarcensis 7 10 17 N 75 45 47 N Colombia Antioquia 2816 V cundinamarc
237. vement of the manual s content Finally the production of this manual would never have been possible without the financial support of INIA Spain through the project Strengthening Regional Collaboration in Conservation and Sustainable Use of Forest Genetic Resources in Latin America and Sub Saharan Africa and of the Austrian Development Cooperation through the project Developing training capacity and human resources for the management of forest biodiversity We would like to acknowledge Ricardo Alia and Santiago Martinez of INIA for their positive and encouraging comments throughout the development process Introduction Plant diversity is vital for the survival and well being of humanity A number of domesticated plant species are critical to global food security while other species are of great importance for purposes such as wood and biofuel production In addition to the cultivated species many wild plants still play an important role in meeting local needs for food fuel medicine and construction materials crop wild relatives are also of special interest for crop breeding programmes There are currently hundreds of underutilized plant species and varieties displaying traits of interest to meet present and future needs while the value of many other plant species is yet to be discovered The Convention on Biological Diversity CBD established in 1992 calls for a global strategy for plant conservation CBD 2009a In ad
238. ves Make pictures of predictions Quadratic features Do jackknife to measure variable importance __ Output format Logistic La Threshold features ri i Hinge features Product features oo Output directory E GIS Tutoriali6 2 Potential distribution v Auto features Projection layers directoryffile Settings In order for Maxent to process the environmental rasters all rasters must have the same parameters in terms of their properties resolution and coordinates of corners or vertices If this is not the case Maxent will generate an error and will not be able to run Note Maxent only uses the information in the first three columns see 2 1 3 If more columns Fields are included in the CSV file csv the Visual warnings sign will be displayed automatically In this case check OK in each field Alternatively select the Suppress similar visual warnings option As mentioned in Step 3 Maxent also allows one to select the raster output files in grd format The advantage of using grd files as compared to raster files in ASCII format asc is that these can be opened directly in DIVA GIS However errors have sometimes been observed in Maxent when using grd output files Therefore using ASCII output files is recommended in this manual Maxent will read all raster files in the folder selected as the Directory File under Env
239. w to import climate data to DIVA GIS from CLM files clm DIVA GIS uses the CLM format to store and read spatial climate data Steps 1 Copy or extract climate data from the compressed file diva_worldclim_2 5min zip to a folder on your computer In this example the data are saved using the file path C Program Files DIVA GIS environ 2 Goto Tools Options in DIVA GIS and select the Climate tab ws DIVA GIS 7 3 0 Project Data Layer Map Analysis Modeling Grid Stack BE Help EA TCawEE bwivie a Projection Graticule Shift Shape Georeference Image Geo Calculator Toolbars 3 Under the Folder box indicate the location of the folder containing the climate data you would like to import In this example the data are found in the folder C Program Files DIVA GIS environ 4 Indicate worldclim_2 5m as the climate database 5 Click Apply and Save to make this the default database for the analysis of climate data in DIVA GIS 6 Click Close to close the Options window Chapter 2 7 Return to Jools Options Climate to view the climate data selected and to check if the climate data have indeed been added Options Layer Climate je Folder E SGIS Tutorial 2 2 Importing climate data diva_wordclim_2 5min o worldclim_2 5rn Columns A ows E40 S600 a Min 180 i Cell size 0 04 EGGEEEEF Mas 180 Index indes_z 5m Altitude alt _2 5m Min Temperature trin_2 5m Mas
240. w vector file In the Data menu click on Selection to New Shapefile to save the selection Vasconcellea species in Ecuador as an additional layer which can now be seen on the legend bar Hide this new layer Import Points to Shapefile Import Text to Line Polygon Draw Shape Polygon to Grid Points to Convex Polygon Selection to New Shapefile Extract Values by Points Climate Assign Coordinates W Check Coordinates Export Gridfile Importto Gridfile Write VRT file Export Shapefile W W File Manager Download 16 Sometimes it is important to remove selected features To do this select the layer which holds the selected features and go to Clear Selection in the Layer menu or click on the corresponding button in the toolbar Pa The selected points in yellow are no longer highlighted SS DIVA GIS 7 3 0 Project Data PE Map Analysis Modeling Grid Stack Tools Help Co ae ta Laer nek x SE Ola 5 X Remove Layer Latin Americe Properties abc Add Labels identify Feature Table Filter 3 Select Records i Select Features Vasoncellea oe Clear Selection Copy Alt C E Paste Alte Z Hide Show Legend Chapter 3 17 Now make a new selection by selecting the points directly on the map using the Select Features option in the Layer menu For this analysis continue to work with the Vasconcellea species l
241. y To Find Businesses Directions Fly to e g 37 25 19 1 N 122 05 06 W Qj v Places 4 vasconcellea b UE v candicans gt OUO v cauliflora v crassipetala v cundinamarcensis rt v glandulosa v goudotiana v microcarpa b LIE v monoica v parviflora v pulchra Ww v quercifolia gt LE v sphaerocarpa D v stipulata gt UIE v weberbaueri gt UE v x heilbornii WIS pkesiya_cut Layers Earth Gallery gt gt gt ES primary Database EA t ATi gt gt Google 2010 ee ae i r 17 06 15 00 N 109 30 2 v Eye alt 3243 81 mi gt 3 3 Editing maps and finalizing a project If you wish to display maps in documents or reports you can customize the combination of generated layers under the Design option located in the lower right hand corner of the screen Selecting this option allows you to carry out basic editing of the current visualization e g adding scale adding north arrow adding text es DIVA GIS 7 3 0 Oia a alee Map Jf OK Location x0 yi0 Cut white space v Right v Bottom Scale 1 41882404 456 Y 0 Data h Design Chapter 3 After customizing the map the image can be saved as a png tif or omp file or can be copied directly to the clipboard The Map menu under the Data tab includes the Map to Image option which allows you to save the edited m
242. y identified sites administrative units or raster cells of any chosen size In this manual raster cells are used to represent subunits of diversity The advantage of using raster cells is that these allow the comparison of species trait or allele presence absence between subunits of similar geographical size throughout the extent of the study area In some cases vector data is used for example to compare the number of species between different countries The most direct measurement of alpha diversity results from counting the number of observed diversity units e g the number of species or the number of alleles per subunit of the study area Referred to as richness this type of measurement is straightforward and fairly easy to interpret see Sections 5 1 and 5 3 The analyses in this chapter will focus mainly on this type of measurement as it is widely used to assess diversity at the species level The richness of alleles is also considered a key measurement for analyzing the conservation of the genetic diversity of a species of interest Frankel et al 1995 Petit et al 1998 A drawback of the richness measurement however is that it depends on the number of samples taken within each subunit of the study area It is common to find higher levels of diversity in instances where many samples have been collected while under sampled areas often appear to have lower levels of diversity such results are not always accurate Chapter 5 Hijma
243. yer s legend double click on the layer to be modified For this analysis click on the layer Latin America Countries The Properties window will immediately be displayed showing the Single tab as the first option This window will allow you to make changes to the legend The Single tab displays a transparent rectangle A double click on the rectangle opens the Symbol window The attributes of the Symbol option allow you to change certain properties of the polygons a Outline thickness b Outline colour c Polygon filling style d Polygon colour Try to change the default values of these properties and observe what happens after clicking OK in the Symbol window and Apply in the Properties window For the best visualization the recommended filling style is Solid Fill Properties j Label e Latin America Countries ears e ais tutonaly3 7 basic elements latin america count Type Polygon 117 30 32 72 29 30 56 11 Single Preview AA Apply Kl Clase M Cancel It is best to use different colours for different attribute values In the Properties window select the Unique tab Under the Field tab select the field to be used to create the layer s new legend It is important to select a field with a limited number of classes If you use a field containing numeric values e g area you may cause DIVA GIS to hang or lag Since the programme will take each value as a different class For this exercise

Training manual on spatial analysis of plant diversity and distribution

Contents

Download Pdf Manuals

Related Search

Related Contents