unicor user manual - Computational Ecology Laboratory
Contents
1. 1 3 What s new in Version 2 0 Similar to connectivity thresholding we have added a resistant kernel approach for predicting habitat connectivity and corridor paths for the given resistance surface as conducted by Compton et al 2007 Resistant kernel connectivity modeling has a number of advantages as a robust approach to assessing population connectivity for multiple wildlife species First unlike most corridor prediction efforts it is spatially synoptic and provides prediction and mapping of expected movement rates for every pixel in the study area extent rather than only for a few selected linkage zones Compton et al 2007 Second scale dependency of dispersal ability can be directly included to assess how species of different vagilities will be affected by landscape change and fragmentation under a range of scenarios e g Cushman et al 2010a Third it is computationally efficient enabling simulation and mapping across vast geographical extents for a large combination of species e g Cushman et al 2010b Cushman et al 2011 The resistant kernel approach to connectivity modeling uses the framework of the modified Dijkstra s algorithm Instead of calculating one shortest path derived from source to source nodes the resistant kernel approach builds a least cost dispersal around each source cell The resistant kernel approach implemented here uses the Dijkstra s2. xpected species movemen guidance on identifying maintaining network con fragmentation and corri graph theory metrics e Hagberg et al 2008 a into popular landscape p McGarigal et al 2002 The program is written i module plug in docking modular development The name value pairs ina st are parsed using the Rip users may specify alternative connectivity scale dependency of dispersal ability will ange and fragmentation under a range of et al 2010a Watts et al 2010 Outputs tat patches that can be used to display t routes and can provide managers with visual corridors that are likely critical for ectivity Quantification of changes to habitat dor connectivity is enabled through outputs of g density number of nodes radius etc d connectivity outputs that can directly input attern analysis programs e g FRAGSTATS r be r n Python 2 6 UNICOR is built on a driver architecture that allows for ease of future program s input parameters are organized as anza oriented text file format The inputs Mgr package a flexible symbol table manager UNICOR user manual 6 for science models that includes special parsing capabilities Glassy 2010 UNICOR has been debugged as carefully as possible by testing all combination of simulation options The program is freeware and can be downloaded at http cel dbs umt edu software UNICOR
3. 2 pixel values around each path is used to create the kernel density buffer SciPy Number of Levels For kernel density estimation of normal paths this is the number of categories used to display the kernel density buffer map If 3 then this option will take the kernel density buffer created and categorize the values into 3 equal interval classes low medium high SciPy Output Default Example Description Dependency UNICOR user manual 13 Input Save Path Output TRUE The surface of paths in ascii format with header file space delimited addpaths Save Individual FALSE The list of individual path values and Paths Output length from every point to every other point comma delimited paths This can be a very large file depending on the size of your problem Save Cost TRUE The resistance distance matrix of all Distance Matrix the source destination connection Output lengths comma delimited cdmatrix Save KDE Output FALSE The surface of kernel density buffered paths in ascii format with header file space delimited kdepaths Save Levels FALSE The categorical surface created from the Output KDE output in ascii format with header file space delimited levels Save Graph FALSE Path graph theory metrics comma NetworkX Metrics Output delimited graphmetrics 5 General is
4. E modified function to produce a map of movement cost from each source up to a given specified threshold Each source cost map is then inverted and scaled with a given transformation function such that the maximum value for each individual kernel is one Once th xpected density around each source location is calculated the kernel maps surrounding all sources are summed to give the total expected density at each location on the landscape The results of the resistant kernel approach are surfaces of expected density of dispersing individuals at any location in the landscape 2 Getting started 2 1 Dependencies 2 1 1 Baseline Requirements UNICOR requires the Python2 6 x interpreter or greater but less than Python3 0 NumPy package and SciPy package Several optional Python module packages if enabled facilitate additional UNICOR functionality Remember that Python modules usually require particular Python interpreters so be sure the version ID for any external Python module or package e g NumPy or others matches the version of your Python interpreter normally v2 6 x 2 1 2 Python on Non Windows Platforms UNICOR user manual 7 Some common computer platforms come with Python installed These include MAC OS X and most Linux distributions To determine which Python a MAC or Linux workstation has installed start a terminal console and ente
5. LLA TURo A a a LAA ag e 8 Bi DW PUE A A AA a le ates AD weird gy Pe Satie A A Sek Res 9 Bi Resistance Grud wise de pedi OAM eee ad tae E hae oe os 9 Sil XY LOCA CTO S s gael tbe Bid i Se cared yore Sah one aes Linge SSA 9 3237 Types of paths wine aide ok A Sah seed 10 3 4 37 Diykstha s shortest pablito wee LO 3 3 2 Kernel density estimation on shortest path 10 3 33 THRESH OUGLNG oe erra tas Sind eee wate eG a LO 3 3 4 Resistant kernel connectivity 10 324 Number oF PrOCeSSOLSis ee roen a ec bane Wile ees ee 11 A gt OWE Dd Ges daa Ri amp age See atten soe ahs 11 5 GESNSTCIT SIESTA AA DA AAA o EA 13 5 1 HOw to Obtain UNICO Reca ais a hala nai BS 13 EZ Debugging and troubleshootinG oooooooooo nooo ee eee 13 523 UNTCOR Limitations As Oh ek Bet hi DA o a 13 5 4 HOW bOrsCi Ces UNICO Ri a a Sune eee ida 13 5x9 Disclaimer ma Sey Rates ee Rade tek dude Sine sil 13 6 RETSESMCES ADA A A rs 14 T EEKnNOWLeagements ais si ss a eee adas 16 1 Introduction Habitat loss and its effects on populations of vulnerabl among the most urgent problems in conservation ecology ective tools to evaluate th on the extent and conn we introduce UNIversal that managers and scientists have eff effects of landuse and climate change To address this need F of populations UN F network simulator COR a species identification tool movement corridors UNICOR applies Dij to
6. individual based simulations incl ding and key features with threshol calculation of graph theory metrics efficiency is greatly improved computational buffering kernel large extents individuals 1 1 What can UNICOR UNICOR is inte com munity and will be It provides new connectivity i invaluable abi conservation a the largest cum can be used to populations a F 1 ded for use by land ma d identify corridors a grid dimensions of thousands in the thousands do a valuable tool Through paral ndscapes UNICOR user manual 3 qu ation connectivity an allowin This speci TELAS C s is itical e ectivity CORridor connectivity and corridor kstra s shortest path Outputs can be used to desi identify isolated populations climate and management changes on popul prioritize conservation plans to maintain population connectivity ude a driver module framework gnat effe d antify g analyses algorithm e cts of The connectivity mapping resistance thresholding llel processing and of and large populations nagers as well as the research in applied conservation biology in turn unctionality to increase understanding of species current and future la provides ulation connectivity source or ity to quantitatively compare spat
7. py small _test rip 6 Check for successful simulation run completion The program will provide a log file in your UNICOR home directory Once completed output files will be created in UNICOR home directory See section 4 for the description of outputs 3 Input See Table 1 for UNICOR inputs and outputs 3 1 Resistance grid Prior to running UNICOR users must create a resistance surface where each cell value pixel represents the unit cost of crossing each location Pixels are given weights or resistance values reflecting the presumed influence of each variable to movement or connectivity of the species in question Resistance surfaces could be parameterized to reflect different costs to movement associated with vegetation types elevation slope or other landscape The filename for the resistance surface must be in an ascii format with header file any file extension is acceptable must be space delimited The example simulation run resistance grid is small test rsg and will provide format example 3 2 XY locations Point locations define starting and ending nodes of paths connecting pairs of individuals The points must be referenced on the landscape resistance surface with any user specified placement pattern e g uniform random or placement in habitat suitability and density The filename for the individuals with x y locations can have any file exten
8. scipy org Download This will install NumPy and SciPy in your Python site packages directory Note if you used Enthought as a distribution you already have NumPy and SciPy installed and do not need this step 2 2 2 Unpack the UNICOR Archive Navigate to the directory on your PC where you wish to install UNICOR and unpack the supplied zip archive file using a free archive tool like 7Zip 7z exe Pkunzip Unzip or an equivalent Seven Zip 7Z exe is highly recommended since it can handle all common formats on Windows MAC OS X and Linux On Windows it is best to setup a project specific modeling subdirectory to perform your modeling outside of any folder that has spaces in its name like My Documents 2 2 3 Install UNICOR Next install the UNICOR software itself by unpacking the zip archive supplied At this point you should be able to execute the supplied test inputs Alternatively and only for Windows operating system you may double click on the unicorn setup ex xecutable file for an automatic download 2 2 4 Optional Python Extension Modules As UNICOR is supplied in the archive it does not require any additional contributed Python modules to run However several additional Python modules are needed if you want the following functionality NetworkX is required for graph theory metrics and can be obtained from http networkx lanl gov wxPython is required to run the G
9. UI and can be obtained from http www wxpython org 2 3 Example run 2 3 1 Command line run The example run is for 10 points representing individuals on a simple landscape resistance surface To run the following example follow these steps 1 Locate the directory that you installed UNICOR to and open the unicor folder 2 The included rip file specifies the parameters that can be changed and used in the sample UNICOR run Open small test rip in your editor of choice e g notepad or wordpad to view these inputs 3 This file is the stanza format following RipMgr documentation All UNICOR user manual 9 signs are comments followed by variable names with a tab to the parameter specified The parameter can be changed for running UNICOR but downloaded parameters will run as is See section 3 for more details on each parameter along with its dependency 4 Start the program at the command line If you use Python from the command line then open a terminal window and change your shell directory to the UNICOR home directory For example in Windows you can go to the Start Menu gt and open cmd exe by searching for it or locating it in Accessories After gt type cd C LOCATIONOFYOURINSTALLATION 5 Run the program There are a number of ways to run this program For example if you are using a command shell you can run the program by typing python UNICOR
10. UNICOR user manual 1 UNICOR USER MANUAL Version 2 0 Updated 2012 03 22 Authors E L Landguth B K Hand J M Glassy S A Cushman 1 University of Montana Division of Biological Sciences Missoula MT 59812 USA 2 Lupine Logic Inc Missoula MT 59802 USA 3 U S Forest Service Rocky Mountain Research Station 2500 S Pine Knoll Dr Flagstaff AZ 86001 USA UNICOR user manual 2 Table of Contents Tk ENE LOOUCE UO NA A A a eis en tha 3 tel What Can UNITCOR CO secede na seme rape Ze ld 3 Led How does UNICOR WOOL Ks jew antes uaea sd ee a 3 1 34 Whats new Tn Version AD oi Bale Ee lee a 6 2 Getting Started mestasi it a dia Ven A A te as Se Doe aS 6 2 1 DEPENGEN SUES esis tri keina E a o E ee we e a 6 2 1 1 Baseline requirementS ooooooooooo eee ee eee 6 2 1 2 Python on non windows platforms 6 Field PYtAON ON MIEDOS ds AA a hehe es AE Sars Ge 6 2 1 4 Obtaaning INUOMPy and SCALP Y ias aliada e seas 7 2 2 Installation set als e ale SAN Ghd tte thd oe 0d 7 2 2 1 Installing Python NumPy and SciPy 7 2 202 Unpack the UNLCOR archives diern ia Sethe as 8 Liza Las tall UNA COR wetter ia A A te DS A Hale 8 2 2 4 Optional Python extension modules 8 230 Example LUNA LS Wee OS Shera OF Ge Oe are Ww Odi EOS 8 2 ros SOmman Gs
11. and corridor network simulator Ecography 34 1 6 5 5 Disclaimer The software is in the public domain and the recipient may not assert any proprietary rights thereto nor represent it to anyone as other than a University of Montana produced program version x x UNICOR is provided as is without warranty of any kind including but not limited to the implied warranties of merchantability and fitness for a particular purpose The user assumes all responsibility for the accuracy and suitability of this program for a specific application In no event will the authors or the University be liable for any damages including lost profits lost savings or other incidental or consequential damages arising from the use of or the inability to use this program We strongly urge you to read the entire documentation before ever running UNICOR We wish to remind users that we are not in the commercial software marketing business We are scientists who recognized the need for a tool like UNICOR to assist us in our research on landscape ecology issues Therefore we do not wish to spend a great deal of time consulting on trivial matters concerning the use of UNICOR However we do recognize an obligation to provide some level of information support Of course we welcome and encourage your criticisms and suggestions about the program at all times We will welcome questions about how to run UNICOR or interp
12. andscape resistance s COR xel Pixels are given nce ues 20 nt r et al F on slope F of tion 10 costs or UNICOR user manual 4 other landscape features UNICOR Conceptual Workflow e DEFINE PROBLEM Step 1 Assemble Step 2 Set Input Geospatial Inputs Parameters Landscape Resistance Surface Point oa ES Step 5 Synthesis Figure 1 UNICOR conceptual workflow diagram Steps 1 and 2 define the inputs and problem Steps 3 and 4 execute the program Step 5 provides synthesis and post processing Point locations define starting and ending nodes of paths connecting pairs of individuals The points must be referenced on the landscape resistance surface with any user specified placement pattern e g uniform random or placement in habitat suitability and density From graph theory and network analysis we can then represent the landscape resistance surface as a graph with nodes and edges Diestel 2010 Every pixel is considered to be a node The graph edges which represent possible movement paths between each node are weighted by the resistance value of the cell times the distance to the next pixel center times the distance to the next pixel center which gives the total edge length in terms of raster cell units cost distance Dijkstra s algorithm is modified in the UNICOR code to find all shortest paths to all destination nodes a
13. ce paths with no thresholding threshold given by Apply the threshold value Edge Distance to all shortest cost distance paths all paths Calculate resistant kernels for all XY locations and use the Edge Distance value as a threshold for each source resistant kernel Transform function Linear Specify the scaling away from the sources point used in the resistant kernel model linear Scale resistant kernel values to be between 0 and 1 using a linear function inverse square Scale resistant kernel values to be between 0 and 1 using an inverse square function vol Constant kernel Tru Keep resistant kernel volume constant for each XY location and equal to 1 Option True or False Kernel Volume 10000 Set the resistant kernel volume to this value when Const kernal vol is False Number of Processors For parallel computing the number of processors that are used in a simulation Linux KDI El Function Gaussian For kernel density estimation of normal paths this is the probability distribution used to calculate the kernel density buffer Gaussian Epanechnikov Uniform Triangle Biweight Triweight Cosine SciPy KD GridSize GA For kernel density estimation of normal paths this is the kernel buffer window used to calculate the buffered maps For example if 2 then
14. ed as the maximum ability This enables UN dispersal abilities of a maximum dispersal distan Euclidean distance Dijkstra s base algorith individuals However th thresholds These connectivity thresholds are path length for a species given its dispersal ICOR to realistically reflect the biological particular species Users can specify the based either on cumulative cost distance or E m assumes the optimal is followed by all is is unlikely to realistically represent the behavior of organisms T multiple low cost paths density function such as Pinto and Keitt 2008 the application of a var kernel density functions biweight triweight and density estimations Li by the program show the by kernel density estima following a distribution hus it is beneficial to consider either or to smooth output paths using a probability a Gaussian bell curve Cushman et al 2008 UNICOR implements the latter and allows for iety of smoothing functions referred to as Gaussian Epanechnikov uniform triangle cosine functions can be used for the kernel amp Racine 2007 The outputs that are produced cumulative density of optimal paths buffered tions see Silverman 1986 Scott 1992 around frequency of common connections Through batch capability thresholds to assess how affected by landscape ch scenarios e g Cushma include paths among habi
15. ey and Sons New York Schwartz M K et al 2009 Wolverine gene flow across a narrow climatic niche Ecology 90 3222 3232 Silverman B W 1986 Chapter 3 In Density Estimation for Statistics and Data Analysis Chapman and Hall New York Spear S F et al 2010 Use of resistance surfaces for landscape genetic studies Considerations for parameterization and analysis Molecular Ecology in press Urban D and Keitt T 2001 Landscape connectivity A graph theoretic perspective Ecology 82 1205 1218 Watts K et al 2010 Targeting and evaluating biodiversity conservation action within fragmented landscapes an approach based on generic focal species and least cost networks Landscape Ecology 25 1305 1318 7 Acknowledgements This research was supported in part by funds provided by the Rocky Mountain Research Station Forest Service U S Department of Agriculture and by the National Science Foundation grant DGE 0504628
16. frequency dispersal abilities users connectivity thresholds are species given its dispersal to realistically reflect Users can ither on cumulative cost based To set the threshold for and us make sure that Euclidean distance in the field cost units fied Edge Distance a speci is TRUE Use and ED Threshold ED Distance T 3 3 4 Resistant kernel connectivity Similar to kernel density estimation modeling does not assume an optimal path produced for each source point and added together to giv resistant kernel connectivity Instead kernel maps are th xpected density of dispersing organisms at any location on the landscape E surtace F function set field Edge See Compton et al Edge Type Distance note that Edge Distance Transform function 2007 for more details To use this to can be specified kernel is scaled A all paths to calculate a resistant is in cost units Constant Kernel Volume This function will use the kernel to a threshold In addition a for how each source s resistant can be enforced as well A constant kernel volume coul species For example UNICOR user manual 11 ld be used if you are comparing different a mobile species can travel farther producing a larger resistant kernel tha
17. ially explicit d restoration scenarios and prioritize actions to have ulative effects on pop designate sites as potential sink address prioritizing areas of greatest concern nd barriers change on wildlife populations climate or landuse change 1 2 UN species location on a landscape to every other sp provides a step by step conceptual location Figure How does UNICOR work ICOR simulator uses a modified Dijkstra s al to solve the single source shortest path prob em requires two input fil les as the first step surface and 2 point locations for each popul locatio see section users m represe weights n or e g Dunning et al Resista to movement associated with vegetation types ust create a resistance surf nts the unit cost of crossing each location reflecting the presumed influe each variable to movement or connectivity of the species in q resistance values 3 for program input e 1992 Cushman et al ee lgorithm rom every specified Dijkstra GE 1 a ation or ied species workflow individua Prior to running UNI ace where each cell val 2006 Spea nce surfaces could be parameterized to reflect differe elevati lue pi The results Simulations could effects of climate or habitat fragmentation under future 1959 UNI COR l
18. lgorithm assumes the optimal However ndividuals this is unl Thus ultiple nd al E niform i behavior of organisms m low cost paths density function a lows for the application of referred to as kernel density functions u triangle triweight or to smooth pl biweight used for the kernel density program show the cumulative density of optimal kernel ikely to realistical it is bene such as a Gaussian bell path is followed by all ly represent the icial to consider either output paths using a probability UNICOR implements the latter va a variety of smoothing functions Gaussian Epanechnikov and cosine functions can be The outputs that are produced by the paths buffered by of common connections 3 3 3 To re 7 Thresholding ect species specific differences in These can specify connectivity thresholds expressed as the maximum path length for a ability in cost units dispersal the biological abil F specify the maximum dispersal Euclidean distance y Edge Type is in cost units or your resistance grid To set the threshold for Euclidean distance 1 distance or specif that E specify a distanc to threshold This enables UNICOR ities of a particular species density estimations following a distribution around
19. n a less mobile species If constant kernel volume is enforced then the volume of the kernel is essentially population size and species that have different mobility can be modeled at the same population size When const kernel vol is False then the kernel volume parameter is used on the transformed kernel resistant distance following kernal volume 3 math pi kernel resistant distances 2 When const_kernel vol is True then no volume transformation is applied 3 4 Number of processors In essence this approach becomes a pes large graph problem for the conservation biology problems faced today In analyses involving large numbers of individuals across a large and fine grained environment computational time becomes intractable However parallel processing allows for efficient use of increasingly ubiquitous modern multi core processors Dijkstra s breadth first search algorithm is ideal for running in parallel for sets of source and destination points because pairwise distances can be calculated independently We have implemented paral lel processing in UNICOR using the multiprocessing module from Python version 2 6 and is currently only available in the Linux operating system 4 Output See Table 1 to specify formatted surfaces for matrices and optional For parallel comp
20. r python You ll see the version number on the top line enter Control D to exit Replacing an older Python interpreter pre v2 4 with a newer one v 2 6 x on a Linux or MAC OS X machine can be tricky so ask a System Administrator for help if you re not sure which packages depend on the current Python installed 2 1 3 Python on Windows Windows 7 XP 2000 Server does not come with Python installed so follow the instructions below to obtain and install Python ona computer running the Windows operating system Get a windows installation of the base Python installation current v 2 6 x at http www python org download releases or see more details on other python distributors below e g ActiveState or Enthought 2 1 4 Obtaining NumPy and SciPy We recommend using the superpack Windows installer available from the SourceForge website http sourceforge net project Note that more complete information for NumPy is available at www scipy org where the SciPy module is also presented Another source is http www enthought com products epd php for a free academic and educational usage in a single downloadable installer that has everything and then some Python Numpy Scipy Matplotlib and 70 modules for python 2 2 Installation 2 2 1 Install Python NumPy and SciPy Make sure that Python and NumPy are installed and available to you You can test this by typing python a
21. ret the output only after you have read the entire documentation This is only fair and will eliminate many trivial questions Finally we are always interested in learning about how others have applied UNICOR in ecological investigation and management application Therefore we encourage you to contact us and describe your application after using UNICOR We hope that UNICOR is of great assistance in your work and we look forward to hearing about your applications 6 References Bunn A G Urban D L and Keitt T H 2000 Landscape connectivity A conservation application of graph theory Journal of Environmental Management 59 265 278 Compton B et al 2007 A resistant kernel model of connectivity for vernal pool breeding amphibians Conservation Biology 21 788 799 Cushman S A et al 2006 Gene flow in complex landscapes testing multiple hypotheses with casual modeling The American Naturalist 168 486 499 Cushman SA McKelvey K S and Schwartz M K 2009 Use of empirically derived source destination models to map regional conservation corridors Conservation Biology 23 368 376 UNICOR user manual 15 Cushman S A Chase M J and Griffin C 2010a Mapping landscape resistance to identify corridors and barriers for elephant movement in southern Africa In Cushman S A and Huettman F eds Spatial complexity informatics and wildlife con
22. servation Springer Tokyo pp 349 368 Cushman S A Compton B W and McGarigal K 2010b Habitat fragmentation effects depend on complex interactions between population size and dispersal ability Modeling influences of roads agriculture and residential development across a range of life history characteristics In Cushman S A and Huettman F eds Spatial complexity informatics and wildlife conservation Springer Tokyo pp 369 387 Dale V H et al 2001 Climate change and forest disturbances BioScience 51 723 734 Diestel R 2010 Graph Theory Springer Verlag Heidelberg Fourth Edition Dijkstra E W 1959 A note on two problems in connexion with graphs Numerische Mathematik 1 269 271 Dunning J B Danielson B J and Pulliam H R 1992 Ecological processes that affect populations in complex landscapes OIKOS 65 169 175 Fall A et al 2007 Spatial graphs principles and applications for habitat connectivity Ecosystems 10 448 461 FAO 2006 Global Forest Resources Assessment 2005 Main report Progress towards sustainable forest management FAO Forestry Paper 147 Rome p 320 Hagberg AA et al 2008 Exploring network structure dynamics and function using NetworkX In Varoquaux G et al eds Proceedings of the qe Python in Science Conference SciPy2008 Pasadena CA USA pp 11 15 McGarigal K et al 2002 FRAGSTATS Spatial Pattern Analysis Program for Ca
23. sion but must be comma delimited In addition points must fall in a unique pixel of the resistance grid i e the point spacing should be greater than the resolution of your resistance grid times the square root of two If you have overlapping points the program will display the following error message There are overlapping points around x and y please check for points that are too close together This point will not be included in the run The example simulation run XY locations file is small test l10pts xy and will provide format example 3 3 Types of paths 3 3 1 Dijkstra s shortest path UNICOR can calculate all pairwise shortest path or least cost paths E normal the specifying no need to run Warning Use for the XY locations specified when th UNICOR user manual 10 is set to You number o recommend running this option in parall ndividual Edge Type this is a large graph program and if you have a large number of points and or large resistance grid then w See section 3 4 for more details Euclidean distance between all i F pixels in your el can also use UNICOR to calculate XY locations by ED Threshold to TRUE This is a fast calculation and in parallel 3 3 2 Kernel density estimation on shortest paths Dijkstra s base a
24. ssociated with the same starting node This provides a substantial boost in computational efficiency where all pairwise combinations are found for the same starting node before clearing the search space from memory All paths found are optimal paths of movement computed for every paired combination of starting and ending nodes The combination of these shortest paths create a path density map or connectivity graph In essence this approach becomes a large graph problem for the applied landscape connectivity assessments In analyses involving UNICOR user manual 5 large numbers of individ environment computationa However parallel proces ubiquitous modern multi uals across a large and fine grained processing time becomes intractable sing allows for efficient use of increasingly core processors Dijkstra s breadth first ext search algorithm is idea and destination points b independently We have i the multiprocessing modu processing in UNICOR is 1 for running in parallel for sets of source ecause pairwise distances can be calculated mplemented parallel processing in UNICOR using le from Python version 2 6 Parallel currently only implemented under th Linux operating system To reflect species speci Pes ic differences in dispersal abilities users can specify connectivity express
25. sues 5 1 How to obtain UNICOR The program is freeware and can be downloaded at http cel dbs umt edu software UNICOR with information for users including manual instructions FAQ publications ongoing research and developer involvement 5 2 Debugging and troubleshooting For help with installation problems please check first for postings at our web site Otherwise please report problems including any bugs to me at erin landguth mso umt edu 5 3 UNICOR limitations The following is a list of the current as we know of limitations with UNICOR 1 The resistance surface is in ASC format header file with 6 lines of information and space delimited 2 The point locations have a header row and file is comma delimited 3 Point locations must fall inside the resistance grid extent The code will run when points lie outside of grid but no paths will be calculated 4 The point locations must lie in a unique pixel or cell in the resistance surface 5 Graph metrics will only run on small problems and not implemented in parallel 5 4 How to cite UNICOR This program was developed by Erin Landguth Brian Hand and Joe Glassy GUI development was done by Mike Jacobi Ross Carlson assisted with graphics data set and website creation The reference to cite UNICOR user manual 14 is Landguth EL Hand BK Glassy JM Cushman SA Sawaya M 2011 UNICOR A species connectivity
26. t a command window If python is available you ll get the python prompt gt gt gt If it is nota recognized command it means either that python is installed but is not in your command shell s paths or that python is not installed In the first case ask an administrator to add it to your command paths or search for a tutorial on How to set environmental path variables If your shell locates and loads python type import numpy Similarly type import scipy If python does not complain that there are no such modules all is well The following instructions assume Python NumPy and SciPy are not yet available on your computer if they are skip to section 2 2 2 First run the Python executable installer you ve chosen either from www python org ActiveState or Enthought accepting defaults for the installation directory On Windows this will typically place the executables and libraries in c Python2 6 bin and the site packages package tree for user installed Python modules in c Python2 6 lib site packages If you are installing it on a network on which you do not have administrative privileges you may need to ask a system administrator to install python and the NumPy and SciPy UNICOR user manual 8 packages in their default locations Next install NumPy and SciPy using the supplied executable superpack installer or visiting http www
27. tegorical Maps Computer software program produced by the authors at the University of Massachusetts Amherst Available at the following web site http www umass edu landeco research fragstats fragstats html McRae B H and Beier P 2007 Circuit theory predicts gene flow in plant and animal populations Proceedings of the National Academy of Science USA 104 19885 19890 McRae B H et al 2008 A multi model framework for simulating wildlife population response to land use and climate change Ecological Modelling 2197 TWO Lo Li Q and Racine J S 2007 Nonparametric Econometrics Theory and Practice Princeton University Press Opdam P and Wascher D 2003 Climate change meets habitat fragmentation linking landscape and biogeographical scale levels in research and conservation Biological Conservation 117 285 297 Pinto N and Keitt T H 2009 Beyond the least cost path evaluating UNICOR user manual 16 corridor robustness using a graph theoretic approach Landscape Ecology 24 253 266 Riitters K et al 2000 Global scale patterns of forest fragmentation Conservation Ecology 4 online URL http www consecol org vol4 iss2 art3 Sawyer H et al 2009 Identifying and prioritizing ungulate migration routes for landscape level conservation Ecological Applictions 19 2016 2025 Scott D W 1992 Chapter 6 In Multivariate Density Estimation Theory Practice and Visualization John Wil
28. uting specify the Processors that are to be used in a simulation which files to output Table 1 UNICOR inputs all Number of Files include ascii connectivity path options cost distance graph theory metrics and outputs with description and dependencies Input Name Default Description Dependency Example Input Grid Filename small_te The filename for the resistance surface st rsg in ascii format with header file any file extension is acceptable must be space delimited XY Filename small t The filename for the individuals with st_10pts x y locations any file extension is XY acceptable must be comma delimited Use ED Threshold False Option for calculating Euclidean distance between all pairwise points Use thresholding in next field as option ED Distance 50000 If Use ED threshold is True then this is Euclidean distance in map units to apply to the x y point locations If you want all Euclidean distance values specify this value to be greater than the maximum distance on your map Edge Distance 100000 The resistance distance threshold in terms of edge distance or cost units to UNICOR user manual 12 the path lengths between each XY locations Use this value Edge Type threshold and all paths apply to pair of with Edge Type Normal normal All pairwise shortest cost distan
Download Pdf Manuals
Related Search