| About GBIF MAPA |
|
|
Help
Frequently Asked Questions about MAPA Why do we need to optimize the design of surveys? What is the Survey Gap Analysis (SGA) tool and what can you do with it? What is the Species Richness Assessment (SRA) tool and what can you do with it? What is the Environmental Values Extraction (EVE) tool and what can you do with it? What do I need to understand when choosing an ROI for SGA? How big an area (ROI) should I use for SGA? How big an area (ROI) should I use for SRA ? What should I consider when selecting taxa for SGA?
The overall process The MAPA application is built around a five step process. To design a survey, carry out a species richness assessment or extract environmental values from raster layers requires following this five step process. If you just want a list of species from an area or to create a map then you may only need a subset of the available five steps. The first step to using MAPA is to select, on the front page, the activity you wish to carry out. Once you have selected the activity you will be directed to the first page of the tool that will assist you with that activity. Each page of each tool has an inbuilt process flow that leads you through the steps required to achieve the result you are after. This process flow is visible at the top of each page after you select the activity you wish to carry out from the front page. The five steps:- Select an ROI - using one of the three options select the area for which you wish to carry out your activity. Option 1 - choose a country from the drop down list. This then establishes the bounding box for that country as the ROI. Many larger countries such as the USA have external territories which make the country ROI much larger than just the mainland. If you wish to only use the mainland as the ROI you may wish to use Option 2 to define your ROI. Option 2 - draw a box on the map to establish the area covered by your box as the ROI. You can pan and zoom before drawing your box to better define your ROI. Option 3 - Enter your ROI manually using the lower left and upper right latitude and longitudes. Choosing Taxa - enter part or all of a taxon name (eg not both genus and species but either genus or species) into the search box at the top left of the page and click Choose. The Catalogue of Life is then searched for any hits.The results will be returned in the box below the search box. Next to each of the returned names is the word "add". By clicking on "add" you can add the taxon name to the "My Search Criteria" box. You can add as many taxa as you like to this box from as many searches of the CoL as you wish. To search the CoL again using a different string simply enter the string (name) into the search box and repeat the process to add taxa to your search criteria. Searching the GBIF cache - once you have finished selecting your taxa of interest you may wish to choose other criteria for filtering the search by. Do this by clicking on the Advanced Search button. Once you have finished selecting your filter criteria click the Search button to activate the search of the GBIF cache. Mapping the search results - the search results are listed in a table with the count of the number specimens of each taxa that have coordinates found, along with a drop box for selecting a symbol for each taxa. To assign a symbol to each taxon you can select a single symbol for all taxa in the auto assign drop down box at the top of the page or assign a symbol to each taxon independently using the drop down boxes on the same row as the taxon name. If you have many taxa where you would like to assign one symbol to the majority and a few different symbols to a small minority simply use the auto assign to give all taxa the same symbol then change those few that you wish to have separate symbols using the adjacent drop down boxes. Once you have assigned your symbols click on the Map button to proceed to the map. The use of manual selection of symbols is to enable you to visualize your data more effectively than would be the case if you were limited to one symbol or to automatic assignment of symbols. Analysing the data - when you have a map of your data on the screen you can visually validate the data to assess whether it is what you want to use for your analysis, if it isn't then you can use the process buttons at the top of the screen or the back button to change your ROI and/or your taxa of interest and rerun the database search. You can also move or delete data points on the map before submitting the points for analysis. This point editing enables you to refine your dataset by removing spurious data or outliers or any points you know are incorrect. Once you are happy with the data then you can run the analysis with the default settings or customize the settings before running the analysis.
Frequently Asked Questions about MAPA MAPA stands for Mapping and Analysis Portal Application. It has been built using funds made available t hrough the Global Biodiversity Information Facility's Demonstration Project process. MAPA consists of :-
MAPA has been developed through a collaboration between the Australian Museum (AM), University of Colorado (UC), and the New South Wales Department of Environment and Conservation (DEC). Principal Investigators for the project are Paul Flemons (AM) and Rob Guralnick (UC). Software developers and system architects are Ajay Ranipeta (AM) and David Neufeld (UC). Statistical analysis developer is Jon Kreiger (UC), Geographic Information Systems Analyst Gareth Carter, Survey Gap Analysis architects Simon Ferrier (DEC) and Dan Faith (AM) and the Survey Gap Analysis developer was Glenn Manion (DEC). MAPA has been developed using Open Source technology and modular service oriented architecture. The technology suite is Java, JSP, MySQL, PostgreSQL, MapServer, Mathematica. The existing software for Survey Gap Analysis and Mathematica was optimized and wrapped as a web service. Why do we need to optimize the design of surveys? Biodiversity surveys are expensive undertakings requiring careful planning, and specialized resources in terms of personnel skills and equipment. It is essential then that the data obtained through new surveys complements existing data and maximizes the usefulness of the new data for conservation planning purposes. Many, if not all, museum collections are characterized by biased sampling resulting from either ad-hoc collection techniques or from planning that is based on ease of access and which only considers geographic, rather than environmental, coverage when locating survey sites. Survey Gap Analysis can be instrumental in reducing bias and thereby more effectively answering the question "If one is interested in obtaining an overall knowledge of the biodiversity (or of a taxon) of an area, and if there are insufficient data, then where should survey data be gathered?" Funk et al 2005. What is the Survey Gap Analysis (SGA) tool and what can you do with it? The Survey Gap Analysis application enables a user to utilise locations of existing specimen records (from GBIF standard data sources eg DiGIR providers) and mapped environmental variables to create a mapped complementarity surface indicating the relative priority for additional survey or collection effort throughout the region of interest (ROI). The priority being based on the potential for an area (based on climatic conditions) to complement existing survey effort in the region of interest. Using a range of contextual GIS data layers (eg roads) the user can select a range of new survey sites that will optimise their survey's representation of the environmental and geographical variation inherent in their ROI. The Survey Gap Analysis technique used in this application has been derived from that previously prototyped by the NSW Department of Environment and Conservation (NSW NPWS 1998, Ferrier 2002, Graham et al 2004, Funk et al 2005). This excerpt from Funk et al 2005 briefly describes the Survey Gap Analysis tool: "A new tool developed by the GIS Unit at the New South Wales National Parks and Wildlife Service (now Department of Environment and Conservation) in Armidale (NSW NPWS, 1998; Ferrier, 2002)...by extending an analysis technique pioneered by Faith & Walker (1996). ...it analyses the survey coverage of a region in relation to the underlying continuous environmental and geographical space, rather than in terms of arbitrary classes. Faith & Walker's (1996) environmental diversity (ED) measure, based on the p-median criterion, was developed for selecting sets of sites that represent regional biodiversity by providing best possible coverage of regional environmental variation. It functions by measuring how well a set of sites covers the continuous environmental space and evaluating the potential improvement that any given site would make if added. The technique, based on the finding that sampling different parts of the overall environmental space yields a good representation of the biological diversity of a region (Faith & Walker, 1996), can equally be applied to the problem of selecting survey sites." Background information and examples for the ED approach underlying SGA can be found in Faith et al (2004). The basic calculations can be described as follows. Given a partial set of survey sites, SGA identifies a new site that would be expected to contribute the greatest number of additional species. Note that this is not the same as finding a site that would have the greatest total number (richness) of species. Instead, it is the site with the greatest complementarity to the existing survey sites. Complementarity is widely used in biological conservation, and conventionally refers to some count of the number of additional species provided by a new site. In the SGA context, we cannot explicitly count such gains. Instead, these complementarity values are estimated using ED (providing ED-complementarity values; Faith et al 2004; Funk et al 2005). In GBIF MAPA the SGA tool produces a map showing the distribution of current ED complementarity values for all possible new sites. Darker colours indicate larger values; the red flag is the suggested choice for a new site, as it is the one sitting at the "highest peak". Using the complementarity surface as a guide the user can move the suggested site to take into account access constraints (using roads and rivers GIS layers). Once accepted as a new site the red flag turns green and cannot be moved again. Once a site is accepted and the analysis run again to select another site, a new complementarity surface is created showing a different pattern of ED complementarity values based on the use of the original survey sites you started the analysis with and the new sites chosen by SGA. This highlights the dynamic nature of the survey gap analysis - the complementarity value of a site always depends on the set of sites already selected. What is the Species Richness Assessment (SRA) tool and what can you do with it? Use this tool to provide an estimate, from GBIF data, of the number of species in an area; and to gain insight into the adequacy of sampling based on abundance distributions for those species. The SRA tool replicates many of the features of EstimateS (http://viceroy.eeb.uconn.edu/estimates), calculating a variety of species richness estimators. This tool builds upon Gareth Russell's Estimating Species Richness eco-tool (http://eco-tools.njit.edu/ webMathematica/EcoTools/index.html). The analyses include: 1. Various summary statistics, 2. Estimators of true species richness, including Chao 1, Chao 2, ACE, ICE, first-order jackknife, second-order jackknife, and bootstrap (see Colwell and Coddington 1994 for an overview of these estimators). 3. Variances (and hence confidence intervals) based on analytical results (for Chao 1 and Chao 2) 4. Variances based on multiple random resamples with replacement (for everything else). 5. A sample-based rarefaction curve, with confidence intervals, using the analytical method described in Colwell et al. 2004, and 6. The individual-based rarefaction curve for comparison. These curves are plotted, and both the figures and the curve data are provided for download. Items 2-6 are presented in two csv-format tables, calculated t hrough resampling with and without replacement. For more detailed information on SRA see here What is the Environmental Values Extraction (EVE) tool and what can you do with it? EVE allows users to extract values for point locations from underlying environmental layers such as rainfall, temperature, and elevation. The result is a matrix with the rows representing each specimen and the columns representing the values of the environmental data layers requested for where that specimen was collected. This tool will work at any spatial scale. Biodiversity change is partly due to changes in abiotic context where species occur. EVE provides a way to determine this abiotic context for multiple species occurrences. This information is the raw data usable for modeling approaches (eg. species distribution modeling like GARP pr Maxent) and is also useful in more heuristic or statistical examinations of limits to species distributions. What do I need to understand when choosing an ROI for SGA? When choosing an ROI it is important to understand the relationship between your ROI and your taxa of interest. SGA does not identify the areas where you are most likely to find your taxa of interest, it identifies those areas that are most complementary to your existing survey sites (represented by your taxa's point locations on the map) in terms of environmental conditions (in the case of MAPA this is in terms of the 19 climatic variables). It is important then to consider the likely or known geographic and environmental range of your taxa of interest when choosing your ROI. Example - if you are looking to design a survey for taxa that is well known to be limited to low lying areas it would be unwise to include in your ROI a mountain range where your taxa of interest would simply not be found. How big an area (ROI) should I use for SGA? Theoretically SGA could be applied to any sized land area. However realistically it is best suited to regional and state scale analyses. It may also be applicable to nation scale for those nations covering relatively small geographic areas. In Europe for example it would still be suitable for use at national scale but in countries as large as Australia, the USA and Canada it would seem to be less advisable to do so for the reasons explained above. How big an area (ROI) should I use for SRA ? Regardless of the size of the ROI used for SRA, the number of grid cells will always be limited to a user selected choice of 100, 650 or 900. This means that the bigger the area you choose the bigger the analysis cell size. What should I consider when selecting taxa for SGA? SGA works best where each point represents a systematic survey site for a taxa or group of taxa meaning that for each location, the prescence or absence of a range of taxa is recorded. The nature of the majority of GBIF data is that is ad hoc collection based. This means that each data point represents usually only a single species or small number of species. References Colwell, R. K. and J. A. Coddington. 1994 Estimating terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society (series B) 345: 101-118. Faith, D. P. & Walker, P. A. (1996) Environmental diversity: on the best-possible use of surrogate data for assessing the relative biodiversity of sets of areas. Biodiversity and Conservation5, 399-415. Faith, D. P., Ferrier, S. & Walker, P. A. (2004) The ED strategy: how species-level surrogates indicate general biodiversity patterns through an "environmental diversity" perspective. Journal of Biogeography31, 1207-1217. http://www.amonline.net.au/systematics/pdf/jbi_faith_2004.pdf Ferrier, S. 2002. Mapping spatial pattern in biodiversity for regional conservation planning: where to from here? Systematic Biology 51: 331-363. Funk, V.A, Richardson, K.S., and Ferrier, S 2005. Survey-gap analysis in expeditionary research: where do we go from here? Biological Journal of the Linnaen Society, 2005, 85, 549-567. Graham, C.H., Ferrier, S., Huettman, F., Moritz, C. and A.T. Peterson 2004. New developments in museum-based informatics and applications in biodiversity analysis. Trends in Ecology and Evolution 19: 497-503. NSW NPWS. 1998. Vertebrate Fauna Survey. NSW Comprehensive Regional Assessment Project Report. Sydney: NSW NPWS.
|