Nature and Human Society: The Quest for a Sustainable World (1997)
Chapter: The National Biodiversity Information System of Mexico
Page 586
The National Biodiversity Information System of Mexico
Introduction
Despite its novelty and sheer scope, the concept of biodiversity has already been used as a basis of multilateral treaties, global funds, national strategies, and many other political and scientific initiatives. Foremost among the actions that countries are expected to undertake to preserve biodiversity is the creation of inventories (Reid and others 1992; SA2000 1994) and information systems (Olivieri and others 1995; United Nations Environmental Programme, UNEP; WCMC 1996; http://www.unchs.unon.org) to organize the huge body of biodiversity data already in existence. Without powerful informational tools, the task of protecting, managing, and using biodiversity at national levels is impossible. Many peasant and indigenous cultures manage their biological resources successfully with nonmodern information tools (Berlin and others 1973; Haverkort and Millar 1994), but the increase in the temporal, spatial, and taxonomic scales implied by the full concept of biodiversity leads us to the use of scientific, modern tools for the knowledge and use of biodiversity.
In March 1992, the Mexican government created a national commission, (Comision Nacional para el Conocimiento y Uso de la Biodiversidad [CONABIO] 1992, http://www.conabio.gob.mx), with the task of coordinating the national biodiversity inventory and the associated databases and information systems. In this presentation, we outline its conceptual framework and current status, focusing on the role of the users and providers of the data. Mexico's National Biodiversity Information System is not yet finished in a strict sense, but many of its
Page 587
components are already in operation and providing many services, which we describe here.
A Biodiversity Information System
An information system can be defined as a structured set of processes, personnel, hardware, and software to turn data into usable information (WCMC 1997). That definition forces us to focus on a number of issues. First, data are provided by channels usually under the control of people. In the biodiversity field, the people are the geographers, taxonomists, ecologists, geneticists, foresters, traditional physicians, wholesalers in natural products, and so on, that generate raw or aggregated data about any of the levels of biodiversity. A biodiversity information system (BIS) must have clearly defined and operative relations with the providers of the data.
Second, what “information” means is determined by a set of potential or actual users. For example, the list of the Latin binomials of medicinal plants in a given municipality (or region) for which a national market exists might be useless to inhabitants who lack scientific training, whereas the same data presented in the form of a guide with common regional names and illustrations might be highly informative to them. Similarly, to a national-level decision-maker, a map of endemic species richness might be much more useful than an equation fitted to the raw data, which might be packed with information for a macroecologist; and the raw data themselves could be useful to a taxonomist. The output of a BIS should be defined in close contact with the main users of the system.
Third, the technical specifications of a BIS are likely to be exacting and difficult. Because of its multiscale features, data will appear in a number of formats, from those of geographical information systems (GISs) to text files, images, taxonomic datasets, genomic information files, and so on. Also, some categories of data are large. For example, the Digital Elevation Model of Mexico at 1:250,000 is almost 500 megabytes (Mb). Some curatorial databases are also of significant size. Two examples of CONABIO illustrate this: a database with 403,507 records of specimens of vascular plants with 26 fields from 85 projects is 111 Mb and the bird database with 250,283 records of specimens and 164 fields is 88 Mb. The structure of a BIS should respond to the complexity and magnitude of the problem at hand.
Many BISs are already in operation (WCMC 1996). Among the better known are the Nature Conservancy Heritage Program (Jenkins 1988; http://www.tnc.org/), the World Conservation Monitoring Centre (WCMC, http://www.wcmc.org.uk/), the Australian government's Environmental Resources Information Network (ERIN, http://kaos.erin.gov.au/erin.html), and the Costa Rican Institute Nacional de Biodiversidad (InBio) system (http://www.inbio.ac.cr/). One of the main criteria for characterizing a BIS is whether it is based on raw, “atomic data” or on interpreted information. For example, the set of locations where a species has been recorded is less interpreted than a researcher's rendering of the area of distribution of the species. BISs that are explicitly based on atomic data are the
Page 588
Australian ERIN (Chapman and Busby 1994) and Costa Rica's InBio. The system being developed in CONABIO belongs to this class.
The National Biodiversity Information System of Mexico
The main task given to CONABIO by the presidential act that created it was to coordinate the inventory of Mexican biodiversity and to develop and maintain the information system for it. CONABIO started by holding workshops with potential users and providers of the information. Such consultations have been maintained, formally or informally, to the present. We also reviewed a number of existing BISs.
Most of the needs of users were related to variations on the symmetrical themes of “What entities are present here?” and “What exists and where?” In those two questions, “entities” and “what” can mean a species, a higher-level taxon, or an attribute of them, such as “mammals,” “federally listed butterflies,” “medicinal plants,” or “migratory birds.” “Where” and “here” refer to arbitrary polygons and regions naturally or politically defined, such as a given ecoregion, a state, a municipality, or a protected area. Those questions call for simple “distributional” information. For example,
• What endangered species exist in a state, municipality, protected area, or ecoregion?
• What are the holdings of particular taxa in particular museums or collections (generally speaking, curatorial data for taxonomic groups present in Mexico)?
• What tree species in a given ecoregion are present in one of the major army nurseries?
• What butterfly and bird species are present in a municipal natural park?
• In what region(s) can a given species produced in the army nurseries be planted?
• In what region(s) is there a high likelihood of the presence of some endangered (or otherwise defined) species?
• What municipalities cover a given ecoregion?
A second class of information is not distributional but is associated with particular taxa. Examples of this more complex, “verbal” information are
• information on production technology for particular species (mainly useful tree species);
• chemical or clinical data on medicinal plants;
• indigenous knowledge about species;
• toxicological and first-help data on poisonous species;
• markets (buyers, certifiers, and exporters) for useful species, such as nontimber forest products;
• demographic or genetic data on particular species or groups, such as whales, cacti, or other “charismatic” taxa;
• traffic data on species regulated under the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES);
Page 589
• images, pictures, illustrations, recordings, and multimedia data related to conspicuous species; and
• general information about protected areas, including images and tourist data.
A third class of information has a temporal component. It is related to trends in the sizes of regions or populations. Obvious examples are
• rates of change of particular types of vegetation into other types (such as in deforestation) and
• changes in the sizes of populations of particular species.
The provenances, updating regimes, quality, and structure of the data required to fulfill the needs described above are highly heterogeneous. A BIS capable of responding to such a set of demands probably does not exist, although many existing systems are quite capable of answering questions on some levels in the biodiversity scale. Some systems are based entirely on bibliographic information, such as the Indian Indira Gandhi Conservation Monitoring Centre (http://www.wcmc.org.uk/igcmc/) and Napralert at the University of Illinois (Farnsworth 1988); metadata systems, such as the National Biodiversity Information Infrastructure of the Department of the Interior in the United States (http://www.nbs.gov/); mixed systems, such as The Nature Conservancy in the United States and several Latin American countries (Jenkins 1988; http://www.tnc.org/); systems of state scope, such as the Gap Analysis Program (GAP, http://www.gap.uidaho.edu/gap/); systems oriented wholly to taxonomic information, such as the Expert Center for Taxonomic Identification (ETI, http://turboguide.com/data2/cdprod1/doc/cdrom.frame/002/607.pub.Expert.Centre.for.Taxonomic.Identification.ETI.html) and the PLANTS National Database (http://plants.usda.gov/plants/); and at least one system with a world scope, WCMC (http://www.wcmc.org.uk/). A scheme of a hypothetical, ideal BIS capable of answering all types of biodiversity questions is depicted in figure 1.
The hypothetical scheme that appears in figure 1 has a realistic interpretation. The core element is the data associated with the specimenthe atomic data referred to before. Particularly important from the perspective of a BIS are the georeference and taxoreference associated with a specimen (Colwell 1996). The georeference is the data that specify the locality where the specimen was collected. It is subject to a variety of errors and imprecision. Without resorting to modern geographic positioning system (GPS) technology, the georeference might be obtained to a precision of a few kilometers. GPS increases the resolution to a few hundred meters or better. The georeference expressed in coordinates provides the most flexible link to information that is spatially structured.
The taxoreference is the expression of a hypothesis on the current position of a specimen within the system of biological taxonomy (Bisby 1995). Besides being hypothetical, the taxoreference is also subject to errors and imprecision. Therefore, this taxonomic information must be updated and modified periodically by specialists. The scientific names are the essential language to communicate about biodiversity (May 1995; Patrick 1996; Thompson 1996), and the taxoreference is
Page 590
Figure 1
Scheme of ideal biodiversity information system. Information should be available on
different scales, temporal and spatial, and from different perspectives. Georeferenced and
taxoreferenced specimen databases provide links required to move among scales and
points of view.
the link to the world of bibliographic data, legislation, markets, population information, and so on.
Taken together, the georeference and the taxoreference provide the links between sets of data that have a geographical structure and sets that are normally associated with a Latin binomial. From the perspective of a national BIS, the information on the millions of specimens in herbariums, museums, and scientific collections constitutes the backbone that allows movement along the many levels in the structure of biodiversity (Soberón and others 1996). That is why the atomic data on a specimen are so powerful and why CONABIO has since its creation been working on assembling the specimen-data backbone.
Specimen Information on Mexican Species.
The biological specimens that have been collected in Mexico are deposited in 190 Mexican institutional collections and in other countries. According to an inventory of the taxonomic activities that CONABIO organized in 1996, there are about 10 million holdings in Mexican collections, in different stages of curation and data capture; the sizes and geographical distributions of the collections in Mexico are uneven. Depending on the taxonomic group, the proportions of specimens held in Mexican and foreign institutions vary. Some collections of Mexican specimens in foreign institutions are very important. For example, the Birds Database (Peterson and others 1997), which has information on about 300,000 specimens, is 80% foreign in origin, with perhaps a further 10% of Mexican
Page 591
collections still to be included. Foremost among the countries with holdings of Mexican specimens are the United States, Canada, and the United Kingdom; several European countries also have important collections in some taxa.
CONABIO's botanical databases, in contrast, are being compiled mainly in two Mexican collections, principally in the National Herbarium of the Instituto de Biologia, UNAM (MEXU, IBUNAM), and the Herbarium of the Escuela Nacional de Ciencias Biológicas of the National Polytechnic Institute (ENCB, IPN). Those two collections contain 1.5–2.0 million specimens.
The task of computerizing and georeferencing the data in the collections is significant. Since its creation, CONABIO has obtained about 230 databases, which contain data on more than 4,250,000 specimens of plants and animals.
The cost of obtaining specimen information is roughly constant per specimen, so computerizing large holdings is much more cost-efficient than small collections.
The information on vertebrates is extensive and probably covers most of the world collections of specimens that have been collected in Mexico. Despite this, large gaps and aggregation of collection sites along roads are still present (figure 2). This is important in countries with a large spatial turn component of diversity that requires extensive sampling over most of the country (Sarukhán and others 1996).
Figure 2
Collecting localities of terrestrial vertebrates. Data came from 43 databases with 514,178
records (26 projects on mammals, 142,923 records; eight projects on amphibian and reptiles,
80,741 records; and nine projects on birds, 290,514 records). Gaps and roads become
apparent by aggregation of dots.
Page 592
Major botanical collections have been more difficult to computerize, and it was only in 1996 that the ENCB herbarium began the task. IBUNAM has partial computerization of some taxonomic groups or regions. Medium-size herbariums like the Instituto de Ecologia of Xalapa (XAL, around 250,000 specimens) and the small herbariums of the Asociacion Mexicana de Orquideología (AMO, around 110,000 specimens) and Centro de Investigaciones Cientificas de Yucatán (CICY, almost 45,000 specimens) are almost totally computerized. XAL and AMO were pioneers in computerizing collections in Mexico. Other collections have been partially computerized, but the addition of data from computerized collections to CONABIO's databases is slowing down because a large proportion of Mexican specimens are still not curated or determined and because a few significant collections both in Mexico and abroad have not started comprehensive computerization of their specimens. When this task is finished, new increases in specimen data will have to come mostly from new explorations.
Quality Control
Information for the specimen databases is obtained from projects undertaken by universities or research groups and is externally peer-reviewed. But an independent process of quality control is required because data can be subject to faulty determination, unstable taxonomy (McNeill 1993), equivocal georeferencing (Chapman and Busby 1994), and other problems (WCMC 1996). It is possible to spot a large number of those by “inconsistency analysis” (Murguía 1996), whereby records are checked for intrarecord consistency, proper spelling and synonymy, taxonomic nestedness, or geographic consistency (Margules and Austin 1995; Margules and Redhead 1995; Stockewell, in Hart 1997).
The coordinator role of CONABIO is important although all the information that has been integrated came from specialists. CONABIO has been responsible for maintaining and updating the data, obtaining authority files, developing inconsistency-spotting routines, and organizing a network for sharing updated data.
The databases that have gone through the full process of data-quality control are included in the large “container” called BIOTICA (http://www.conabio.gob.mx), which has links to the GIS and to bibliographic information. Soon it will be linked to information on markets and legislation.
The data model for BIOTICA was developed in CONABIO with the assistance of a number of users (mainly taxonomists). The increasing number of providers using the same model, already more than 105, is greatly reducing the number of inconsistencies. Some data providers still use their own systems or commercial data managers; their information can be included in BIOTICA, but the number of inconsistencies tends to be larger than when BIOTICA is used.
Uses of the Mexican National Biodiversity Information System
Although the Mexican National Biodiversity Information System is still unfinished as an integrated system, many of its components are fully operational, and it is already providing services. Most services answer requests for information about the distributional questions noted earlier. Every month, around 30 requests are made by telephone, e-mail, or fax or personally. The providers of the data
Page 593
are informed about who used their information and what for, and the users can ask the providers for more details. (Some information is protected for various reasons.) A single request might require information from up to 20 databases, each the product of years of expert work and sometimes the result of centuries of accumulated institutional efforts.
CONABIO's home page (http://www.conabio.gob.mx) receives an average of 2,000 hits per day. The page is linked to some databases already released, many GIS thematic maps of Mexico, and the results of a number of analyses like a biodiversity priority-setting of Mexico and a guide to species illegally traded through Mexico.
The Future of the Mexican National Biodiversity Information System
User demands require that CONABIO's BIS include information about protected species (including trends in populations) and useful or marketable species and provide a higher level of resolution of cartography, including time series for vegetation cover in some areas. Therefore, the next steps should probably focus on
• higher resolution for priority areas (the priority-regions workshop [http://www.conabio.gob.mx/textos/prior.htm] yielded 155 regions covering about 20% of Mexico as those still promising for conservation efforts; the workshop used 1:4,000,000 cartography and pinpointed areas on which there was a serious lack of information);
• monitoring (this may be done on different scales, the simplest using satellite images to monitor changes in vegetation);
• updating of databases and catalogs;
• completing the computerization of the national collections and promotion of the extensive use of the system as a powerful tool for scientific purposes, management, and communication;
• repatriation of information and strengthening relations with foreign museums to ensure collaboration;
• data on useful species (trees, medicinals, and ornamental, food and nontimber forest products), which are especially important for peasant communities, small ranchers and farmers, and national and international biotechnology industries (CONABIO has started a project to create an information system for 600 such species. It will be based on data already obtained for the reforestation program, but it will also include data on uses, production techniques, ecological requirements, images, and so on); and
• a similar effort for 300 protected species to add to the current database on the CITES species.
Perspectives on a Regional Biodiversity Information System
An example of cooperation and information-sharing among neighbor countries was recently initiated with the support of the Commission for Environmental Cooperation of the NAFTA countries. Following the lead of previous efforts developed by Julian Humphrys, formerly of Cornell University, a pilot study was finished in 1997 that was based on data from the Mexican Bird Atlas (300,000
Page 594
records), the Breeding Bird Survey (160,000 and 100,000 more in collections), and the US Breeding Bird Survey (15,000 records). This pilot example, the North American Biodiversity Information Network (NABIN), demonstrates the feasibility of accessing data distributed in several independent institutions. It is an example of the benefits of sharing information (Peterson and others 1997). One of the goals achieved was the development of a common catalog based on previous efforts (American Ornithological Union, http://www.itis.usda.gov/itis/). Another interesting development was the capacity to do bioclimatic modeling by sending the results of queries to the machine at the San Diego Super Computing Center. NABIN demonstrated the feasibility of creating a large-scale, multicountry distributed BIS. It will open the doors to larger efforts, such as the Inter-American Biodiversity Information Network being discussed by several countries.
Conclusion
We have described an information system based on specimen data and the uses that nonbiologists might have for the information. In Mexico, assembling the data required the participation of hundreds of Mexican taxonomists, ecologists, agronomists, and geneticists. The Mexican Government, through CONABIO, had to support not only the creation of databases, but also a large part of the basic activities of the researchers, such as purchase of cabinets and equipment, visits to foreign institutions, and field trips. Maintaining the information system will require continued support for the creators and maintainers of the information. The cost, although great for the country, has remained moderate relative to the large increase in the value of the information in the collections, which is now available and being used by an unprecedented number of people.
The specimen data are the core of a multiscale BIS. Despite the enormous holdings of systematic institutions all over the world, large gaps in our knowledge about the biota of the planet remain. Therefore, we will need more funds and concentrated efforts to computerize existing collections and increase the pace of exploration (SA2000 1994). The example of CONABIO shows that the task is feasible and should be tackled on a global basis.
Acknowledgments
We are grateful to the many people at CONABIO who worked to make this presentation possible. We thank especially Raul Jimenez, CONABIO's director of systems, for the GAP analysis; Hesiquio Benitez, subdirector of external services; Carlos Alvarez and all the personnel in the Biotic Inventories Area who worked on the GIS. We are also very grateful to the providers of the data we have presented here. Their collective effort is what makes biodiversity information systems possible.
References
Berlin BD, Breedlove E, Raven P. 1973. General principles of classification and nomenclature in folk biology. Am Anthro 75:214–42.
Page 595
Bibby CJ, Collar NJ, Crosby MJ, Heath MF, Imboden C, Johnson TH, Long AJ, Sttatersfield AJ, Thirgood SJ. 1992. Putting biodiversity on the map: priority areas for global conservation. Cambridge UK: International Coun for Bird Preservation.
Bisby FA (coord). 1995. Characterization of biodiversity. In: Heywood VH (ed). Global biodiversity assessment. Cambridge UK: Cambridge Univ Pr. p 21–106.
Butterfield BR, Csuti B, Scott JM. 1994. Modeling vertebrate distributions for gap analysis. In: Miller RI (ed). Mapping the diversity of nature. Oxford UK: Chapman & Hall. p 53–68.
Chapman AD, Busby JR. 1994. Linking plant species information to continental biodiversity inventory, climate modeling and environmental monitoring. In: Miller R (ed). Mapping the diversity of nature. London UK: Chapman & Hall. p 179–94.
Colwell RK. 1996. Biota. The biodiversity database manager. Sunderland MA: Sinauer.
Farnsworth NH. 1988. Screening plants for new medicines. In: Wilson EO, Peter FM (eds). Biodiversity. Washington DC: National Acad Pr. p 83–97.
Hart D. 1997. New communities will benefit from HPC technology. Gather/Scatter 13(3):14–5.
Haverkort B, Millar D. 1994. Constructing biodiversity: the active role of rural people in maintaining and enhancing biodiversity. Etnoecologica 2(3):51–64.
Jenkins Jr RE. 1988. Information management for the conservation of biodiversity. In: Wilson EO, Peter FM (eds). Biodiversity. Washington DC: National Acad Pr. p 231–8.
Margules CR, Austin MP. 1995. Biological models for monitoring species decline: the construction and use of databases. In Lawton J, May RM (eds). Extinction rates. Oxford UK: Oxford Univ Pr. p 183–96.
Margules CR, Redhead TD. 1995. BioRap. Guidelines for using the BioRap methodology and tools. Dickson Australia: CSIRO.
May RM. 1995. Conceptual aspects of the quantification of the extent of biological diversity. In: Hawksworth DL (ed). Biodiversity measurement and estimation. London UK: Chapman & Hall. p 13–20.
McNeill. 1993. Instability in biological nomenclature: problems and solutions. In: Bisby FA, Russell GF, Pankhurst RJ (eds). Designs for a global plant species information system. Oxford UK: Clarendon Pr. p 94–108.
Murguía M. 1996. Jerarquización de las metodologías de validación de datos de georreferencia. CONABIO. Mexico.
Peterson T, Navarro A, Warner R, Pisanty Y, Kennedy J. 1997. Pilot project North American bird information network. North American Biodiversity Information Network. CCA.
Olivieri ST, Harrison J, Busby JR (Coord.) 1995. Data and information management and communication. In: Heywood VH (ed). Global biodiversity assessment. Cambridge UK: Cambridge Univ Pr. p 21–106.
Patrick R. 1996. Systematic: a keystone to understanding biodiversity. In: Reaka-Kudla ML, Wilson DE, Wilson EO (eds). Biodiversity II: understanding and protecting our biological resources. Washington DC: Joseph Henry Pr. p 199–211.
Reid W, Barber C, Miller K. 1992 Global biodiversity strategy. Washington DC: World Resources Inst, The World Conservation Union, UNEP.
SA2000. 1994. Systematics agenda 2000: charting the biosphere. Technical report, produced by systematics agenda 2000. A consortium of the American Society of Plant Taxonomists, the Society of Systematic Biologists, and the Willi Hennig Society, in cooperation with the Association of Systematics Collections.
Sarukhán J, Soberón J, Larson J. 1996. Biological conservation in a high beta-diversity country. In: di Castri F, Younès T (eds). Biodiversity, science, and development: towards a new partnership. Cambridge UK: CAB International.
Soberón J, Llorente J, Benítez H. 1996. An international view of national biological surveys. Ann Missouri Bot Gard 83: 562–73.
Thompson FC. 1996. Names: the keys to biodiversity. In: Reaka-Kudla ML, Wilson DE, Wilson EO (eds). Biodiversity II: understanding and protecting our biological resources. Washington DC: Joseph Henry Pr. p 199–211.
Wilson EO. 1996. Introduction. In: Reaka-Kudla ML, Wilson DE, Wilson EO (eds). Biodiversity II: understanding and protecting our biological resources. Washington DC: Joseph Henry Pr. p 1–3.
WCMC [World Conservation Monitoring Centre]. 1996. Guide to information management in the context of the convention on biological diversity. Nairobi Kenya: UNEP.
WCMC [World Conservation Monitoring Centre]. 1997. Darwin initiative handbooks. Cambridge UK: WCMC.
