• DOI:10.1111/J.1523-1739.2007.00784.X
• Corpus ID: 32169747

Categorization systems of threatened species.

@article{Sobern2007CategorizationSO,
  title={Categorization systems of threatened species.},
  author={Jorge Sober{\'o}n and Rodrigo A. Medell{\'i}n},
  journal={Conservation biology : the journal of the Society for Conservation Biology},
  year={2007},
  volume={21 5},
  pages={
          1366-7; discussion 1368-70
        }
}

• Jorge Soberón, R. Medellín
• Published 1 October 2007
• Computer Science
• Conservation biology : the journal of the Society for Conservation Biology

de Grammont and Cuar´on (2006) provide interestinginformation about the way several countries of the West-ern Hemisphere address the issue of developing systemsfor threatened species categorization. Nevertheless, wedisagree with these authors on some fundamental pointsand think their conclusions that “the country categoriza-tions [they] evaluated have serious defects” and that “theIUCN (2001) system is the most adequate” are not sup-ported by their arguments. All our examples will comefrom the Mexican (SEMARNAT 2002) system, with whichwe are familiar.First,wedisagreewiththeauthors’implicitassumptionthat all the systems they analyzed have the same purpose.Threatened species lists can have several different objec-tives (Possingham et al. 2002) and therefore may need tobe evaluated differently. An example of this assumption isthe authors’ use of a single vector of weights (a weight of1 for all but characteristics 11 and 15, which received avalue of zero) for all the systems they analyzed. They didthisdespitethefactthatthe“othersystemcharacteristics”in their Table 2 are mainly of a practical, institutional, orsocial character and therefore are of a much more rela-tive nature than the characteristics related to categoriesand criteria of threat. Therefore, there are several validand different weighing vectors corresponding to differ-ent institutional and national situations, and it is difficultto accept that a single vector of weights can be appliedacross the board. In addition, de Grammont and Cuar´ononly allow for binary scores even though a more nuancedsystemwouldhavebeenbetter.Therefore,theirvaluationis just one among many possibilities.For example, in the Mexican case, one of the mainpoints of the method now in use by the Mexican Fed-eral Government (SEMARNAT 2002) is to accept onlywell-documented proposals of lists of threatened species.Therefore, characteristic 15 (“Systems include systematicand explicit records of information used”), to which deGrammont and Cuar´on assign a weight of zero, was infact a very important one for the Mexican government.Another characteristic of high importance for somecountriesisoneofpracticality(characteristic11:“Systemis flexible in terms of required information”). A govern-ment that has to grant or deny permits, fight for or con-test court rulings, assign fines, and even send people tojail, all within legally established periods of time, cannotstrive for theoretical perfection. It is impossible to basedecisions on conceptually satisfactory but fundamentallyimpracticalsystems,andthereforeacharacteristicofprac-ticality is important for governments. de Grammont andCuar´on acknowledge several times that the IUCN system(probably among the best from the perspective of an in-ternationalnongovernmentalorganizationwithoutactuallegalresponsibilities)isdifficulttoimplementinpractice,andevenmoresoinmegadiversecountrieslackingexten-sive scientific systems. Nevertheless, they do not followthis point to the conclusion that for some countries hav-ing a system that is impractical may be virtually valueless,whereas having a practical applicable system (relative toitseconomy,scientificapparatus,andbiologicalcomplex-ities) is very useful.This means that a country may have good reasons toassignahigherweighttocharacteristics11and15thantocharacteristic 5 (“Categories measure species probabilityof extinction”) or 13 (“System is applicable at differentgeographic scales, from regional to global”), for example.This fact was crucial in the process to decide whetherMexico should or could use the IUCN system. The de-cision was made that, being a megadiverse developingcountry, even with a relatively significant scientific base,Mexico was in no position to apply the IUCN categoriesand criteria to assess threat risk of most of its biodiversity.By assigning a weight of zero to characteristic 11 (whichthe authors did just for reasons of data availability), deGrammont and Cuar´on implicitly state that, in their val-uation, this practical consideration is irrelevant, whereasin real life it is essential.Our second disagreement, although this is not reallyfundamental, is with some of the scores assigned to SE-MARNAT (2002), the one system we know well. We dis-agreewith,andcouldprovidedetailedargumentsagainst,the score of zero given to characteristics 3 (“Definitionsof categories do not include subjective elements”) and