Five selection criteria were used to determine whether a species qualified for national listing. A species could qualify by meeting any one of the following criteria:
(1) Delphi. Species assigned a Delphi score of Moderate or High concern by more than 50% of all (minimum of 5) respondents. Of approximately 600 species evaluated, 72 met this criterion for listing; 63 were identified in the survey of American Birds regional editors and RMNBCs, and 9 were added in the followup survey of Alaska, Hawaii, and Puerto Rico/U.S. Virgin Islands "experts."
(2) Breeding Bird Survey(BBS). Species has an apparent long-term (1966-1993) continental population decline documented by the BBS that equals or exceeds 2.5%/year (P < 0.05). This is an effort to highlight species suffering chronic population declines that were not identified in the Delphi exercise. An estimated trend of -2.5%/year amounts to a cumulative decline of about 50% over a span of 27 years. A decline of this magnitude is considered to be biologically significant, even for species that are widely distributed and relatively abundant. A review of the approximately 424 eligible species in the BBS database revealed 81 species with significant long-term (1966-1993) population declines at a continental level [NOTE: To avoid biases due to small sample sizes, all species with a degree of freedom (i.e., total number of routes minus total number of states or provinces) of less than 50 were excluded from consideration. Also, two species that otherwise qualified for listing under this criterion (white-throated swift and boreal chickadee) were dropped from consideration because of concerns about the reliability of the BBS trend estimates (B. Peterjohn, pers. comm.)]. The logical next step would be to identify that subset of species with the most extreme population declines. However, when sampling variation is a substantial proportion of the total variation in a collection of estimates (as is the case with BBS data; Link et al. 1994), "the species identified by their extreme trend estimates are likely to be a poor representation of the species actually undergoing the greatest rates of population change" (W. A. Link and J. R. Sauer, unpublished manuscript), even when the assessment is limited to those species with statistically significant trends. To avoid the potential biases of sampling variation, we used constrained empirical Bayes estimates of population trend (W. A. Link and J. R. Sauer, unpublished manuscript) for the 81 species for which there was evidence of significant declines. Twenty-one species met this criterion for listing (including 12 that already qualified on the basis of Delphi scores).
(3) Aubudon Christmas Bird Count(ACBC). Species has an apparent long-term (1959-1988) continental population decline documented by the ACBC that equals or exceeds 2.5%/year (P < 0.01). As with the BBS, this is an effort to highlight species suffering chronic population declines that were not identified in the Delphi exercise. A review of the approximately 369 species in the ACBC database revealed 49 species with significant long-term (1958-1988) continental population declines. [NOTE: To avoid potential biases in the database, only species with sample sizes equal to or greater than 100 were considered. Also, four species that otherwise qualified for listing under this criterion (semipalmated sandpiper, black-legged kittiwake, "Baltimore" oriole, and rusty blackbird) were dropped from consideration because of concerns about the reliability of the ACBC trend estimates]. Five species met this criterion for listing (including 4 that already qualified on the basis on Delphi scores or BBS trends).
(4) Partners in Flight (PIF). Species has a composite PIF rank score of at least 24 in (a) 2 or more USFWS regions or (b) the USFWS region that contains at least 50% of the U.S. breeding range or population. Prioritized regional lists were reviewed for USFWS Regions 1-6 (unpubl. data provided by J. Bradley and M. F. Carter in November 1993), USFWS Region 7 (Alaska) (unpubl. data provided by B. Andres in October 1994), and Puerto Rico/U.S. Virgin Islands (unpubl. data provided by W. C. Hunter in November 1994). Each of these lists was prepared using a standardized format and ranking criteria (e.g., Carter and Barker 1993). PIF rank scores ranged from 7-34, with higher scores indicating greater degrees of concern. Within each region, about 10% of the species had scores of 24 or more. Thus, this criterion is designed to incorporate only those species of neotropical migrant landbirds for which the greatest degree of concern has been expressed by PIF (i.e., the highest ranking 10%). Of 254 species reviewed (see Research Working Group 1992), 68 had a score of 24 or more in at least one region, and 44 met this criterion for listing (including 20 that already qualified on the basis of Delphi scores, BBS trends, or ACBC trends).
(5) Candidate List. Species listed as a Category 1 or Category 2 candidate in a geographical area covering at least 10% of the U.S. breeding range. Of 92 taxa (species, subspecies, and populations) of birds listed as Category 1 or Category 2 candidates (USFWS 1994), 67 taxa (61 species) were eligible for consideration (i.e., they fell within the focus of this exercise) and 40 species met this criterion for listing (including 24 that already qualified on the basis of Delphi scores, BBS trends, or PIF scores).
[MBMO Home Page][Main Contents]