Predictors of avian and mammalian translocation success:reanalysis with phylogenetically independent contrastsC. Magdalena Wolfa,*, Theodore Garland, Jrb, Brad Grithc,yaDepartments of Zoology and Wildlife Ecology, 430 Lincoln Drive, University of Wisconsin, Madison, WI 53706-1381, USAbDepartment of Zoology, 430 Lincoln Drive, University of Wisconsin, Madison, WI 53706-1381, USAcUS National Biological Service, 101 12th Avenue, Box 15, Fairbanks, AK 99701, USAReceived 27 November 1996; received in revised form 20 October 1997; accepted 23 November 1997AbstractWe use the phylogenetically based statistical method of independent contrasts to reanalyze the Wolf, C.M., Grith, B., Reed, C.,Temple, S.A. (1996. Avian and mammalian translocations: update and reanalysis of 1987 survey data. Conservation Biology 10,1142±1154). translocation data set for 181 programs involving 17 mammalian and 28 avian species. Although still novel in con-servation and wildlife biology, the incorporation of phylogenetic information into analyses of interspeci®c comparative data iswidely accepted and routinely used in several ®elds. To facilitate application of independent contrasts, we converted the dichot-omous (success/failure) dependent variable (Wolf et al., 1996; Grith, B., Scott, J.M. Carpenter, J.W., Reed, C., 1989. Transloca-tions as a species conservation tool: status and strategy. Science 245, 477±480) into a more descriptive, continuous variable with theincorporation of persistence of the translocated population beyond the last release year, relative to the species' longevity. Forcomparison, we present three models: nonphylogenetic multiple logistic regression with the dichotomous dependent variable (themethod used by Wolf et al. 1996 and Grith et al. 1989), nonphylogenetic multiple regression with the continuous dependentvariable, and multiple regression using phylogenetically independent contrasts with the continuous dependent variable. Results ofthe phylogenetically based multiple regression analysis indicate statistical signi®cance of three independent variables: habitat qualityof the release area, range of the release site relative to the historical distribution of the translocated species, and number of indivi-duals released. Evidence that omnivorous species are more successful than either herbivores or carnivores is also presented. Theresults of our reanalysis support several of the more important conclusions of the Wolf et al. (1996) and Grith et al. (1989) studiesand increase our con®dence that the foregoing variables should be considered carefully when designing a translocation program.However, the phylogenetically based analysis does not support either the Wolf et al. (1996) or Grith et al. (1989) ®ndings withrespect to the statistical signi®cance of taxonomic class (bird vs mammal) and status (game vs threatened, endangered, or sensitive),or the Grith et al. (1989) ®ndings with respect to the signi®cance of reproductive potential of the species and program length.# 1998 Elsevier Science Ltd. All rights reserved.Keywords: Birds; Comparative method; Mammals; Phylogenetically independent contrasts; Reintroduction; Translocation1. IntroductionNumerous methodological, environmental, species-speci®c, and population-level factors in¯uence whetherthe intentional release of wild-caught or captive-rearedanimals into new locations will result in the successfulestablishment, re-establishment, or augmentation of awild population. Because speci®c causal factors andtheir relative importance vary widely among suchrelease programs (herein referred to as translocations,following Grith et al., 1989; see also Wolf et al., 1996),it is dicult to identify general trends associated withsuccess. Nevertheless, both theoretical considerationsand empirical evidence suggest that some methodologi-cal and biological factors are of general importance. Forexample, such methodological details as raise andrelease procedures, number and composition of animalsreleased, and choice of source stock for the releasedanimals were shown to in¯uence trans location outcomesin single-species studies (Beck et al., 1991; Allen et al.,1993; Reed et al., 1993; Bright and Morris, 1994; Velt-man et al., 1996). Environmental factors perceived asBIOLOGICALCONSERVATIONBiological Conservation 86 (1998) 243±2550006-3207/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved.PII: S0006-3207(97)00179-1* Corresponding author at: P.O. Box 52918, Riverside, CA 92517.yPresent address: Alaska Cooperative Fish and Wildlife ResearchUnit, 216 Irving I Building, University of Alaska, Fairbanks, AK99775.important to success include general habitat quality andclimatic conditions (Lindenmayer, 1994; Veitch, 1994),as well as the absence of predators or competitors(Crawley, 1986; Short et al., 1992). Some species-speci®cand population characteristics considered favorable forsuccessful invasion of a new location include a relativelyhigh reproductive potential, an omnivorous diet, smallbody mass, and high genetic diversity (Mayr, 1965;Laycock, 1966; Berger, 1972; Smith et al., 1976; Craw-ley, 1986; Ehrlich, 1986; O'Connor, 1986). The questionremains: how universal is the in¯uence of such metho-dological and biological factors? Also, in light of sub-stantial case-by-case variations, are generalizationsacross a range of taxa possible?Grith et al. (1989) (see also Grith et al., 1990,1993) used a comparative app roach to test for generalpatterns underlying the success vs failure of transloca-tions among species of birds and mammals. They con-ducted surveys of translocation programs throughoutNorth America, Australia, and New Zealand; coded theoutcomes as either a success (reported establishment ofa self-sustaining population), a failure, or incomplete;and used multiple logistic regression to identify sevenstatistically signi®cant predictors of success: (1) taxo-nomic class (bird vs mammal), (2) legal status of thetranslocated species (native game vs threatened, endan-gered, or sensitive species), (3) habitat quality of therelease area (excellent, good, or fair/poor), (4) locationof the release area relative to the historical range of thespecies (core vs periphery or outside), (5) number ofanimals released (log10transformed), (6) programlength (number of years over which releases occurred),and (7) potential productivity of the translocated species(high vs low). Wolf et al. (1996) conducted a follow-upsurvey, in