|
<br />/
<br />
<br />/.-~
<br />
<br />~
<br />
<br />r
<br />
<br />animals (3, 19) are also considered prerequisites of a successful
<br />translocation.
<br />We found thar several rfaaors were associated with success of
<br />translocarions (fable 1). Native game species were more likely to be .
<br />successfully tranSlocated than werethrc:atened. endangered. or
<br />sensitive species. Increased habitat quality was associared with
<br />greater success. Translocations into the core of species historical
<br />ranges were more successful than were those on the periphery of or
<br />outside historical ranges. Herbivores were more likc:ly to be success-
<br />full)' translocated than either carnivores or omnivores. Trans!oca-
<br />tions intQ areas with potential competitori of similar life fann were
<br />less successful than rrans1ocations into areas \\ithout competitors or
<br />areas, with a congeneri~ potential competitor. Early breeders with
<br />large clutches were slightly more likely to be successfully translocat-
<br />ed than were species that bred late and had small clutches.
<br />Translocations of exclusively wild-caught animals were more
<br />likely to succeed than were those of exclusively captive-reared
<br />animals (fable 1). Among translocations of exclusively wild-caughr
<br />animals, success depended (P:S 0.10) on whether the source popu-
<br />lation density was. high (779ft success, " = 109), medium (78%,
<br />II = 37), or low (37%, " = 8). Success ohranslocations of wild:-
<br />caught animals was also associated (P s 0.10) with whether the
<br />source population was increasing (83% success, " = 93), stable
<br />(63%, n = 49), or declining (44%,,, = 9). Successful translocations
<br />released more animals than unsu(:cessful translo-::ations (160 com-
<br />pared to 54, respecrively;P = 0.024).
<br />Our resullts are consistent with analyses of naturally invading or
<br />COlOnizin~' pc:cies that show (i) larger founde:r populations ,are more
<br />successful ( ,21), (ii) that habitat suitability is important (21), and
<br />(iii) incr ed number and size of clutches enhances successful
<br />invasion (2f)' Our data also support the h}'J>Othesis that herbivores
<br />
<br />
<br />Table 1. perentage success ofintcnrional introductions or reintroductions
<br />(tr3llslocatio s) of narive birds and mammals 10 the wild in AUstralia,
<br />C3Ilada, Ha '3ii, New Zc.1bnd, and the United States between 1m and
<br />1986, Data . ere obtained from a survey conducted in 1987 (15), The data
<br />include 134 ~ransJocations ofbirds and 64 translocations of mamnuls_ For ;ill
<br />\'ariables list~d, ,twu statistically signiliant (P :S 0.10),implyingtnJC,
<br />, dilTen:ncesiq the percentages of suCccssful tt;iiU\ocations among the catego-
<br />ries. Anim~that first give binh at age 1 or kss with average clutch sizcof
<br />three or mor~ are considered arly brccdcrs with large clutches; all othcrsare
<br />late brcc:dersiwith small clutches.
<br />
<br />Variabk
<br />
<br />T cans- Succcss
<br />locations
<br />(II) (%)
<br />80 44
<br />118 86
<br />63 84
<br />98 69
<br />32 38
<br />133 76
<br />54 48
<br />163 75
<br />34 38
<br />40 48
<br />145 77
<br />13 38
<br />102 7S
<br />96 62
<br />39 72
<br />48 52
<br />105 75
<br />
<br />
<br />T1uC'atened, ~ndangered, or ~nsitive species
<br />:-:ati\'e gamei
<br />Rdeasc: area Ihabitat
<br />Excellent i
<br />Good .
<br />_Fair or pobr
<br />Location of tdcase
<br />" Core of h~toric range
<br />, Periphery pr outside
<br />Wild-caught i
<br />Captive-rear+t
<br />Adult food Habit
<br />Cami~'bre l
<br />,H~rblvorci'"v:
<br />_Omnivorei
<br />Early brccdct, laric clutch
<br />Late breederl small clutch
<br />Potential corhpctitors
<br />Congeneri~
<br />Similar i
<br />Neither i
<br />I
<br />i
<br />
<br />I
<br />
<br />478
<br />
<br />are more successful invaders than carnivores (17) and the conclusion
<br />that, for birds, morphologically similar speCies have a greater
<br />depressing effect on successful invaSi9n than do congeneric species
<br />(~. ,
<br />We found no consistent association oftraI1slocation success with
<br />number of releases, habitat i~provcment, whether the release was
<br />hard (no food and shelter provided on site) or soft,immediate: or
<br />delayed release on site, or average physical condition of animals at
<br />release. We were un.able to directly evaluate genetic he:terogeneity.
<br />sex and age composition, o~ specific rearing and handling proce-
<br />dures for released animals because of inadequate response: to survey
<br />questions.
<br />
<br />Evalua~g Alternative Strategies'
<br />
<br />Analyscs of individual factors associated with translocation suc-
<br />cess dci not adequately reflect the multivariate' nature of actUal
<br />translocations. To overcome this problem, we used stepwise logistic
<br />regression (24, 25) to develop preliminary predicti\'e' equations for
<br />estimating the success of trans locations (Table 2). An cxpandeddata
<br />set or independent sample would probably rield diffen:nnegression
<br />coefficients and estimates of success than we report. As a result,
<br />extrapolation to conditions much diffe:rent than those represented
<br />by our data and applications to individual species are discouraged.
<br />The: coefficients from Table 2 can be used to plot predicted succesS
<br />of different kinds' of translocations as a function of continuous
<br />\'ariable:s such as the: number released, We: prescnt an example for a
<br />threatene:d,endangered,or sensitive bird (Fig. 1).
<br />This exercise (Fig, 1) illustrates that the increase in success
<br />associated with releasing larger numbers of organisms' quickly
<br />becomes asymptotic. Releases larger than 80 to 120 birds do little: to
<br />increase: the chances that a translocation will be successful for this
<br />particular set of conditions. The asymptotic property is consistent
<br />across other classifications of the data but the inflection point ,.aries_
<br />For large nati\'e game m:lmmals the asymptote is reached at releases
<br />of 20 to 40 animals with aconcurrentlv higher prediC:i:ed,sllccess.,'
<br />The as}'mptoricproPerty of thca'ssociation of 'translocation
<br />success and number released (Fig. 1) is consistent with theoretical
<br />predictions (13) and analytical treaune:ncs (26) to'lat suggest a
<br />threshold population size below which extinction is li~e1y, primarily'
<br />due to chance ~ents affecting birth and death of in4ividuals;,The
<br />, existence of the inflection (Fig. 1) is also consistent with the
<br />prediction ofathrcshold density below whichpopuJationsocial,
<br />inte:ractions and mating succc.sSare disrupted (27), again Ie:ading to
<br />diminished population viability.
<br />The coefficients from Table 2 and relationships presented in Fig. 1
<br />can be used to asse:ss alternative strategics-Suppose: 300 threatened
<br />and endangered birds are available for a translocation program and
<br />they must be released during a 3-year time frame:.-Funher suppose
<br />that two pote:ntial translocation are:aSare available within the core of
<br />the species historical range. If the goa,1 of the: translocation is to
<br />'establish at least one geographically disjunct population to reduce
<br />the risk of catastrophic loss of the species, how should the birds be
<br />distributed betwee:n the two potential translocation areas to mini-
<br />mize the: probability that botlltranslocations will f~il?
<br />If both releasearcas havc-cxcellent, habitat quality, and the: are:as
<br />arc inde:pendent,the answer-is obvious. The birds should be divided
<br />betwe:en the: areas. The coefficients from Table 2 allow us to estimate
<br />the probability that a single: release of 300 birds will fail (1.0 minus
<br />probability of success) is 0.257. Two rele:asc:s of 150 birds~a~h have
<br />individual probabilities of failure of 0,312. The probability, that
<br />both will fail is 0.312 X 0.312 = 0.097; substantial gain is achieved
<br />by splitting the birds between areas.
<br />
<br />c::rn=l'\J'rJ:' vnT ?A(
<br />
<br />
|