|
<br />/
<br />
<br />, ,..-'.
<br />
<br />~ If ~~'~mplicate the pictUre and say that one potential area has
<br />excellent habitat quality and the other has only good habitat quality.
<br />we see that it remains slighdy. advantageous to split the birds
<br />bctweell areas. Predicted probabilities of failure are 0:312 for
<br />excellent and 0.69S'for good habitat, res~vely. The pf<?bability
<br />that both translocations will fail is 0.312 X 0.698 = 0.218 com-
<br />pared to 0.257 for putting all birds in a single excellent habitat .
<br />quality area. In this example,. slight advantage to splitting the
<br />tranSlocated birds between areas is maintained down to a total
<br />release of 40 bird:s~ However. with so few birds released the
<br />probability that both ~Iocations will fail is increased to about
<br />0.42.
<br />The model coefficients in Table 2 may be used, to evaluate other
<br />scenarios. For ex3lJ)ple. given two alternatives. should a given
<br />number of birds be released in good habitat quality in the: core of the
<br />historical species range or in excellent habitat quality 'on the: '
<br />periphery or outside: the historical range? Good habitat: quality in
<br />:he core of the: range: is the bener choice re:gardless of the: number of
<br />birds released. This suggests that the physiological amplitude of a
<br />species may influence local population viability.
<br />
<br />Enhancing the Chances of Succc;ss
<br />
<br />Without high habitat quality. translocations have low chances of
<br />success regardless of how many organisms are released or how well
<br />~l,cy arc: prepared for the release. Active management is required:
<br />Limiting faci:orsmwt be identified and controlled and assurances of
<br />:n.1intenance of habitat quality obtained prior to translocation.
<br />Identification and retention of adequate habitat will require a
<br />combined species and ecosystem approach, Ecological information
<br />will be necessary to identifY critical life history traits. factors
<br />cctermining habitat quality. species interactions. and minimum
<br />
<br />Table ~ Stepwise logistic regression (24) model coefficients for predicting
<br />probablbty (P = 1/(1 + t-")) of success of intentional introductions or
<br />reintroductions (mnslocations) of native birds and mammals in Australia.
<br />Canada, Hawaii.,. Ncw' ~and the: United States between 1973 and
<br />198~; .~is the s~ of ~pplicable coefficients for categorical variabJes p~us the
<br /><lppbcablecocffiClcot nmcs the value of continuous variables. The model is
<br />t>.1SCd on 155 mnslocations; 100 wcrc: of birds and 55 were of mammals.
<br />Data were obtained from a survey condDCtedin 1987(15). The stepwise
<br />procedure was run at the: II = 0.10 level for enny of terms and the II = 0.15
<br />level for ranovaJ of terms. Probability of brger test statistics for the model
<br />were x~ P = 0.90 (24); Hosmer-Lcmcshow X2. P = 0,121 (24); BroWn's X2,
<br />~, = 0.::>37 (24). The model correcdy classified 81.3% of observed transloca-
<br />cons based ona cutpoint of 0,50 inprcdiaCd probability of success, '
<br />
<br />Variable
<br />
<br />Coefficient
<br />(SE)
<br />
<br />-1.418 (0.738)
<br />-0.9n (0.253)(1]*
<br />0.9n (0.253)[1)
<br />-0,919 (0.374)[6]
<br />0.919 (0.374)(6)
<br />
<br />1.681 (0.438)[2]
<br />0.053 (0.314)(2]
<br />-1.734 {O.450)(2]
<br />
<br />1.028 (0.267)[3]
<br />~ 1.028 (0.267)[3)
<br />1.080 (0.355)(5]
<br />-1.080 (0.355)(5]
<br />0.887 (0.405)(7]
<br />0.181 (0,074)[4)
<br />
<br />, Threatened, endangered, or sensitive species
<br />~ative game
<br />Birds
<br />!>famma/s
<br />Release area habitat
<br />Excellent
<br />Good
<br />Fair or poor
<br />Release area
<br />Q;>reof historic range
<br />Periphety or outside
<br />wly breeder. large clutch
<br />Late breeder, large clutch
<br />Log(nwnber rcJcased)
<br />Program length (years)
<br />
<br />.~wnbcis in brackets ~t order of entry.
<br />
<br />+ AUGUST 1989
<br />
<br />Fig. 1. Prcdiaed proba-
<br />bility of$UCCcssful trans-
<br />location as a function of
<br />the nwnbcr of arumals
<br />released during a 3-yeac
<br />: period in the core of the:
<br />historic, species range in
<br />either excellent (solid
<br />linc:)'orgoOd' (dashed
<br />line) habitat qUality for a
<br />threatcnc:d, endangered.
<br />~r sensitive: bird species
<br />400 that first ',' breeds at 2
<br />years' of age or more
<br />with average clutch size:
<br />ofthrcc or Icss. Probabilities are based on stepwise logistic regreSsion model
<br />coefficients (Table 2).
<br />
<br />1.0
<br />
<br />:: 0,8
<br />.
<br />..
<br />..
<br />~ 0,6
<br />'0 ",
<br />:0-
<br />S 0.4
<br />A
<br />.3.;.. . ~
<br />~ 0.2 ,/----
<br />
<br />
<br />
<br />--'~-.----:--,-.-'--:---:-:--
<br />
<br />0.0
<br />2
<br />
<br />100' 200 300
<br />, Number 01 birds releasitd
<br />
<br />"
<br />habitat fragment size (28). ,Regional approaches to maintaining
<br />diversity (29) will be essential to ensure that existing species and
<br />habit;lt assemblages are identified. their interactions arc: understood,
<br />and, remnant habitats are protected. The laner approach may
<br />ultimately reduce the number of species that require, translocation if
<br />it enhances understanding ofth~ effects,ofhabitat fragmentation on
<br />persistence of multiple disjunct populations.:, ,
<br />We may reduce the ,need foraqd incrc;:ase the success oftransloca-
<br />tions if we can improve our ability, to identify potentially tenuous
<br />sirnations and act before we ar~ faced with a rescue. Simulation
<br />modding (28, 32) of the behavio~ of small populations of species or
<br />of groups of species with simil:tr reproductive strategies can provide
<br />guidance for establishing min4num population and vital rate goals,
<br />Sif!lulations will be most productive if set in a regional context that
<br />addresses the, interaction' among, metapopulations and the spatial
<br />rel~tion among reserves or poterltial release sites (28).
<br />The asymptotic nature of die relation between tranSlocation
<br />success and number of animals ~e1eased emphasizes the point that
<br />releasing large numbers of anil11~ does little to increase the success
<br />of translocations. Lack of demo1'JStrated succress after translocating
<br />large' numbers of animals is caufe for reevaluating other variables
<br />associated with success.
<br />'. 'The asyniptotic levels do sugg~tthat there: isa'minimumnumbcr
<br />'of 3nUnalsthat should be 'id~c:d, Because longer tranSlocation
<br />programs are more successful (T*ble 2). the minimUm number may
<br />be released over severaJ years if in~fficient animals are available for a
<br />single' releaSe: Capnvc rearing pn!>gramsthat are focused on translo-
<br />cation should have the goal of qrablishing multiple self-sustaining
<br />populations so they can provide sufficient animals ()ver a number of
<br />years and increase' the successotlthese~pe~ive(2, 3) programs.
<br />Those planning rranslocationS~hould'adopt rigorous data record-
<br />ing procedures (19, 30). Dcuils <i'f tranSlocation attempts should ~
<br />assembled in a database. It is criqcal that both failures and successes
<br />be adequately documented. Pe~t-granting' agencies may need to
<br />assume, the role of ensuring tha~ adequate records are kept so the
<br />database can be incrcasc:dand p*dictability of success enhanced.
<br />Because of the low success of p-anslocatior)s' of small numbers of
<br />endangered. threatened, or sensipve species, even in excellent habi-
<br />tat quality, it is clear that rrans'ocation must be considered long
<br />before it becomes a last resort for these speci~ore density has
<br />., become low and POPulati9ns arf in decline; Both these traits are
<br />associated with low chances of Sl\ccessful traJ1Slocation.' In addition.
<br />obtaining 'sufficient' num~ :~f ~a1stoachieve" reasc:'nable
<br />"chances' of success may bCunpcpsslble.'The. greatest potentIal for
<br />establishing "satellite populatio$ ,may. ,ocCUr when a candidate
<br />population is expanding and nwPbers are n$1erate to high. These
<br />conditions arc: the ones that t~d to make endangered species
<br />biologistsrclax; our analysis sUggests that ,these conditions may
<br />point oqt the PIne fQr.~()n.,! . ' _
<br />
<br />ARTICLES 479
<br />
<br />.
<br />!
<br />.
<br />i
<br />t
<br />I
<br />"
<br />
<br />I
<br />I
<br />I
<br />
<br />
<br />
<br />f
<br />
<br />1.
<br />I
<br />
<br />I'
<br />i
<br />I
<br />I
<br />I
<br />~
<br />~
<br />
|