Reintroduction: Challenges and Lessons for Basic Ecology

PERSPECTIVES

This kind of misunderstanding prob- ably explains the poor level of involvement Reintroduction: challenges and by ecologists, and particularly population biologists, in reintroduction programs - lessons for basic ecology only slightly more than half of the reintroductions involve professional biologists’. Franqois Sarrazin Academic laboratories generally face constraints preventing them from sustaining Robert Barbault long-term and costly applied projects on their own. On the other hand, as Lindburglr Ecology is a subject where theoretical predictions are often difficult to test emphasized, many biologists ‘outside the experimentally in the field. To address this challenge, the Ecological Society community of zoo professionals’ argue of America suggested exploltlng large-scale envlronmental management decisions in that the large amount of money necessary a scientific way. This ‘adaptive management’ constitutes one of the purposes of the for single species reintroduction could be Sustainable Biosphere Initiative. Meanwhile, In the current context of the blodiversity spent far more usefully to protect whole crisis, translocations and particularly reintroductions of threatened species are ecosystems from destruction. They also becoming more numerous. It Is time for ecologists and wlldllfe managers underline that environmental conditions to collaborate on these unique opportunltles for large-scale studles. at the time of release are rarely exactly the same as those prevailing before extinc- Francois Sarrazin and Robert Barbault are at 1’UniversitCPierre et Marie Curie, tion, and the extinction causes are often Laboratoire d’Ecologie,URA 258 CNRS,BZt A. 7eme Ctage,7 Quai Saint Bernard, difficult to determine and to quantifyis. BP 237,75252Paris Cedex 05, France ([email protected]). Furthermore, reintroductions in devel- oped countries - particularly reintroductions of large vertebrates - generally sup he threats to biodiversity resulting real-scale hypothetico-deductive experi- pose that the future population will be Tfrom human activities, and the under- ments in ecology. artificially maintained at least for some standable reluctance to perturb their time because of continuing human impact study subject, often prevent ecologists The mutual scepticism of field on the ecosystemia. This is against the from developing their own large-scale ex- conservationists and population IUCNrecommendations, but often remains periments. Therefore, long-term monitor- biologists difficult to prevent. In many cases, the ing or experiments on model species are Despite the recent development of con- reintroduced populations contain few in- often used, but this may limit the range of servation biology as a science’a-12,the gap dividuals, at least during the first years, testable hypothesesi-3. Meanwhile, faced between field conservationists and scien- and they may be difficult to observer. with the increasing numbers of trans- tists persists. Caughley” distinguished two Moreover, some behavloural traits exhib- location&s, the International Union for approaches in conservation biology: the ited by these individuals can be altered by Conservation of Nature (now IUCN- The declining population paradigm and the captivity or human imprinting, or by low World Conservation Union) proposed in small population paradigm. The first inves- density of conspecifics. These negative 1987 to define reintroduction as the intro- tigates extinction processes empirically effects may vanish in the first wild-born duction of a species in a previously occupied case by case, and the second theoretically generation, at least for species in which area in order to improve the conservation considers the consequences of smallness social learning and cultural transmission of the species. Reintroduced individuals on population viability. In restoration ecol- are low. However, demographic and even may come from captive breeding pro- ogy1sJ4, reintroduction suffers from the behavioural studies require long-term ef- grams or be translocated from natural same discrepancy between practical works fort, to achieve sufficient sample sizes and populations. The IUCNalso recommended published in grey literature (see Ref. 15) relevant observations in the second gen- that reintroduction programmes incorpo- and a comparatively poor theoretical eration. The possibility of controlling and rate feasibility studies and preparatory, background. replicating should also be considered. To introduction and monitoring period@. Practical reasons explain this situation. assess the particular effect of population Moreover, restoration of the original habitat Non-governmental organizations, which foundation on any biological trait, it is de- and amelioration of causes of extinction usually initiate reintroduction programs, sirable to study a control sample in natu- were considered as essential conditions are supported by sponsors requiring maxi- ral conditions. This is possible in the case for these projects. Unfortunately, the mum efficiency in terms of time and cost. of local reintroduction -that is, where rem- monitoring period that should follow re They aim at protecting and restoring bio- nant populations persist in other areas - introductions often remains neglected4JJ diversity, but the understanding of the but not when reintroduction constitutes or is documented only in ‘grey’ literature. mechanisms of population extinction and the only way to restore a species that is Our purpose is to show that, in order growth is not their first priority. Moreover, extinct in the wild. to combine the recommendations of the Kleiman et al.6 suggested that reintroduc- Overall, the scepticism of scientists IUCN and the Sustainable Biosphere tion success seemed more linked to bio- results from the basic problem of reintro- Initiativeg,reintroduction programs should political conditions and long-term funding duction: the initial decision concerning re- systematically include ecologists who are than to scientific rigour. Obviously reintro- introduction is usually a pragmatic choice proficient in population biology and gen- duction in a hostile human contexti6, and and not a scientific one. Indeed, even when etics, behavioural ecology and evolution. with low funding, would be very unlikely the IUCN conditions are respected, the They could then generate a fruitful ap- to succeed whatever the biological back- decision of release remains a political one. proach for testing hypotheses in basic ground, but once funding and local agree- Because the role of biodiversity in gen- ecology, and particularly in population ment are obtained, the success of reintro- eral, and of one species in particular, is biology (Box 1). Without overlooking con- duction remains a question of population often hard to establish from a strictly servation issues, reintroduction programs viability involving demographic, genetic, scientific point of view, philosophical, aes- could provide important opportunities for behavioural and ecological processes. thetic or more economic reasons usually

474 0 1996, Elsevier Science Ltd PII: SO169-5347(96)20092-S TREE vol. I I, no. II November I996 lead to such projects. Therefore, ecolo- ment of reintroduction methods, which demographic parameters can be used to gists generally have fewer arguments in can be useful in other project@. Recently, determine what kind of individuals should favour of reintroduction than against it. Cade and Temple’s reviewed the success be protected or which measures should We want to show that the planning and of numerous bird reintroduction programs be adopted to maintain and encourage the monitoring of reintroduced populations according to the methods involved. It expansion of a population. may be relevant for both conservationists appeared, however, that even the success Again, population dynamics models and ecologists provided that they work criterion was not clearly defined among that integrate demographic parameters togetherGJ9. authors. For example, Craig and Reed (in and behavioural data recorded in the field Ref. 15) think that a three-year breeding can lead to simulations and tests of a priori Reintroduction preparation and population with a recruitment rate that is hypotheses. The use of previous results knowledge of species biology higher than the death rate of adults de- to design further decisions is a central Whenever reintroduction managers scribes successful reintroduction. Accord- concept of adaptive management, that is, use scientific tools, the questions they aim ing to Minimum Viable Population stud- learning by doing2. In that purely applied to answer are generally focused on the ies, Beck et af.7 consider 500 free-living context, population ecologists should preparation phase of translocations. The individuals as representing good success therefore collaborate on reintroduction challenge of captive breeding20 requires and regret the low availability of published programs. studies of the genetic consequences of in- results using this criterion. However, with- Some projects on Arabian oryx (Oryx breedingzi, behavioural consequences of out taking into account life history traits, feucoryx)23, black-footed ferrets (Mustela captivity22, and population consequences habitat quality or the eventual metapopu- n@-ipes)16~35or golden lion tamarin (Leonto of infectious diseasesi8, involving mainly lation structure (which varies widely pithecus rosafia)30are now recognized for veterinary and zoo biologists. Many efforts between reintroduced populations), this having been precisely designed and well are dedicated to the maintenance of high threshold of 500 individuals seems rela- documented and many others are just too levels of genetic variation in captive-bred tively arbitraryis. Since the settlement of a recent to give relevant results. For both and reintroduced populations. This may self-sustaining population corresponds to preparation and monitoring of reintroduc- be achieved by selecting individuals with a dynamic process, extinction probability tions, increasing contacts between aca- known pedigreeszs, individuals with high estimates that combine population size, demic population biologists and wildlife allozyme heterozygosity or individuals growth rate and growth rate variance28 managers is encouraged within the Sur- coming from geographically separated should be the main criteria for assessing vival Species Commission (SSC, IUCN), populations24. However, in the last case, this success. This necessitates the accu- especially the Conservation Breeding Spe- the consequences of outbreeding de- rate estimate of demographic parameters cialist Group (CBSG) and the Reintroduc- pression must be consideredzs. Moreover, such as survival rates.29and the modelling tion Specialist Group (RSG) (Ref. 37). Such reintroduced individuals lack locally se- of various dynamics scenarios, including forums, including Conservation Assess- lected traits that are likely to have existed unexpected catastrophic events. A shorter ment and Management Plans (CAMPS), in the extinct population@. In the case of term criterion to assess this success could Global Captive Action Plans (GCAPs) and population reinforcement, the impact of re- be the breeding of the first wild-born gen- Population and Habitat Viability Assess- leases on the residual population implies eration in the release areaso. Clearly, elab- ment (PHVA), have already covered vari- caution regarding the origin of translo- orating success criteria for reintroduction, ous aspects of status assessment, captive cated individuals (e.g. Ref. 26). at least similar to Mace’s31 criteria for breeding, and reintroduction opportuni- This bias towards genetics should not threatened species, is an urgent challenge ties and organization for more than 2100 mask the need for a priori knowledge about for theoretical population biology. taxa38.They are often dedicated to whole life history traits of reintroduced specie&6. Results from already reintroduced groups of endangered species with inter- Caughleyn underlined that endangered populations can also be important for fur- national involvement, but they constitute species, which should be the primary tar- ther project preparation. By recording as good examples for more permanent scien- get of translocations, are also among the much relevant information as possible on tific collaborations at the local scale. least understood. He also argued that half potential release candidates (such as ori- the books on conservation biology pub- gin, sex, age, breeding status), it becomes Reintroduction as a way to lished in the 1980s were about genetics. possible to model the efficiency of different experiment in ecology However, demographic and environmen- release strategies using simulations with The two previous points concern what tal stochasticity are also likely to have a various survival and reproductive rates ecologists can contribute to reintroduc- short-term impact on very smallilJ~J7 and obtained from the monitoring of reintro- tions. However, reintroduction monitor- reintroduced populations. Moreover, all duced populations 32,33.Similarly, surro- ing can also constitute a framework for biological factors able to play a role in gates have already been used to improve more-theoretical ecological studies in an population viability do so by acting on sur- release strategies of the Californian con- extended view of ‘adaptive management’. vival and fecundity rates, whatever their dor (Gymnogy~scalifornianus) and black- First, restoration has already been genetic, behavioural or environmental footed ferrets (Mustela nigripes)35. Outside identified as an ‘acid test for ecology’ origin. Therefore, a good knowledge of the the conservation area, numerous game spe- since its success checks our understand- demography of a species is crucial to as- cies translocations4 could be exploited as ing of ecosystems39. Reintroduction could sessing the possible effects of all factors experiments in order to improve our under- therefore become an ‘acid test for popu- affecting reintroduction success. In that standing of reintroduction requirements. lation biology’. way, comparative studies in population Moreover, once the new population is Second, reintroduction monitoring biology can provide useful information. settled, wildlife managers have to judge could be a good opportunity for evolution- the feasibility of further releases, the re- ary biologists to complement their knowl- Reintroduction monitoring as an laxation of help for already released in- edge of processes among the wide range of assessment of reintroduction dividuals, and the long-term viability of plant and animal models. Indeed, the char- methods the population. In that way, in stage or acteristics that lead some species to be Monitoring of reintroduced popu- age-structured population models, the reintroduced make them interesting mod- lations after releases provides an assess- analyses of growth rate sensitivity28JG to els for biologists. Reintroduction concerns

TREE ool. II, no. II November 1996 475 PERSPECTIVES species that disappeared recently under Californian condor (Gymnogyps califomi- concerning their behavioural, population human impact on the environment. In many anus), white-tailed sea eagle (Haliaeetus and community ecology. cases, extinctions are due to the disturb afbicilla), griffon vulture (Gyps f&us), Third, behavioural ecologists may use ante of whole habitats - particularly iso- Przewalski’s horse (Equus przewalskii), reintroductions to study newly founded lated island habitats. Although knowledge white rhinoceros (Ceratotheriumsimum) or populations in which the origin of each of species status varies among taxa40, orang-utan (Pongo pygmaeus)+5J5. These individual is known. They can be aged, it appears that extinction rates resulting large species often display the most ex- sexed and individually identified by gen- from interactions between environment treme life history traits among the pool of etic methods of fingerprinting, which allow, and species characteristics are higher for endangered species such as low repro- for example, studies of mating strategies. habitat specialists or long-lived species ductive rate, late age at maturation, long Morton43 planned experimental introduc- exhibiting extreme life history traits11a41. generation times, low density, and so on. tions of wrens (Cyphorhinusphaeocephalus Moreover, whatever the motivations of Because of these characteristics, they and Henicorhina leucosticta) to study the their choice, conservationists often choose cannot be studied in the laboratory and settlement abilities, habitat selection and flagship species to be reintroduced. De- natural populations cannot easily be ma- vital requirements of these species at an spite the warnings of taxonomists42 that nipulated and studied considering their individual level. Komdeura assessed the many invertebrates are disappearing each endangered status. Consequently, reintro- role of habitat saturation and territory qual- year, most reintroduction efforts concern duction programs constitute unique ity in the evolution of cooperative breed- ‘charismatic’ megafauna: for example, opportunities for experimental work ing by translocating Seychelles warblers (Acrocephalus sechellensis) in a recovery context. Release strategies can allow the Box 1. How reintroductions may contribute to ecology comparison of groups coming from different environmentsss. The study of the ap pearance of social transmission in newly Basic ecology Applied ecology founded groups may also be considered. \ Fourth, reintroduced populations may Understanding of show many characteristics induced by ecologicalprocesses colonization and founder effects that remain rarely studied in the wild, although being strongly emphasized in metapopulation studiesds. To mimic the natural colonization that is likely to occur in metapopulation, three conditions should be fulfilled: (1) the biology of released individuals should not be altered by captivity and hypotheses or translocation; (2) the released individ- + uals should mimic propagules (age and sex of dispersers); and (3) their number should be likely to occur in nature. The as an experiment impact of translocation on the biology of propagules may be assessed by comparison with second generation individuals29s46J0, lntroducbon+ j--j Success aFment 1 and amelioration of translocation effects remains one of the challenges of reintro- + ductions, Similarly, the intrinsic value of individuals as dispersers can be discussed Monitoring - introduction experiments only test settle ment ability not dispersalhr. Furthermore, numbers of releases are often maximized in order to increase reintroduction success inducing a breach of the third point. However, assuming these conditions, the a priori constitution of future released co- horts permits tests of different hypotheses concerning settlement mechanisms, such ‘Extended adaptive management’ ‘Adaptive management’ as the importance of conspecific attrac- [email protected], demographic and genetic Learning ecology Learning management monitoring of the reintroduced population through management through management may provide potential assessments of

By preparing hypotheses and designing releases as experiments, the monitoring of the reintroduced popu bottleneck and founder effects. In the same lation can provide good data and modelling opportunities. These analyses can lead to both understanding way, the appearance of density-dependent (dashed arrows) and action (solid arrows). The success of the reintroduction can be assessed, and this pro processes resulting from increasing intra- vides feedback on the feasibility, preparation and introduction methods used previously. The reintroduced specific or interspecific competition, and population can be managed in order to avoid its future extinction. Moreover, in a basic context, behavioural, the revision of the controversial carrying demographic and genetic hypotheses on population and community ecology can be explored via such a compromise between laboratory and nature. In the long term, this could help us to understand how extinction capacity concept could be assessed by occurs in a natural context. monitoring reintroduced populations. Finally, reintroduction programmes may facilitate studies of the role of

476 TREE vol. 11, no. II November 1996 PERSPECTIVES keystone species in community ecology. References 21 Hedrick, P.W. (1992) Genetic conservation in Rosenzweigdg argued that it is generally 1 Roush,W. (1995)When rigor meets reality, captive populations and endangered species, impossible to study community ecology Science 269,313-315 in Applied Population Biology (Jain, S.K. and by looking at small perturbations. Only 2 Walters, C.J.and Helling,C.S. (1990) Botsford, L.W., eds), pp. 45-68, Kluwer removing or adding species in the wild Large-scale management experiments and 22 Shepherdson, D. (1994) The role of would be likely to give significant patterns. learning by doing, Ecology71,2060-2068 environmental enrichment in the captive 3 Nichols, J.D. (1991)Science, population breeding and reintroduction of endangered Obvious ethical arguments usually pre- ecology, and the management of the americau species, in Creative Conservation.. Interactive vent ecologists from doing this outright, Black Duck, J. Wildl. Manage. 55,790-799 Management of Wild and Captive Animals but reintroductions provide a way. By 4 Griffith,B. et al. (1989) Transiocation as a (Olney, P.J.S., Mace, G.M. and Feistner, A.T.C., monitoring ecosystems during the fea- species conservation tool: status and eds), pp. 167-177, Chapman &Hall sibility and preparatory periods advised strategy, Science 245,477-480 23 Stanley-Price, M.R. (1989) Animal by the IUCN,and considering, for example, 5 Wilson, A.C. and Stanley Price, M.R. (1994) Re-introductions: the Arabian Oryx in Oman, disturbance in food webs or competition Reintroduction as a reason for captive Cambridge University Press in any niche dimensions induced by the breeding, in Creative Conservation: interactive 24 Leberg, P.L. (1993) Strategies for Population reintroduced species, it seems possible to Management of Wild and Captive Animals Reintroduction: Effects of Genetic assess the role of these species at the (Olney, P.J.S., Mace, G.M. and Feistner, A.T.C., Variability on Population Growth and Size, eds), pp. 243-264, Chapman &Hall Conserv Biol. 7, 194-199 community and perhaps the ecosystem 6 Kleiman, D.G.,Stanley Price, M.R.and Beck, B.B. 25 Templeton, A.R. (1986) Coadaptation and level. (1994) Criteria for reintmductions, in outbreeding depression, in Conservation Deliberate or accidental invasions have Creative Conservation: Interactive Management Biology: The Science of Scarcity and Diversity already constituted opportunities for eco of Wild and Captive Animals (Olney, P.J.S., (Soul& M.E., ed.), pp. 105-l 16, Sinauer logical studiesls. However, all individuals Mace, GM. and Feistner, A.T.C., eds), pp. 26 Hedrick, P.W. (1995) Gene flow and genetic of a reintroduced population may be of 288-303, Chapman &Hall restoration: the Florida Panther as a case known origin. The size of the original popu- 7 Beck, B.B. et al. (1994) Reintroduction of study, Conserv Biol. 9,996-1007 lation is known exactly and this allows captive-born animals, in Creative Conservation. 27 Lande, R. (1988) Genetics and demography in predicted models of population dynamics Interactive Management of Wild and Captive biological conservation,Science 241,1455-1460 and genetics to be tested. Moreover, demo- Animals (Olney, P.J.S., Mace, G.M. and 28 Ferriere, R., Sarrazin, F. and Legendre, S. Feistner, A.T.C., eds), pp. 265-286, Chapman Matrixpopulation models applied to graphic parameters can be estimated, at & Hall viability analysis and conservation: Theory least during the first years, for the overall 8 Scott, J.M. and Carpenter, J.W. (1987) Release and practice with ULMsoftware, Acta Oecol. population and not only for an indetermi- of captive-reared or translocated (Spec. Vol.) (in press) nate part of it. endangered birds: what do we need to 29 Sarrazin, F. et al. (1994)High survival estimates know? Auk 104,544-545 of Griffon vultures (Gyps fuuluus fufous) in a What timing for reintroduction 9 Lubchenco, J. et al. (1991) The Sustainable reintroduced population, Auk 111,853-862 studies? Biosphere Initiative: an ecological research 30 Kleiman, D.G. et al. (1991) Costs of a It is quite clear that the timescale in- agenda, Ecology 72,371-412 re-introduction and criteria for success: volved in these different approaches varies 10 SOI& M.E. (1986) Conservation Biology. accounting and accountability in the Golden strongly (Box 1). The preparation of re- The Science of Scarcity and Diversity, Sinauer Lion Tamarin Conservation Program, Symp. 11 Simberloff, D. (1988) The contribution of Zool. Sot. London 62, 125-142 introduction, including a priori knowledge population and community biology to 31 Mace, G.M. (1994) Classifying threatened of reintroduced species and the reintrc- conservation science, Annu. Rev. Ecol. Syst. species: means and ends, Philos. Trans. R. duction success assessment, is a matter of 19,473-511 Sot. London Ser. B 344,91-97 urgency because of the increasing number 12 Caughley, G. (1994) Directions in 32 Wootton, J.T. and Bell, D.A. (1992) of reintroduction projects. The management ConservationBiology, J. Anim. Ecol. 63,215-244 A metapopulation model of the Peregrine of reintroduced populations implies long- 13 Jordan, W.R., III,Gilpin, M.E. and Aber, J.D. falcon in California: viability and term monitoring with direct interactions (1987) Restoration Ecology: A Synthetic management strategies, Ecol. Appl. 2,307-321 between ecologists and managers. The Approach to Ecological Research, Cambridge 33 Ginsberg, J.R. (1994) Captive breeding, theoretical use of reintroduction, as basic University Press reintroduction and the conservation of experiments, must, by definition, not be 14 Bowles, M.L. and Whelan, C.J. (1994) canids, in Creative Conservation: Interactive Restoration ofEndangered Species, Cambridge Management of Wild and Captive Animals constrained by time, although it will University Press (Olney, P.J.S., Mace, G.M. and Feistner, A.T.C., eventually lead back to microconservation 15 Cade, T.J. and Temple, S.A. (1995) eds), pp. 365-383, Chapman &Hall biology42 by allowing a better knowledge Management of threatened bird species: au 34 Toone, W.D. and Wallace, M.P. (1994) The of metapopulation functioning and extinc- evaluation of the hands-on approach, Ibis extinction in the wild and reintroduction of tion mechanisms and finally limiting the 137(Suppl. l), 161-172 the California condor (Gymnogyps need for reintroductions. 16 Reading, R.P. and Kellert, S.R. (1993) Attitudes californianus), in Creative Conservation., Restoration must remain the main aim toward a proposed reintroduction of Interactive Management of Wild and Captive of reintroductions. Nevertheless, reintro- Black-Footed ferrets (Mustela nigripes), Animals (Olney, P.J.S., Mace, G.M. and ductions offer a unique opportunity for Conserv. Biol. 7,569-580 Feistner, A.T.C., eds), pp. 411-419, experimental studies on ecological pro- 17 Lindburg, D.G. (1992) Are wildlife Chapman &Hall reintroductions worth the cost? Zoo Bio. 11, 35 Miller, R. et al. (1994) Reintroduction of the cesses. It behoves us as ecologists and l-2 Black-footed ferret (Mustela nigripes), in conservationists not to let this opportu- 18 May, R. (1991) The role of ecological theory Creative Conservation: Interactive Management nity pass. in planning re-introduction of endangered of Wild and Captive Animals (Olney, P.J.S., species, Symp. Zool. Sot. London 62,145-163 Mace, G.M. and Feistner, A.T.C., eds), Acknowledgements 19 Salwasser, H., Shonewald-Cox, C. and Baker, R. pp. 455-464, Chapman &Hall We are very grateful to (1987) The role of interagency cooperation in 36 Caswell, H. (1989) Matrti Population Models, Thierry Boulinier, Denis Couvet, managing for viable population, in Viable Sinauer Etienne Danchin, David C. Houston, Populalion for Conservation (Soule, M.E., ed.), 37 Stanley-Price, M.R. (1991) A review of Jane Lecomte and Robert Wielgus for pp. 159-173, Cambridge University Press mammals re-introductions, and the role of helpful comments on the manuscript. 20 Ebenhard, T. (1995) Conservation breeding as the Re-introduction Specialist Group of Torbjiirn Ebenhard and two anonymous a tool for saving animal species from lUCN/SC, Symp. Zool. Sot. London 62, referees provided useful suggestions. extinction, Trends Ecol. Evol. 10,438-443 145-163

TREE vol. II, no. II November 1996 477 PERSPECTIVES

38 Seal, U.S., Foose, T.J. and Ellis, S. (1994) Systematics, Ecology and the Biodiversity Crisis Grtffon vulture Gyps tidvus population, Ibis Conservation Assessment and Management (Eldredge, N., ed.), pp. 121-143, Colombia 138,315-325 Plans (CAMPS)and Global Captive Action University Press 47 Massot, M. et al. (1994) incumbent advantage Plans (GCAPS),in Creative Conservation: 43 Morton,E.S. (1987)Rehttroductfon as a method in common lizards and their colonizing Interactive h4anagement of Wild and Captioe of studying bird behavior and ecology, in ability, J. Anim. Ecol. 63,431-440 Animak (Olney, P.J.S., Mace, GM. and Restoration Ecology: A Synthetic Approach to 48 Reed, J.M. and Dobson, A.P. (1993) Feistner, A.T.C., eds), pp. 312-325, Ecological Research (Jordan, W.R., III,Gilpin, Behavioural constraints and conservation Chapman &Hall M.E. and Aber, J.D., eds), pp. 166-172, biology: conspeciftc attraction and 39 Bradshaw, A.D. (1987) Restoration: an acid Cambridge University Press recruitment, Trends Ecol. Euol. 8,253-255 test for ecology, in Restoration Ecology: 44 Komdeur, J. (1992) importance of habitat 49 Rosenzweig, M.L. (1987) Restoration ecology: A Synthetic Approach to Ecological Research saturation and territory quality for evolution a tool to study population interactions? in (Jordan W.R., III,Gilpin, M.E. and Aber, J.D., of cooperative breeding in the Seychelles Restoration Ecology; A Synthetic Approach to eds), pp. 23-29, Cambridge University Press warbler, Nature 358,493-495 Ecological Research (Jordan W.R., III, 40 Smith, F.D.M.et al. (1993) How much do we 45 Ebenhard, T. (1991) Colonisation in Gilpin, M.E. and Aber, J.D., eds), pp. 189-203, know about the current extinction rate? metapopulations: a review of theory and Cambridge University Press Trends Ecol. Evol. 8,375-378 observations, in Metapopulation Dynamics: 41 Tracy, C.R. and George, T.L. (1992) On tbe Empirical and Theoretical Investigations Reference added in proof determinant of extinction, Am. Nat. 139, (Gilpin, M. and Hanski, I., eds), pp. 105-121, 50 Saltz, D. and Rubenstein, D.I. (1995) 102-122 Academic Press Population dynamics of a reintroduced 42 Barrowclough, F. (1992) Systematics, 46 Sarrazin, F. et al. (1996) Breeding biology . Asiatic wild ass Equus hemionus herd, Ecol. biodivemity and conservation biology, in during establishment of a reintroduced Appl. 5,237-335

evolutionary and population genetics. The lack of reciprocal exchange between the Sewall Wright meets Artificial Life: fields results from the different questions each is trying to address. These differ- the origin and maintenance ences in central questions lead the fields to differ in the emphasis placed upon the of evolutionary novelty evolutionary forces that consume or generate genetic and phenotypic variations. Finally, we discuss the conceptual conver- Yukihiko Toquenaga gence between some A-Life models and the evolutionary genetic theories of Michael J. Wade Sewall Wright, and indicate some unex- plored avenues for two-way interactions A decade ago, Langton coined the term Artificial Life (A-Life) to identify the between genetic algorithms and popu- new field of research that is attempting to create and characterize open-ended lation genetics. evolving systems using diverse computer-based methods. Fruitful interactions between A-Life research and that of conventional biological sciences (B-Life) Evolutionary aud population are rare. Using the framework of molecular and evolutionary genetics, genetics we discuss some of the reasons for this lack of conceptual cross-pollination The goal of population genetics theory between the disciplines and we identify some potential areas for is to characterize and to discriminate interdisciplinary collaboration. among the several evolutionary forces that operate simultaneously in natural Yukihiko Toquenaga is at the Institute of Biological Sciences, populations. The mathematical tools of Universityof Tsukuba, Japan ([email protected]); the theory vary from simple linear algebra MichaelWade is at the Dept of Ecology to statistical and stochastic diffusion mod- and Evolution, University of Chicago, els. A major focus is the balance among 1103E. 57th Street, Chicago,IL 60637,USA ([email protected]). evolutionary forces that maintains genetic variation in natural populations. The sig- nature of natural selection is detected by c hat is life?’ and ‘what could life be?’ engineering. Modern texts in evolutionary comparison with the predictions of null are central questions in the inter- genetics rarely cite any A-Life pubii- models where only the neutral forces disciplinary field of Artificial Life (A-Life). cation+ and some biologists have con- (mutation and random drift) are operat- It tries to subsume all of the conventional sidered these genetic algorithms as coming. Two extreme views coexist within domain of the biological sciences (B-Life) puter exercises in pseudo-genetics: ad hoc formal evolutionary theory’. and it can make fundamental contribu- treatments with a fragmentary theoretical tions to our understanding of B-Life pro- basis of B-life genetics. Computer scien- The fisherian worldview cessesrI*. However, a decade after the tists, on the other hand, have found few if R.A. Fisher proposed that adaptation birth of A-Life, there are few reciprocal any useful theorems in evolutionary gen- typically takes place in large and ran- contributions between the A-Life and B- etics that improve the adaptive perfor- domly mating population+s17~a. Random Life sciences. These two sciences con- mance of genetic algorithms or that per- drift is negligible in this context and mu- verge most conspicuously in the devel- mit the solution of complex problems (cf. tation and selection are sufficient to ex- opment of genetic algorithms, which are Ref. 6). plain patterns of genetic variation. Fisher’s computer-based models for creating open- We discuss the general structure and fundamental theorem of natural selection ended evolving systems in A-Life in ad- findings of genetic algorithms (see defini- (FFf’)s predicts that the rate of adaptive dition to solving complex problems in tion in Box 1) in the context of classical evolution of a population is determined

478 0 1996, Elsevier Science Ltd PII: SOlSS-5347(96)20075-8 TREE uol. I I, no. I I Nouember 1996