- Email: [email protected]
Genetic control of migratory behaviour in birds
TREE vol. 6, no. 8, August
34 Read, A.F. (19881 Trends Ecol. Evol. 3, 97-102 35 Park, T. (1948) Ecol. Monogr. 18, 267-307 36 Price, P.W., Westoby, M. and Rice, B. 119881 Am. Nat. 131, 544-555
37 Freeland, W.1. ( 1983) Am. Nat I2 I, 223-236 38 Scudder, G.G.E. (1983) Hydrobiologia 105, 143-154 39 Smith, B.P. 11988) Anno. Rev. Entomol.
1991
33, 487-507 40 Osborne, P. ( 19851 I. Appl. Eco/. 22, 681-691 41 Scott, M.E. and Dobson, A.P. ( 19891 Parasitol. Today 5. I 76- I83
GeneticControlof Migratory Behaviour in Birds PeterBerthold Although it has long been suspected that biannual migration in birds has a direct genetic basis, only in the last decade have details of the inheritance of behavioural traits such as migratory activity and directional preferences been demonstrated. A model has now been developed to estimate how inexperienced first-time migrants manage to reach their unknown winter quarters on the basis of inherited spatiotemporal programs. Furthermore, in obligate partial migrants the decision to migrate or not has been shown to have a strong genetic base. Migratoriness and sedentariness in partial migrants have been shown to have a high potential for rapid evolution. A recent set of results has suggested that novel migratory habits can evolve in less than 25 years. A possible consequence is that environmental changes, including ‘greenhouse’ effects, might considerably alter avian migration systems bg acting on genetic variation for migratory tendencies.
Before genetic control mechanisms for migration in birds had been demonstrated, ‘endogenous’ factors affecting long-distance intercontinental migration had been postulated. These factors had been documented in the 1970s and were termed ‘circannual rhythms’ or ‘internal calendars’l,2. About ten years ago it became possible to trace these endogenous factors back to their genetic base. One prerequisite for these investigations was the establishment of a system for breeding free-living migratory birds on a large scale in captivity. This was accomplished for a European warbler, the blackcap
Peter Berthold is at the Max-Planck-lnstitut haltensphysiologie, Vogelwarte Radolfzell, Schloss Miiggingen, Germany.
254
fiir VerD-7760
(Sylvia atricapi/fa)3. This species has an extraordinarily wide range of migratory habits (Fig. I I and is thus especially suitable for long-term programs aimed at elucidating the genotype-environment interactions that occur during migration. Migratory behaviour in nocturnal migrants such as the blackcap can easily be measured quantitatively as nocturnal migratory restlessness or zugunruhe*. Heritability of basic behavioural traits A migratory species has to know when to leave the breeding grounds, where to go, when to end the journey and how to organize the trip. Although all these essential features should be regulated and affected by both external and internal factors, a direct genetic control has now been demonstrated for all of them. The urge to migrate With the exception of young cuckoos (Cuculus canorus), which never meet their parents, the first departure of all obligate migrants could theoretically be triggered by experienced conspecifics. In singly migrating species, however, it has generally been hypothesized that endogenous factors are crucial and that the urge to migrate is innate. However, the actual genetic basis and heritability of the urge to migrate have only recently been demonstrated experimentally. When blackcaps from a fully migratory population (from Central Europe) and a resident population (from the Cape Verde Islands) were crossbred, 40% of the F, hybrids were found to be migratory4. Thus, the urge to migrate can be bred into a nonmigratory bird population. Similarly, the urge to migrate can be
enhanced or reduced by selectivebreeding experiments in partially migratory populations. There are, however, limits to this inheritance. The fact that in the cross-breeding experiment not all hybrids became migratory indicates that it is not a single locus that determines the urge to migrate. It is more likely to be a multilocus system with a reaction threshold4. Endogenous time programs and migratory distance Many migrants are thought to migrate singly to distant wintering areas, even though in such species winter quarters are known to be well defined (and hence effectively predetermined). Thus, inexperienced individuals of such species must somehow have ‘knowledge’ of these areas. Several hypotheses have been proposed to account for the means of reaching unknown winter quarters. Innate knowledge of specific local factors, such as star patterns, endogenous distance and energy, have been suggested, as have endogenous distance and time programs2. A substantial amount of evidence has supported the idea that endogenous time programs play a major role5. The most important finding was that the amount of migratory activity displayed by caged individuals of various species and populations was correlated with the distance needed to be covered. Again, cross-breeding experiments with birds of populations with different migratory habits have shown that the amount of migratory activity can be inherited. When blackcaps from Central Europe (long-distance migrants) were cross-bred with conspecifics from the Canary Islands (short-distance migrants) or with
TREE vol. 6, no. 8, August
1991
conspecifics from the Cape Verde Islands (nonmigrants), the offspring showed intermediate amounts of migratory activityb”. Thus, migrants fat least in this species) appear to be equipped with inheritable population-specific time programs for migration that enabled them to cover the distance between breeding grounds and wintering areas without any previous experience (the vector-navigation hypothesis2). Programmed migratory directions It is a common belief among scientists working on orientation that many singly migrating species have innate information about their speciesor population-specific migratory direction7. Ringing recoveries have clearly demonstrated that individuals of the same population regularly move along specific geographic sector9. Captive individuals, tested in orientation cages, often show directional preferences in agreement with free-living conspecifics2. recent cross-breeding Two studies have now demonstrated that the migratory direction can indeed be inherited and thus preprogrammed. When resident blackcaps (from the Cape Verde Islands) were cross-bred with exclusively migratory individuals (from Central Europe) those F, hybrids that showed migratory activity also preferred the directional axis (northeast-southwest) of their migratory parents4. Similarly, in the same species, individuals crossbred from both sides of the Central European migratory divide, which separates populations migrating southeast from those migrating southwest, yielded offspring with phenotypically intermediate behaviour”. Many species change their migratory direction during the course of migration, for example those that migrate from Central Europe travel first in a southwesterly direction to the lberian peninsular but then shift to a southern direction in order to reach winter quarters in Africa and to avoid the Atlantic ocean. There is strong evidence that such directional shifts can also be preprogrammed genetically since they also occur in captive individuals under controlled experimental conditionslO” i
Otfier migratory traits Cross-breeding experiments with blackcapsr2 have also shown, as expected, that morphological features essential for migration, such as wing length and body weight, are genetically determined. The same holds true for aspects of juvenile development that are adaptive to migration - above all, juvenile moult’2. This has also been shown in the stonechat (Saxicola torquata)13. Finally, there is strong suggestive evidence that seasonal changes in feeding rates and food and habitat preferences, as well as activity patterns during staging en route to wintering grounds, are to a considerable extent regulated by inheritable programsr2. Heritability in obligate partial migration There are two types of so-called partial migrants. In one, part of a population occasionally migrates and in the other, part regularly migrates. In the former ffacultative partial migrants), individuals migrate in some years but not others and thus perform irregular or ‘irruptive’ movements. Tits, waxwings and nutcrackers (Nucifraga) are wellknown examples of this category2. These movements are most likely to be controlled by external factors, particularly food shortage and population density7 (and the accompanying dominance effects14, and to some extent weather factors, which will not be discussed further here). Obligate partial migration, on the other hand, is characterized by regular, annual migration in certain individuals of a population. This occurs in many species in Central Europe, e.g. robins (Erithacus rubecufa 1, blackbirds ( Turdus merula) and starlings (Sturnus
vulgaris)7. For many years there has been theoretical discussionr5-ry about whether endogenous or even genetic or environmental factors play a crucial role in the decision to migrate. Recently, experimental evidence has become available for several species20 that sheds light on basic mechanisms involved in the control of obligate partial migration.
Genetic cotitrol of decisiorrmakitig Selective breeding studies on the blackcap2’ and the robin22 have yielded fairly strong selection re-
Fig. I. Migratory habits of the blackcap in its distributional area from the Cape Verde Islands in the southwest through northern Africa and Europe to central Siberia. Thick lines: schematic demonstration of the main migration routes, thin lines: by-routes, dotted line: southern border of the continental breeding area. M: area of exclusively migratory populations, P: area of partially migratory populations, R?, R!: areas of presumed and certain resident populations, respectively.
From
Ref. 23.
sponses when migratory individuals from partially migratory populations and resident individuals were bred in aviaries (Fig. 2). Heritability estimates were rather high2?, indicating that genetic influences are important for the control of this bimodal pattern in behaviour. Since migratory and nonmigratory parents do not exclusively produce migratory or nonmigratory offspring, a single-locus determination of the two behavioural traits is unlikely. Apparently, in the two species tested, both traits are threshold characters that are determined at multiple loci4r23. This means that decision making is most likely not exclusively based on genetic factors but rather that it includes environmental input. However, in obligate partial migrants, genetic influences may be dominant over environmental factors. This view is supported by a number of facts from both field and laboratory studies. A population study with song sparrows (Melospiza melodia), based on long-term genealogiesr6, also demonstrated predominant genetic contro120, as did results of other field studies on blackbirds24 and possibly those on stonechats25. Handraised blackbirds from free-living 255
TREE vol. 6, no. 8, August
Selection
for
Fig. 2. Results of a two-way selective breeding experiment with partially migratory blackcaps from southern France. Nonmigrants were bred up to the F,generation, migrants up to the F, generation. The numbers indicate how many individuals have been bred in the different generations, the broken lines represent the best-fit mathematical functions to the selection response. From
Ref
used, which was composed of about three quarters migratory and one quarter resident individuals. The selection of migrants resulted in exclusively migratory behaviour after just three generations and that for nonmigrants in a largely sedentary habit after four to six generations (Fig. 2). Both behavioural traits thus have an extremely high evolutionary potential and adaptive nature in response to strong selection pressures.
28.
Current evolution and inheritance of novel migratory habits migratory or nonmigratory parents showed mainly the parental traits when they were subsequently tested in captivity25. Changes from the sedentary to the migratory habit have not been detected in individuals tested over several years in captivityi and they appear also to be extremely rare in the wild26. Changes in the opposite direction, which commonly occur in many species, mainly depend on age and may thus be a matter of maturation processes, which could also be heritable12. In the blackbird, there is strong evidence that neither competition and stress in autumn nor the pituitary-adrenal axis are proxinvolved in inducing imately departure26. Thus, the control of obligate partial migration appears to be primarily based on direct genetic control mechanisms. Nonetheless, the extent to which weakly determined genotypes might be influenced in their decision making by environmental factors remains open. Here, detailed field studies have been restarted to answer this question2’. Evolutionary
potential of Gekaviaural
traits The strong selection responses obtained in the F, generations of selectively bred migrants and nonmigrants in blackcaps and robins indicated the potential for a rapid alteration rate in the migratory behaviour of an obligate partially This was migratory population. tested in a two-way selectivebreeding experiment with blackcaps, which continued over several generations28. For this purpose, an obligate partially migratory population from southern France was
The blackcap has developed a novel migratory habit over the last 25 years. Individuals of Central European populations no longer migrate exclusively to the south, towards Mediterranean and African winter quarters. Increasing numbers have begun to fly northwest, to new wintering grounds in the British Isles29-3i. A key factor allowing for this evolutionary process appears to be a general improvement of the nutritional resources in England during the winter, due to the widespread use of feeders, which became much more numerous after the Second World War. These feeders are regularly utilized by blackcaps30. Selection of this novel migratory habit may be accelerated by increasing intraspecific competition in the Mediterranean due to population increases. There are a number of advantages to the novel habit, including the possibility of earlier departure from the wintering area, premature gonadal development, and perhaps greater chances for reproduction and hence increased fitness32,33. The key factor for the control of this microevolutionary process could be the inheritance of the novel migratory direction. This is presently being tested by breeding blackcaps that have been trapped during winter in Britain and examining the directional preference of their offspring.
Avianmigration and ‘greenhouse’effects Global warming and possible increased primary production over large areas during the next century34.35 would be most likely to favour resident species as a result of reduced winter mortality and increased reproductive outpuP.
1991
If obligate partial migration has a strong genetic basis and an extremely high evolutionary potential, birds in this category would easily be capable of increasing their resident proportions adaptively and becoming more or less nonmigratory. Such an overall increase in resident individuals would be to the detriment of exclusively migratory species due to increased resource competition, which has been demonstrated both statistically37 and experimentally2. This increase in residents could eventually lead to a strong decline in bird migration systems and a depletion of species diversity36, a situation that would severely restrict birds’ capacity to cope with future environmental changes. At present, the resident fraction of many obligate partially migratory populations of the temperate zone may well be about to increase. In this case, when migrants ‘are making the best of a bad job’28, partial migration may not only be to some extent a conditional strategyX8 but may in the long run have to be regarded as a transitional stage.
Conclusion Recent results indicate that regularannual bird migration, including obligate partial migration, is to a considerable extent the expression of inherited genetic programs. It is likely that other mechanisms controlling aspects of migration such as fat deposition, staging, and habitat selection are based on immediate genetic contro12. Since breeding migratory birds in captivity is now a well-established approach to studying genetic mechanisms, we can expect good progress in this field in the near future. However, once the genetic basis of migration has been elucidated, a much more difficult task will still remain: to determine the genotype-environment interactions of migration processes which surely differ greatly among themselves as well as among species, populations and most likely also in sex and age groups. The knowledge of their quantitative relationships will be crucial for a more complete understanding of the control of avian migration and for predictions of future microevolutionary processes in migratory species
TREE vol. 6, no. 8, August
1991
Acknowledgements My work on bird migration is supported the Deutsche Forschungsgemeinschaft.
IO Gwinner,
by
References I Gwinner, E. ( 1986) Circannoal Rhythms, Springer-Verlag 2 Berthold. P. ( 1990) Vogelzug, Wissenschaftliche Buchgesellschaft 3 Berthold, P., Quemer, U. and Schlenker, ( 1990) Die MBnchsgrasmiicke Sylvia atricapilla, Die Neue Brehm-Biicherei, Ziemsen 4 Berthold, P., Wiltschko, W., Miltenberger, H. and Querner, U. (1990) Experientia 46, 107-108 5 Gwinner, E. I I971 ) in Biochronometry (Menaker, M., ed.), pp. 405-427, National Academy of Sciences 6 Berthold, P. and Querner, U. (1981)
Vogelzug-Verlag A. 1. I I99 I 1 Behav.
Ecol.
R.
Sociobiol.
Adam Smith proposed that individuals can share sufficient interest in their society’s welfare to profit by cooperating for its benefit and by jointly suppressing behavior hurtful to it. Simibrly, a genome’s genes share a common interest in ‘honest’ meiosis, which ensures that alleles can spread only by Genefitting their carriers. HoneyGee workers express a common interest in raising their queen’s, rather than their half-sisters’, eggs 6y eating eggs laid 6y half sisters. Can analogous principles explain the evolution of harmony af other levels of biological organization, such as ecosystems or organismic development?
Several authors have considered the issue of the evolution of harmony at various levels of biological and have discussed organization’.2, the obstacles that confront the evolution of such harmony2,3. In this article, I review some of these obstacles, and consider some of the mechanisms by which they might be circumvented or overcome.
Egbert Leigh is at the Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Panama. (Mailing address: Smithsonian Tropical Research Institute, APO Miami, FL 14002-001 I, USA.) 0 1991 Elsevw
Physiol.
I. Ornithol.
Science 2 i 2. 77-79 7 Schiiz, E.. ed. (1971) Crundrifl der Vogelzugskunde, Parey 8 Zink, G. ( 1973-851 DerZug Europlischer
Singvdgel, 9 Helbig, 28,9-I2
E. and Wiltschko, W. (1978) 125, 267-273 II Helbig, A.I., Berthold, P. and Wiltschko, W. ( 19891 Ethology 82, 307-3 I5 I2 Berthold, P. II9901 in Bird Migration: The Physiology and Ecophysiology (Cwinner, E., ed.), pp. 269-280, Springer-Verlag I3 Gwinner, E. and Neusser, V. (1985)
1. Comp.
126, 2 19-220
I4 Smith, H. G. and Nilsson, 104, 109-I I5 15 Miller, A. H. (1931) Univ. Publ.
I-A. (19871 Auk Calif..
Berkeley,
Zoo/.
38, I l-242 M. M. (1933)
16 Nice, 1. Ornithol. 81, 552-595 I7 Lack. D. (1943-44) Brit. Birds 37. 122-130, 143-150 I8 Gauthreaux, S. A. II9781 in Perspectives in Ethology (Bateson, P. P. and Klopfer, P. H., eds), pp. 17-24, Plenum Press 19 Lundberg, P. ( 1988) Trends Ecol. Evol. 3, 172-175 20 Berthold, P. ( 1984) Ring IO, 253-265 21 Berthold, P. and Ouerner, U. (1982) Experientia
38, 805
22 Biebach, 23 Berthold, 195-209 24 Schwabl,
H. (1984) Auk 100, 601-606 P. I 1988) /. Evol. No/. I, H. II9831
1. Ornithol.
124.
101-l I6 25 Dhondt, 155-158 26 Schwabl,
A. A. ( 19831 Ringing H. and
Silverin.
Migration: The Physiology Ecophysiology (Gwinner,
Migr.
B. (19901
4,
in Bird
and
E., ed.1, pp. 144-l 55, Springer-Verlag 27 Adriaensen, F. and Dhondt, A. A. (I9901 1. Anim. Ecol. 59, 1077-l 090 28 Berthold, P., Mohr, G. and Querner, U. I I 990) /. Ornithol. I 3 I, 33-45 29 Langslow. D. R. (19791 Bird Study 26, 239-252 30 Leach, 1. H. (19811 BirdStudy28, 5-14 31 Simms. E. ( 1985) British Warblers, Collins 32 Berthold, P. and Terrill, S B. (I9881 Ringing Migr. 9, 153-l 59 33 Terrill, S. B. and Berthold, P. (19901 Oecologia
79, 266-270
34 Dobson. II9891
Trends
A., lolly, Ecol.
A. and Rubenstein, Evol.
D.
4, 64-68
35 Boer, M. M., Koster, E. A. and Lundberg, H. (1990) Ambio 19, 2-10 36 Berthold. P. ( 19901 Verh. Dtsc,b. Zoo/. Ges. 83. 227-244 37 O’Connor, R. 1. (I981 ) in Animal Migration (Aidley, D. I., ed.), pp, 167-195, Cambridge University Press
Genes,BeesandEcosystems: The Evolutionof a CommonInterest amongIndividuals Egbert Giles Leigh,Jr Reciprocal altruism Let us illustrate the ‘tragedy of the commons’4 - the conflict between individual advantage and the common good - by a pair of conspecific individuals that interact repeatedly. In each interaction, an individual may cheat or cooperate. Cheating minimizes each individual’s potential loss and maximizes its potential gain, yet cooperation is in their common interest, as it maximizes the average gain per interaction for both individuals. How can they enforce their common interest? The best strategy is ‘tit for tat’: cooperate at first encounter, then act as the other did the last time5. Hamlets (Hypoplectrus spp.) are simultaneously hermaphroditic coral reef fish that employ ‘tit for tat’ to avoid competing for mates. Successful sperm contribute as many genes to future generations as successful eggs, so selection usually favors dividing reproductive effort
Science Publishers Ltd iilK 0169~5347,91.$02 00
equally between male and female function&. Organisms with separate sexes usually raise as many sons as daughters; simultaneous hermaphrodites spend equal effort producing and fertilizing eggs6. Males contributing only genes to their young compete for mates by excessive sperm production, expensive displays, or potentially destructive fighting. Redirecting this surplus effort could nearly double the number or quality of young: the effort devoted to male functions represents the ‘50% cost of sex16, ‘Tit for tat’ enables hamlets to avoid this cost7. Fish form pairs, and repeatedly trade eggs for each other to fertilize (Fig. I). Fish that fail to reciprocate are abandoned, and must take the time to find new mates. Cooperative enforcement of group welfare In larger groups, ‘tit for tat’ is less likely to enforce cooperation, 257
34 Read, A.F. (19881 Trends Ecol. Evol. 3, 97-102 35 Park, T. (1948) Ecol. Monogr. 18, 267-307 36 Price, P.W., Westoby, M. and Rice, B. 119881 Am. Nat. 131, 544-555
37 Freeland, W.1. ( 1983) Am. Nat I2 I, 223-236 38 Scudder, G.G.E. (1983) Hydrobiologia 105, 143-154 39 Smith, B.P. 11988) Anno. Rev. Entomol.
1991
33, 487-507 40 Osborne, P. ( 19851 I. Appl. Eco/. 22, 681-691 41 Scott, M.E. and Dobson, A.P. ( 19891 Parasitol. Today 5. I 76- I83
GeneticControlof Migratory Behaviour in Birds PeterBerthold Although it has long been suspected that biannual migration in birds has a direct genetic basis, only in the last decade have details of the inheritance of behavioural traits such as migratory activity and directional preferences been demonstrated. A model has now been developed to estimate how inexperienced first-time migrants manage to reach their unknown winter quarters on the basis of inherited spatiotemporal programs. Furthermore, in obligate partial migrants the decision to migrate or not has been shown to have a strong genetic base. Migratoriness and sedentariness in partial migrants have been shown to have a high potential for rapid evolution. A recent set of results has suggested that novel migratory habits can evolve in less than 25 years. A possible consequence is that environmental changes, including ‘greenhouse’ effects, might considerably alter avian migration systems bg acting on genetic variation for migratory tendencies.
Before genetic control mechanisms for migration in birds had been demonstrated, ‘endogenous’ factors affecting long-distance intercontinental migration had been postulated. These factors had been documented in the 1970s and were termed ‘circannual rhythms’ or ‘internal calendars’l,2. About ten years ago it became possible to trace these endogenous factors back to their genetic base. One prerequisite for these investigations was the establishment of a system for breeding free-living migratory birds on a large scale in captivity. This was accomplished for a European warbler, the blackcap
Peter Berthold is at the Max-Planck-lnstitut haltensphysiologie, Vogelwarte Radolfzell, Schloss Miiggingen, Germany.
254
fiir VerD-7760
(Sylvia atricapi/fa)3. This species has an extraordinarily wide range of migratory habits (Fig. I I and is thus especially suitable for long-term programs aimed at elucidating the genotype-environment interactions that occur during migration. Migratory behaviour in nocturnal migrants such as the blackcap can easily be measured quantitatively as nocturnal migratory restlessness or zugunruhe*. Heritability of basic behavioural traits A migratory species has to know when to leave the breeding grounds, where to go, when to end the journey and how to organize the trip. Although all these essential features should be regulated and affected by both external and internal factors, a direct genetic control has now been demonstrated for all of them. The urge to migrate With the exception of young cuckoos (Cuculus canorus), which never meet their parents, the first departure of all obligate migrants could theoretically be triggered by experienced conspecifics. In singly migrating species, however, it has generally been hypothesized that endogenous factors are crucial and that the urge to migrate is innate. However, the actual genetic basis and heritability of the urge to migrate have only recently been demonstrated experimentally. When blackcaps from a fully migratory population (from Central Europe) and a resident population (from the Cape Verde Islands) were crossbred, 40% of the F, hybrids were found to be migratory4. Thus, the urge to migrate can be bred into a nonmigratory bird population. Similarly, the urge to migrate can be
enhanced or reduced by selectivebreeding experiments in partially migratory populations. There are, however, limits to this inheritance. The fact that in the cross-breeding experiment not all hybrids became migratory indicates that it is not a single locus that determines the urge to migrate. It is more likely to be a multilocus system with a reaction threshold4. Endogenous time programs and migratory distance Many migrants are thought to migrate singly to distant wintering areas, even though in such species winter quarters are known to be well defined (and hence effectively predetermined). Thus, inexperienced individuals of such species must somehow have ‘knowledge’ of these areas. Several hypotheses have been proposed to account for the means of reaching unknown winter quarters. Innate knowledge of specific local factors, such as star patterns, endogenous distance and energy, have been suggested, as have endogenous distance and time programs2. A substantial amount of evidence has supported the idea that endogenous time programs play a major role5. The most important finding was that the amount of migratory activity displayed by caged individuals of various species and populations was correlated with the distance needed to be covered. Again, cross-breeding experiments with birds of populations with different migratory habits have shown that the amount of migratory activity can be inherited. When blackcaps from Central Europe (long-distance migrants) were cross-bred with conspecifics from the Canary Islands (short-distance migrants) or with
TREE vol. 6, no. 8, August
1991
conspecifics from the Cape Verde Islands (nonmigrants), the offspring showed intermediate amounts of migratory activityb”. Thus, migrants fat least in this species) appear to be equipped with inheritable population-specific time programs for migration that enabled them to cover the distance between breeding grounds and wintering areas without any previous experience (the vector-navigation hypothesis2). Programmed migratory directions It is a common belief among scientists working on orientation that many singly migrating species have innate information about their speciesor population-specific migratory direction7. Ringing recoveries have clearly demonstrated that individuals of the same population regularly move along specific geographic sector9. Captive individuals, tested in orientation cages, often show directional preferences in agreement with free-living conspecifics2. recent cross-breeding Two studies have now demonstrated that the migratory direction can indeed be inherited and thus preprogrammed. When resident blackcaps (from the Cape Verde Islands) were cross-bred with exclusively migratory individuals (from Central Europe) those F, hybrids that showed migratory activity also preferred the directional axis (northeast-southwest) of their migratory parents4. Similarly, in the same species, individuals crossbred from both sides of the Central European migratory divide, which separates populations migrating southeast from those migrating southwest, yielded offspring with phenotypically intermediate behaviour”. Many species change their migratory direction during the course of migration, for example those that migrate from Central Europe travel first in a southwesterly direction to the lberian peninsular but then shift to a southern direction in order to reach winter quarters in Africa and to avoid the Atlantic ocean. There is strong evidence that such directional shifts can also be preprogrammed genetically since they also occur in captive individuals under controlled experimental conditionslO” i
Otfier migratory traits Cross-breeding experiments with blackcapsr2 have also shown, as expected, that morphological features essential for migration, such as wing length and body weight, are genetically determined. The same holds true for aspects of juvenile development that are adaptive to migration - above all, juvenile moult’2. This has also been shown in the stonechat (Saxicola torquata)13. Finally, there is strong suggestive evidence that seasonal changes in feeding rates and food and habitat preferences, as well as activity patterns during staging en route to wintering grounds, are to a considerable extent regulated by inheritable programsr2. Heritability in obligate partial migration There are two types of so-called partial migrants. In one, part of a population occasionally migrates and in the other, part regularly migrates. In the former ffacultative partial migrants), individuals migrate in some years but not others and thus perform irregular or ‘irruptive’ movements. Tits, waxwings and nutcrackers (Nucifraga) are wellknown examples of this category2. These movements are most likely to be controlled by external factors, particularly food shortage and population density7 (and the accompanying dominance effects14, and to some extent weather factors, which will not be discussed further here). Obligate partial migration, on the other hand, is characterized by regular, annual migration in certain individuals of a population. This occurs in many species in Central Europe, e.g. robins (Erithacus rubecufa 1, blackbirds ( Turdus merula) and starlings (Sturnus
vulgaris)7. For many years there has been theoretical discussionr5-ry about whether endogenous or even genetic or environmental factors play a crucial role in the decision to migrate. Recently, experimental evidence has become available for several species20 that sheds light on basic mechanisms involved in the control of obligate partial migration.
Genetic cotitrol of decisiorrmakitig Selective breeding studies on the blackcap2’ and the robin22 have yielded fairly strong selection re-
Fig. I. Migratory habits of the blackcap in its distributional area from the Cape Verde Islands in the southwest through northern Africa and Europe to central Siberia. Thick lines: schematic demonstration of the main migration routes, thin lines: by-routes, dotted line: southern border of the continental breeding area. M: area of exclusively migratory populations, P: area of partially migratory populations, R?, R!: areas of presumed and certain resident populations, respectively.
From
Ref. 23.
sponses when migratory individuals from partially migratory populations and resident individuals were bred in aviaries (Fig. 2). Heritability estimates were rather high2?, indicating that genetic influences are important for the control of this bimodal pattern in behaviour. Since migratory and nonmigratory parents do not exclusively produce migratory or nonmigratory offspring, a single-locus determination of the two behavioural traits is unlikely. Apparently, in the two species tested, both traits are threshold characters that are determined at multiple loci4r23. This means that decision making is most likely not exclusively based on genetic factors but rather that it includes environmental input. However, in obligate partial migrants, genetic influences may be dominant over environmental factors. This view is supported by a number of facts from both field and laboratory studies. A population study with song sparrows (Melospiza melodia), based on long-term genealogiesr6, also demonstrated predominant genetic contro120, as did results of other field studies on blackbirds24 and possibly those on stonechats25. Handraised blackbirds from free-living 255
TREE vol. 6, no. 8, August
Selection
for
Fig. 2. Results of a two-way selective breeding experiment with partially migratory blackcaps from southern France. Nonmigrants were bred up to the F,generation, migrants up to the F, generation. The numbers indicate how many individuals have been bred in the different generations, the broken lines represent the best-fit mathematical functions to the selection response. From
Ref
used, which was composed of about three quarters migratory and one quarter resident individuals. The selection of migrants resulted in exclusively migratory behaviour after just three generations and that for nonmigrants in a largely sedentary habit after four to six generations (Fig. 2). Both behavioural traits thus have an extremely high evolutionary potential and adaptive nature in response to strong selection pressures.
28.
Current evolution and inheritance of novel migratory habits migratory or nonmigratory parents showed mainly the parental traits when they were subsequently tested in captivity25. Changes from the sedentary to the migratory habit have not been detected in individuals tested over several years in captivityi and they appear also to be extremely rare in the wild26. Changes in the opposite direction, which commonly occur in many species, mainly depend on age and may thus be a matter of maturation processes, which could also be heritable12. In the blackbird, there is strong evidence that neither competition and stress in autumn nor the pituitary-adrenal axis are proxinvolved in inducing imately departure26. Thus, the control of obligate partial migration appears to be primarily based on direct genetic control mechanisms. Nonetheless, the extent to which weakly determined genotypes might be influenced in their decision making by environmental factors remains open. Here, detailed field studies have been restarted to answer this question2’. Evolutionary
potential of Gekaviaural
traits The strong selection responses obtained in the F, generations of selectively bred migrants and nonmigrants in blackcaps and robins indicated the potential for a rapid alteration rate in the migratory behaviour of an obligate partially This was migratory population. tested in a two-way selectivebreeding experiment with blackcaps, which continued over several generations28. For this purpose, an obligate partially migratory population from southern France was
The blackcap has developed a novel migratory habit over the last 25 years. Individuals of Central European populations no longer migrate exclusively to the south, towards Mediterranean and African winter quarters. Increasing numbers have begun to fly northwest, to new wintering grounds in the British Isles29-3i. A key factor allowing for this evolutionary process appears to be a general improvement of the nutritional resources in England during the winter, due to the widespread use of feeders, which became much more numerous after the Second World War. These feeders are regularly utilized by blackcaps30. Selection of this novel migratory habit may be accelerated by increasing intraspecific competition in the Mediterranean due to population increases. There are a number of advantages to the novel habit, including the possibility of earlier departure from the wintering area, premature gonadal development, and perhaps greater chances for reproduction and hence increased fitness32,33. The key factor for the control of this microevolutionary process could be the inheritance of the novel migratory direction. This is presently being tested by breeding blackcaps that have been trapped during winter in Britain and examining the directional preference of their offspring.
Avianmigration and ‘greenhouse’effects Global warming and possible increased primary production over large areas during the next century34.35 would be most likely to favour resident species as a result of reduced winter mortality and increased reproductive outpuP.
1991
If obligate partial migration has a strong genetic basis and an extremely high evolutionary potential, birds in this category would easily be capable of increasing their resident proportions adaptively and becoming more or less nonmigratory. Such an overall increase in resident individuals would be to the detriment of exclusively migratory species due to increased resource competition, which has been demonstrated both statistically37 and experimentally2. This increase in residents could eventually lead to a strong decline in bird migration systems and a depletion of species diversity36, a situation that would severely restrict birds’ capacity to cope with future environmental changes. At present, the resident fraction of many obligate partially migratory populations of the temperate zone may well be about to increase. In this case, when migrants ‘are making the best of a bad job’28, partial migration may not only be to some extent a conditional strategyX8 but may in the long run have to be regarded as a transitional stage.
Conclusion Recent results indicate that regularannual bird migration, including obligate partial migration, is to a considerable extent the expression of inherited genetic programs. It is likely that other mechanisms controlling aspects of migration such as fat deposition, staging, and habitat selection are based on immediate genetic contro12. Since breeding migratory birds in captivity is now a well-established approach to studying genetic mechanisms, we can expect good progress in this field in the near future. However, once the genetic basis of migration has been elucidated, a much more difficult task will still remain: to determine the genotype-environment interactions of migration processes which surely differ greatly among themselves as well as among species, populations and most likely also in sex and age groups. The knowledge of their quantitative relationships will be crucial for a more complete understanding of the control of avian migration and for predictions of future microevolutionary processes in migratory species
TREE vol. 6, no. 8, August
1991
Acknowledgements My work on bird migration is supported the Deutsche Forschungsgemeinschaft.
IO Gwinner,
by
References I Gwinner, E. ( 1986) Circannoal Rhythms, Springer-Verlag 2 Berthold. P. ( 1990) Vogelzug, Wissenschaftliche Buchgesellschaft 3 Berthold, P., Quemer, U. and Schlenker, ( 1990) Die MBnchsgrasmiicke Sylvia atricapilla, Die Neue Brehm-Biicherei, Ziemsen 4 Berthold, P., Wiltschko, W., Miltenberger, H. and Querner, U. (1990) Experientia 46, 107-108 5 Gwinner, E. I I971 ) in Biochronometry (Menaker, M., ed.), pp. 405-427, National Academy of Sciences 6 Berthold, P. and Querner, U. (1981)
Vogelzug-Verlag A. 1. I I99 I 1 Behav.
Ecol.
R.
Sociobiol.
Adam Smith proposed that individuals can share sufficient interest in their society’s welfare to profit by cooperating for its benefit and by jointly suppressing behavior hurtful to it. Simibrly, a genome’s genes share a common interest in ‘honest’ meiosis, which ensures that alleles can spread only by Genefitting their carriers. HoneyGee workers express a common interest in raising their queen’s, rather than their half-sisters’, eggs 6y eating eggs laid 6y half sisters. Can analogous principles explain the evolution of harmony af other levels of biological organization, such as ecosystems or organismic development?
Several authors have considered the issue of the evolution of harmony at various levels of biological and have discussed organization’.2, the obstacles that confront the evolution of such harmony2,3. In this article, I review some of these obstacles, and consider some of the mechanisms by which they might be circumvented or overcome.
Egbert Leigh is at the Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Panama. (Mailing address: Smithsonian Tropical Research Institute, APO Miami, FL 14002-001 I, USA.) 0 1991 Elsevw
Physiol.
I. Ornithol.
Science 2 i 2. 77-79 7 Schiiz, E.. ed. (1971) Crundrifl der Vogelzugskunde, Parey 8 Zink, G. ( 1973-851 DerZug Europlischer
Singvdgel, 9 Helbig, 28,9-I2
E. and Wiltschko, W. (1978) 125, 267-273 II Helbig, A.I., Berthold, P. and Wiltschko, W. ( 19891 Ethology 82, 307-3 I5 I2 Berthold, P. II9901 in Bird Migration: The Physiology and Ecophysiology (Cwinner, E., ed.), pp. 269-280, Springer-Verlag I3 Gwinner, E. and Neusser, V. (1985)
1. Comp.
126, 2 19-220
I4 Smith, H. G. and Nilsson, 104, 109-I I5 15 Miller, A. H. (1931) Univ. Publ.
I-A. (19871 Auk Calif..
Berkeley,
Zoo/.
38, I l-242 M. M. (1933)
16 Nice, 1. Ornithol. 81, 552-595 I7 Lack. D. (1943-44) Brit. Birds 37. 122-130, 143-150 I8 Gauthreaux, S. A. II9781 in Perspectives in Ethology (Bateson, P. P. and Klopfer, P. H., eds), pp. 17-24, Plenum Press 19 Lundberg, P. ( 1988) Trends Ecol. Evol. 3, 172-175 20 Berthold, P. ( 1984) Ring IO, 253-265 21 Berthold, P. and Ouerner, U. (1982) Experientia
38, 805
22 Biebach, 23 Berthold, 195-209 24 Schwabl,
H. (1984) Auk 100, 601-606 P. I 1988) /. Evol. No/. I, H. II9831
1. Ornithol.
124.
101-l I6 25 Dhondt, 155-158 26 Schwabl,
A. A. ( 19831 Ringing H. and
Silverin.
Migration: The Physiology Ecophysiology (Gwinner,
Migr.
B. (19901
4,
in Bird
and
E., ed.1, pp. 144-l 55, Springer-Verlag 27 Adriaensen, F. and Dhondt, A. A. (I9901 1. Anim. Ecol. 59, 1077-l 090 28 Berthold, P., Mohr, G. and Querner, U. I I 990) /. Ornithol. I 3 I, 33-45 29 Langslow. D. R. (19791 Bird Study 26, 239-252 30 Leach, 1. H. (19811 BirdStudy28, 5-14 31 Simms. E. ( 1985) British Warblers, Collins 32 Berthold, P. and Terrill, S B. (I9881 Ringing Migr. 9, 153-l 59 33 Terrill, S. B. and Berthold, P. (19901 Oecologia
79, 266-270
34 Dobson. II9891
Trends
A., lolly, Ecol.
A. and Rubenstein, Evol.
D.
4, 64-68
35 Boer, M. M., Koster, E. A. and Lundberg, H. (1990) Ambio 19, 2-10 36 Berthold. P. ( 19901 Verh. Dtsc,b. Zoo/. Ges. 83. 227-244 37 O’Connor, R. 1. (I981 ) in Animal Migration (Aidley, D. I., ed.), pp, 167-195, Cambridge University Press
Genes,BeesandEcosystems: The Evolutionof a CommonInterest amongIndividuals Egbert Giles Leigh,Jr Reciprocal altruism Let us illustrate the ‘tragedy of the commons’4 - the conflict between individual advantage and the common good - by a pair of conspecific individuals that interact repeatedly. In each interaction, an individual may cheat or cooperate. Cheating minimizes each individual’s potential loss and maximizes its potential gain, yet cooperation is in their common interest, as it maximizes the average gain per interaction for both individuals. How can they enforce their common interest? The best strategy is ‘tit for tat’: cooperate at first encounter, then act as the other did the last time5. Hamlets (Hypoplectrus spp.) are simultaneously hermaphroditic coral reef fish that employ ‘tit for tat’ to avoid competing for mates. Successful sperm contribute as many genes to future generations as successful eggs, so selection usually favors dividing reproductive effort
Science Publishers Ltd iilK 0169~5347,91.$02 00
equally between male and female function&. Organisms with separate sexes usually raise as many sons as daughters; simultaneous hermaphrodites spend equal effort producing and fertilizing eggs6. Males contributing only genes to their young compete for mates by excessive sperm production, expensive displays, or potentially destructive fighting. Redirecting this surplus effort could nearly double the number or quality of young: the effort devoted to male functions represents the ‘50% cost of sex16, ‘Tit for tat’ enables hamlets to avoid this cost7. Fish form pairs, and repeatedly trade eggs for each other to fertilize (Fig. I). Fish that fail to reciprocate are abandoned, and must take the time to find new mates. Cooperative enforcement of group welfare In larger groups, ‘tit for tat’ is less likely to enforce cooperation, 257