Ex situ preservation involves the conservation of plants or animals in a situation removed from their normal habitat. It is used to refer to the collection and freezing in liquid nitrogen of animal genetic resources in the form of living semen, ova or embryos. It may also be the preservation of DNA segments in frozen blood or other tissues. Finally it may refer to captive breeding of wild plants or animals in zoos or other situations far removed from their indigenous environment.
In situ conservation is the maintenance of live populations of animals in their adaptive environment or as close to it as is practically possible. For domestic species the conservation of live animals is normally taken to be synonymous with in situ conservation.
Ex situ and in situ conservation are not mutually exclusive. Frozen animal genetic resources or captive live zoo populations can play an important role in the support of in situ programmes. The relative advantages and disadvantages of the major systems are therefore reviewed here with a view to identifying the relative strengths and areas of mutual support.
|Ex Situ||In Situ|
|i.||COST - initial set up cost||rel high||low-high|
- maintenance cost
|ii.||GENETIC DRIFT - initial||rel high||low|
|iii.||Applied to all species||no||yes|
|vi.||International access||good||not good|
|ix.||Selection for use||none||good|
The relative cost of collecting, freezing and storing frozen material, as compared to maintaining large scale live populations, has been estimated to be very low (Smith, 1983). In particular, once the material has been collected, the cost of maintaining a cryogenic store is minimal. Such banks require little space and few trained technicians. A very large number of frozen animals from a large number of populations can be stored in a single facility.
Cryogenically preserved populations suffer no genetic loss due to selection or drift. The method places a sample in suspended animation and that sample remains genetically identical from the time of collection to the time of use. (The effects of long term radiation are considered to be negligible.)
Frozen animal genetic resources can be made available to livestock breeding and research programmes throughout the world.
The principal disadvantages of ex situ, or cryogenic preservation lie in the availability of the necessary technology and access to the frozen populations.
Cryogenic stores are not expensive to run but they do have annual capital maintenance requirements. In particular they require a guaranteed supply of liquid nitrogen which must be imported into many countries with expensive foreign currency or aid.
Cryogenic stores have no intrinsic value with respect to financial income unless material can be sold for research and development. They do not produce food or other agricultural commodities and might therefore be deemed to be expensive luxuries in periods of financial austerity.
Cryogenic storage is ideal for the preservation of defined ‘genes’ or recognized characteristics. Quite small samples ensure the inclusion of all but the rarest genes (see Table 7, section 4.3.1). However, the cryogenic method is less effective in the conservation of ‘breeds’ where the relative frequency of genes is important.
The methods of initially sampling and collecting genetic material from a limited number of animals to be incorporated into cryogenic storage can result in an initial genetic drift. Thus there is a shift in gene frequencies between the original population and the cryogenically conserved sample population.
The technology necessary for semen collection and freezing, and for superovulation, ova and embryo flushing and freezing is readily transferred throughout the world, however, it is expensive for countries in which the technology is not yet established. The technologies are not yet developed for all species, viable pig and poultry embryos, for example, cannot currently be successfully thawed after freezing. There are also instances where the technologies may be developed but the livestock themselves are not accessible for semen or embryo collection, for reasons of politics, ownership or their remote location.
There is a potential danger in cryogenic storage, from large scale loss of material due to serious accidents. This could be due to human error, power failure, loss of liquid nitrogen, fire, flood, storm, earthquake or war. Such risks can be reduced by keeping duplicate stores in different regions but remains a serious concern.
Linked to the danger of global loss of cryogenic material due to accident, is the danger that regions or nations might lose access to the material. This could be due to their failure to develop or maintain the technological ability to access the frozen stores. There is also the fear of political change which might affect the rights of access to global or regional banks.
Cryogenically preserved populations cannot be studied characterized or monitored. They are not easily available for comparative trials or for research projects. It takes a number of years to regenerate a cryogenically preserved population to review or re-evaluate it in changed circumstances, or to utilize it as a breeding population.
Cryogenically preserved populations are not able to adapt through gradual selection, to changes in the climate or disease background of the local or global environment.
Finally animal disease control in the future could make the use of frozen material laid down in a relatively disease prevalent period, too dangerous to use. This is already a problem with European semen banks which have been collected under a number of different health regulations and testing regimes over the past fifty years. The majority of these stores will be deemed unsuitable for use in Europe under new European Community animal health directives to be implemented in 1992 (National Cattle Breeders Association, 1991).
The major advantages for in situ conservation relate to the availability of technologies and the utilization of the breeds.
The in situ conservation of live populations requires no advanced technology. There are optimal sampling strategies (see section 4.3.1) and breeding strategies (see section 4.4), but the basic needs of an in situ programme are already available and affordable throghout the world. The farmers of every region and nation know how to manage and maintain their local strains. They already have the capability, all they require is direction.
In situ projects can ensure that financial commitment to the conservation of animal genetic resources involves helping to improve the livelihood of farming communities associated with the breeds targeted for conservation. Live conservation projects involve animal utilization and are net producers of food, fibre and draught power (see table 6). They do not require the importation of expensive materials, skills or equipment.
Live conservation programmes may survive major political or environmental upheaval, wars, or climatic disasters that could eliminate frozen stores, especially those needing imported frozen nitrogen. Sufficient numbers of breeding units must be established and maintained, however, for each conserved population.
In situ projects enable breeds to be properly characterized and evaluated in their own and related localities. They allow for comparative trials, research and crossing experiments.
This method of conservation also allows populations to adapt to changing environmental conditions and endemic diseases.
The maintenance of live herds allows for selection and improvement of populations within the sustainable constraints which will be discussed later (see section 4.3.2).
The disadvantages of in situ conservation are brought about by a lack of complete control over the many factors which influence the survival of individuals and therefore the genetic makeup of the conserved population.
In situ conservation projects require land and people which are limited resources in some regions of the world. Continuation of all conservation projects is dependent upon unpredictable financial and political change particularly if they are government or institutionally run. They do have the capacity to produce agricultural commodities and sell livestock to supplement their budgets (see Table 6).
Genetic drift is an inevitable feature of all live animal conservation projects, even when steps are taken to minimize the problem. Selection and the resultant shift in the gene frequencies within a population are a real possibility, and may even be a legitimate objective of some programmes. Selection is a particular concern when it is applied to populations being maintained under modified environmental conditions and should only be made within locally sustainable conditions (see section 4.3.2).
In situ conservation incurs the possible threat of disease eliminating whole, or substantial parts, of a conserved population, particularly if the conserved herd is in a single or only a few linked locations. Diseases may also act as a major selection pressure within a population, and may substantially change its characteristics.
Finally, live animal conservation programmes do not assist in the easy international transfer of animal genetic resources as compared to the movement of frozen material. Moving live animals is relatively more expensive and there are international restrictions on the movement of animals to control disease.
Cryogenic methods allow for animal genetic resource material to be suspended, unchanged, for long periods of time. Live conservation efforts enable breeds to be properly evaluated, monitored and used in the present changing agro-economic climate as well as being available for future farmers and livestock breeders. The two strategies are not mutually exclusive and should be considered as complimentary strategies which may be easily and beneficially linked.
Economic Production and Recreational Uses Arising From Live In situ Conservation Projects
|Direct production of food||×||×||×||×||×|
|Dams for crossing for meat||×||×||×|
|Production of furs||×||×|
|Production of wool/fibres||×||×||×|
|Use in prison, school and hospital farms||×||×||×||×||×|
|Pasture and lawn management||×||×||×|
|Utilization of harsh environments||×||×||×|
|Utilization of marginal areas||×||×||×||×||×|
|Sera for research||×||×||×||×||×|
|Non-allergic milk production||×||×|
|Veterinary or medical research||×||×||×||×||×|
|Education Sport and Leisure|
|Aid to education||×||×||×||×||×|
|Sport and leisure||×|
|(after Maijala, 1986)|
Collecting and freezing of semen is far simpler in most species than collecting and freezing of embryos. Recent development in the technology to mature ova from the ovaries of slaughtered females has produced a relatively cheap and easy method for the collection of haploid cells from females to parallel the collection of sperm. It is likely that this technique will become increasingly useful as the methods become more widely available.
Maintenance of semen alone does not allow for the recreation of a pure breed, but only for ‘breeding back’ through a crossing programme but, by using a large sample of semen from different males alongside a relatively small population of live females it is possible to maintain an entire population of animals. The use of artificial insemination (AI) in situ conservation of populations enables a much larger number of males to be used in the breeding programme than would be practical if they all had to be maintained as live adult males. This automatically increases the effective population size (Ne) and reduces the minimum total number of live animals that must be maintained to produce an acceptable level of inbreeding and genetic drift (Smith, 1977). This strategy could be particularly useful in species or varieties where the technology for collecting and freezing of ova or embryos is not well developed or available, for example with pigs and poultry; or for endangered species where no alternative host for preserved embryos could be used as is the case with the Indian elephants for example.
Use and replenishment of frozen semen collections alongside a live population will enable breeds to adapt to gradual and permanent changes to the environment. It will allow for the changes necessary to respond to a background of disease and parasites which will gradually mutate over time. It will also allow for current and accurate data to be collected and maintained on breeds. Over time disease control, nutritional knowledge and veterinary care will improve. It is therefore important that breeds are monitored in these changed circumstances and are not continually judged by their production characteristics as measured in less developed situations in the increasingly distant past.
The conservation of endangered species, breeds or populations is an attempt to maintain genetic resources in an identifiable and potentially usable form.
Endangered species must be conserved in separate species units because it is not possible to out cross or pool different species. Endangered breeds of the common species, however, may be maintained as separate ‘breeds’ or may be combined or pooled into groups of breeds or composites for the purposes of conservation.
The advantage of conserving a distinct separate ‘breed’ is that it has a defined set of characteristics and parameters. Its appearance, behaviour, production, and native environment should all be known or can be determined. A breed represents a group of animals with a known range of genetic variation with predictable and characteristic effects. Such a population can be screened for undocumented characteristics in the future and desirable genes can be accessed through conventional breeding techniques or genetic engineering.
The disadvantage of conserving ‘breeds’ separately, is that there are a very large number of them, and that many have very similar characteristics.
The conservation of genetic variation in a gene pool or breed composite requires less resources than individual breed conservation. Many breeds may be combined into a gene pool of a size considerably smaller than that required for separate programmes. There are, however, a number of serious disadvantages with gene pool conservation;
A well described breed is by definition predictable in its appearance and production, while a gene pool or composite population is not predictable in the expression of those characteristics.
The identification of animals carrying a specific gene within a composite may be impossible to determine because expression of the gene may become masked by alternative alleles found in the other breeds in the composite. This may be the case even if the presence of the gene in the pool is known, because it was a feature of one of the breeds included in the original breed mixture.
Valued genetic characteristics may be caused by the interaction of a number of genes always found within one breed. In a gene pool such complementary genes may be separated resulting in the disappearance of the valued characteristic within the composite. This could be the case with some forms of parasite resistance for example, where a physiological adaptation might be linked to dietary preferences or behavioural characteristics. Thus there may be unpredictable genetic interactions between breeds resulting in the disappearance of expected characteristics or the appearance of new unexpected ones.
It is generally believed that a composite conservation population may be considerably smaller than the total size of separate breed programmes for each of the breeds maintained independently. Such a strategy may result in a larger number of the pre-existing genes being lost over time through drift.
Gene pools or composites can be used effectively to conserve genes that affect obvious morphological features which can be easily identified. For example, colour or extreme quantitative traits such as the prolific Boroola gene associated with litter size in sheep. However, such individual traits can be equally well preserved in many species, by cryogenic techniques.
Pooling of breeds should never be considered until the separate breeds or identified populations have been properly characterized. The gene pool is not a useful strategy for the conservation of very varied populations. It may be used for the conservation and selection of a number of closely related breeds with economically important traits, whose physiology and adaptive characteristics are similar. For example, there are four recognized breeds of goat in the desert region of North Eastern Brazil. Each comes from a slightly different area and are distinguishable by their colour patterns. The Moxoto are light cream with black points, the Rapartida are cream with dark forequarters, the Caninde are black with a yellow belly and the Morota or Curaca are solid white in colour. However, beside the name and colour differences initial research suggests that their environment, size, growth rate, production and survivability are very similar (Mariante, 1991). If this is confirmed a more effective conservation, selection and improvement programme could be developed by pooling the breeds rather than maintaining separate strains.
When gene pools are deemed to be beneficial no more than two or three populations should be combined in order to keep the frequency of most the alleles at a useful level.
Before sampling can begin clear objectives must be defined with respect to the objective of the programme. In particular consideration must be given to whether the programme is to conserve unique genes within the population or the breed itself (see section 3.2).
a. Sample Size
As a general rule the larger the sample or founder group, the greater the range of genetic variation that will be incorporated into the conservation programme.
Where the conservation herd is to act as a nucleus which will interact with other farm or village herds, the sample may not be finite. In this case exchange of genetic material will be possible between herds in the future. In breeds where the conservation herd is likely to be all that will survive of the breed, it is essential that as many founders as possible are included. In this case no more diversity can be maintained than is included in the initial sample.
Relatively few unrelated individuals can represent a considerable genetic diversity. The chance that a population sample of size N, will not contain a gene whose frequency in the population is p, may be expressed as (1-p) to the power of 2N.
Thus in a reasonable size sample there is a good chance that all the available genes will be included unless they were at a very low frequency in the original breeding population. A sample of 25 males and 50 females is recommended as a minimum for a live conservation programme. This has been calculated to result in a loss of less than 1% of the possible genetic variation present in the original population (Smith, 1983).
It has also been shown that even quite small founder groups of less than 10 females can survive and produce viable living populations. It is possible to ensure the survival of almost all the genetic variation present in such small founder groups with a carefully planned breeding strategy (see 4.4). In such cases it is advisable to avoid future or frequent bottle necks in the population size, as this will inevitably result in a dramatic reduction in genetic variation and possible extinction. The optimal strategy for conservation is then to increase the population size as rapidly as possible.
|Gene Frequency||Number of sires at a probability of|
The number of sires that must be sampled to reduce the probability of an allele with the frequency P being excluded from a sample semen store to below 0.01 or 0.001 (after Notter and Foose, 1987).
Once a population has reached its holding size, if one is to be imposed, it is important to design a programme to minimize selection, inbreeding and drift so as to maximize the survival of the genetic diversity found within that population and its chance of survival along the same lines as for the conservation of any small population (see section 4.4).
b. Statistical Sampling Techniques
Methods of sampling fall into three major categories; random, stratified and maximum avoidance techniques.
A random sample is one in which every animal has an equal chance of being selected for the sample as every other animal. By definition the sampler has no control over the specific animals selected and can make no judgements about typical or atypical animals being included or excluded from the sample. Relatively small samples collected in this way from populations in which there is a lot of genetic diversity, may result in a shift in gene frequency between the initial population and the sample population, due to chance.
By dividing a population into strata, or groups of similar animals and then sampling the various groups at random, it is possible to ensure a reduced shift in genetic makeup between the sample and the original population.
In situations where the pedigree of the animals within the population is known, it is possible to create a sample which represents the largest possible number of ancestral or founder animals. In this case animals will be selected for inclusion in the sample because they share no common ancestors. No common ancestors is normally taken to mean no common ancestors as parents, grandparents or great grandparents.
c. Practical Sampling Techniques
The principal objective in sampling a population to create a conservation unit, is to attempt to include as much of the genetic variation inherent within that population as possible. Thus animals should be selected from throughout the breeds normal geographical range and should incorporate all the characteristics normally associated with the breed. Furthermore, when breeding records are available, closely related animals should be avoided in order to make room for unrelated individuals.
The major problems associated with sampling a breed in order to establish any kind of conservation programme are: availability of suitable stock due to disease, political restrictions and ownership; conformity to breed description; and degree of dilution by other breeds.
Not all groups of animals that constitute a breed or identified population are equally available for sampling for inclusion in a conservation programme.
Some geographical regions or individual herds may have endemic diseases not found on the conservation farms or co-operating regions. Some herd owners may not be willing or able to participate in supplying animals to a conservation programme. Finally part of a population may be situated on the other side of a political boundary making access difficult or even impossible. In each of these cases the subpopulations excluded from the programme should be very carefully considered with respect to their potential contribution to the programme. If possible it should be determined how different the subpopulation is to the available one by careful comparison of: the environment, which may be more or less extreme; visible phenotypic differences; simple physiological differences; such as blood types or milk proteins; or differences between the structure of the chromosomes or DNA fragments. Any of these studies might suggest that the subpopulation represents a unique and important subsection of the population. If this is not the case the subpopulation may be omitted from the conservation sample. However, if they are deemed to be important, action may need to be taken to establish disease control or parallel conservation efforts.
A pure conservation scientist will look at a conservation programme in terms of maintaining the maximum genetic variation possible within the conservation flock or herd. This conflicts to some extent with the value of conserving a breed with its predictable and known set of characteristics in a known environment. Sampling, for a programme whose objective is ‘breed’ conservation, should reflect this breed description and samples should be taken from within the known parameters of the breed. Individuals exhibiting extreme characteristics might be included in a frozen store with the appropriate information in the associated database. Animals on the fringe of the breed parameters will often be the result of cross breeding with exotic or introduced breeds in the relatively recent past.
There can be serious practical problems associated with historically well defined breeds which have been extensively diluted. In many cases a limited number of animals may remain which exhibit some of the characteristics which were typical of the old breed. In this case it may be possible and desirable to define parameters for the conservation group, increase the population size as rapidly as possible, and then select from within the group to eliminate the ‘foreign influence’ and re-establish a new population which exhibits as much as possible of the original breed characteristics. For small populations selection should be imposed on males only, to reduce the risk of inbreeding. This strategy has been employed with the Cikta sheep in Hungary (Bodo, 1984) and similar programmes have been developed for the Navajo-Churro in the USA (McNeale, 1970).
It is often said that within the conservation of small populations no selection pressure should be imposed because it would reduce the levels of genetic diversity intrinsic within the population. In practice this is an impossible restriction to place on any conservation programme. Selection at some level will inevitably occur in all live conservation programmes and is essential in order to maintain the characteristics of the population.
Wild species are maintained by natural selection and it is recognized that in order for these species to have a good long term survival chance natural selection must be allowed to continue to act upon them. One of the problems of conserving wild species in captivity is that there will be drift in the genetic makeup of the population due to the lifting of natural selection pressures. For example albinoism is a rare recessive gene found in many populations of mammals. Because the gene is rare in the population, individuals exhibiting the albino characteristics are extremely rare. They also stand a much lower chance of surviving into adulthood and reproducing because they are much more likely to be predated than their camouflaged fellows. This fact ensures that the gene remains rare. In a captive situation where there are no predators an albino individual is far more likely to survive and to reproduce, thereby passing its albino genes onto the next generation. The frequency of the albino gene will therefore increase within the population and consequently so will the chance of albino individuals appearing in future generations. This is a very obvious and visible example of genetic drift due to the lifting of natural selection pressure, but it demonstrates the subtle shifts in genetic characteristics which may occur in populations protected from normal selection pressures.
The selection pressures imposed by the environment and by man that have created domestic breeds and populations are equally important and are discussed in chapter 2 of this manual.
In the case of naturally selected characteristics, it is just as important for domestic populations as it is for wild species to be able to continue to exist and adapt within their normal environment. For example, breeds adapted to climatic extremes, unusual diets or heavy parasite or disease infestation should continue to exist under this selection pressure. Animal welfare issues should clearly be taken into account. Allowing natural selection to work does not imply non-intervention and does not require that non-adapted animals be left to slowly die. It does provide the opportunity for the selective culling of those individuals which are clearly not functioning as well as would be expected in the breed's normal environment.
Selection within breeds by man is not fundamentally different from the approach to natural selection described above. In projects established to conserve unique genes or characteristics the appearance of atypical animals may be considered beneficial. If a population has been developed for its own particular production characteristics, for example its wool, milk or draught power, and the objective is ‘breed’ conservation, the population should be conserved with these breed specific characteristics. Thus a limited amount of selection should be an integral part of breed conservation. This selection should be targeted at maintaining the known characteristics and parameters of the breed. It should not be used to reduce the genetic diversity found within a breed being conserved in a small population, but rather to limit the effects of individual outstanding or unusual animals and to prevent traits previously alien to the breed becoming common. The objective should be to conserve an identified group of animals with known parameters. This should not be conservation of just a colour pattern or horn shape, or the conservation of a breed name attached to a herd which has long ago lost the characteristics for which the original breed was valued.
The important general features of such selection are:
Selection should not be carried out in very small populations where inbreeding may be a problem. The population should first be allowed to increase in size (see 3.3.4).
Most livestock breeding programmes involve the use of more females than males. Selection may be imposed on males whilst maintaining the influence of as many founders as possible through the unselected females.
Selection should be carried out within the adaptive environment and should be against characteristics which prevent the animal from functioning well in that environment, or from exhibiting the production characteristics typical of the breed.
Selection for or against features relatively common within a breed should be considered very carefully. Advantageous characteristics may be positively selected in the context of conservation, but this should be done in larger programmes as described in section 4.6 of this manual.
Selection against so called ‘undesirable’ characteristics common to a breed should only be carried out once the real affects and interactions of these characteristics are known. Congenital splitting of the upper eyelid in multihorned breeds of sheep, for example, has been shown to be closely linked to the genes causing the development of the impressive four horns. Selection against the eyelid condition, which has no selective disadvantage in the natural environment of these sheep, resulted in a dramatic reduction in the frequency of four horned individuals within the UK population (Henson, 1981).
Similarly selection was carried out among the seaweed eating sheep of the Orkney Islands to remove monorchid ram lambs which were found to be very common. This selection was done without first identifying if the condition was historically common, why it was so prevalent within the population and if it was linked in any way to the breeds remarkable ability to survive on a diet of the seaweed laminaria. This breed is considered in more detail in 5.4.2.
A similar type of selection has been reported in the Ethiopian programme to supply drought resistant bulls of the indigenous cattle to the devastated regions of Eritrea. These cattle are the only animals that will survive in the region and the project is an excellent one. However, one of the characteristics used to select bulls for redistribution was ‘a straight back’ (Relief Society of Tigray, 1986). Although a feature used to select European cattle it is not a characteristic known to be linked to the ability to survive in extreme drought conditions. Selection, if it is to exist in conservation herds must, therefore, be justifiable with respect to the important and locally valued features of the breed.
Inbreeding is the mating of closely related animals and results in an increase in homozygosity. It reduces the amount of genetic variation within the population as compared to an outbred population. The chance that closely related animals carry the same mis-copied, non-functional or deleterious pieces of DNA inherited from a common ancestor is quite high. Inbreeding will tend to result in more homozygous animals which have inherited two copies of the less efficient gene. For this reason the general affects of continuous inbreeding are seen as a reduction in fertility and viability particularly with respect to survival after birth and growth rate to weaning (Falconer, 1981; Lasley, 1978; Warwick, 1979; Wright, 1977).
Conversely mild inbreeding combined with intensive selection can be used to improve livestock breeds so that superior animals with more effective genetic characteristics have more influence over future generations than inferior ones. In many important developed breeds more than 80 or 90% of the population can trace their pedigrees back to one or two superior individual animals. This selection combined with low level continuous inbreeding concentrates the desired genes and enables deleterious genetic characteristics to be selected out. This method of continuous low level inbreeding and elimination of deleterious genes has resulted in domestic populations that can withstand much higher levels of inbreeding than wild species which are not normally exposed to inbreeding pressures (Frankel & Soule, 1981).
In small populations inbreeding can be controlled by careful planning of the breeding strategy but it remains a function of small population size. In situations of very small populations where close inbreeding is the only option it is better to mate brothers and sisters than parents to offspring. The inbreeding coefficient is the same in both cases but sib matings help to equalize the genetic contribution to the next generation from the two parent lines. In this situation the objective is to carry as much genetic variation from the founder group into the next generation as is possible (see section 4.4).
Within small populations inbreeding is a function of the population size, because the chance that any two individuals mated together will be related, or share common ancestors, is increased. If a strategy of random breeding is assumed within a population, it is possible to estimate the rate of inbreeding, which will vary according to the number of breeding animals in the population.
The rate of inbreeding (δF) in a small population is calculated as
ΔF = 1/2Ne
where Ne is the effective population size. The effective population size is affected by the ratio of males to females, longevity, and variance in family size (see appendix 3.2).
In turn, the rate of inbreeding reflects the drift in genetic variation within the population. Genetic drift in a small population is the loss of genetic variation through random chance. There may be a number of different DNA options at one address on the chromosome which code for a number of different possible phenotypic characteristics. For example blue, brown or green eyes. If there is no selection pressure the likelihood that any one option will be passed onto the next generation is affected by random chance. In a very small population this may result in the frequency of the options ‘drifting’, by chance, until they will become fixed at the frequency of one or zero in the population.
The percentage of genetic diversity conserved over time decreases rapidly with smaller population sizes (see fig. 2). Thus the amount of heterozygosity or genetic variation present in each generation begins to decrease at an accelerated rate once the effective population size (Ne) falls below 100.
The level at which a small population conservation programme can be established is determined primarily by the rate of inbreeding, or the rate of loss of genetic diversity, which is considered to be acceptable over a specified period time. An absolute minimum effective population size of 50 is considered necessary for the survival of zoo populations of wild species where breeding strategies can be very closely controlled (Frankel & Soule, 1981).
The ratio of males to females is very important in the calculation of the minimum of animals needed to conserve a population. This is based on the effective population sizes, or the number of animals contributing genetic material into the next generation (see section 3.3.2).
Small effective population sizes result in an increase in inbreeding which results in a loss of heterozygosity or genetic diversity (see section 4.3.3).
Programmes to conserve endangered domestic animals have been proposed that would result in inbreeding rates of between 1 and 4% per generation. It is possible to maintain inbreeding rates at this level, with populations of between 12–25 males and 100–250 females.
It has been estimated (Smith, 1984) that the following minimum number of animals are required for the conservation by management of endangered breeds of the common agricultural species.
These estimates take into account the number of males and females in the breeding unit and the number of young replacement males and females joining the population each year. They may be taken as absolute minimum for the maintenance of a conservation herd and require carefully planned breeding programmes. Calculations of minimum population size and loss of genetic variation are theoretical, although Wright's inbreeding coefficient, central to all the calculations, does not always correlate with the observable affects in particular populations. This observable difference in real populations is due to the actual number of common ancestors in the group and the severity of the particular deleterious genes they happen to carry.
The American Association of Zoological Parks and Aquaria ‘Species Survival Plans’ for endangered species similarly incorporate minimum population sizes with a breeding strategy which requires the replacement of individual animals by their offspring in a controlled and planned way, in order to maximize the use of limited animal spaces in the zoo programme.
|No.s males||No.s females||Total||Ne||% inc in inbreeding/generation assuming random mating|
|(After Brem, FAO, 1989)|
|Size of breeding unit||10||26||22||60||44||44||72||72|
|No. of breeding animals entering each year||10||5||22||12||44||18||72||72|
In these programmes an effective population size of 500 is considered to be an absolute minimum for the long term survival of a conserved population. This may be obtained theoretically with an actual number of only 250 animals if the sex ratio is held constant at one to one and every animal contributes equally to the next generation (Franklin, 1980).
Begin with an adequate sized sample of animals who should ideally be unrelated, non-inbred and fertile. They should represent the range of genetic types found within the population. If possible a sample of at least 50 males and 50 females should be included.
Expand the population as rapidly as possible, to a minimum effective population size of 500 animals (see section 4.3.5).
Standardizing the longevity.
Equalize the representation of the founders, (i.e. the animals in the original sample). It is important that as many of these founder animals as possible are represented in each generation.
Manage inbreeding, in most cases the best strategy is to keep inbreeding to a minimum. There are situations in sublined populations where alternative strategies might be chosen (see appendix 5.4).
Subdividing the population may be a useful option (see section 4.5.5). In particular this strategy may help to control the possible spread of disease between conservation herds.
The three principal methods for establishing a conservation breeding programme for small populations are based upon the methods of natural breeding, random mating and pedigree breeding.
The natural breeding strategy adopted by wildlife conservation programmes involve ensuring that sufficient animals exist in the conservation area to allow normal mating structures to exist. Thus territorial or harem behaviour is allowed to proceed as it would in a normal wild population, such that the strongest and best adapted males mate with the majority of the females. Intervention may take the form of removing older males after one or two breeding seasons to ensure that younger individuals have the opportunity to reproduce. Action may also be taken to transport males from one ‘island’ reserve to another in situations where wildlife sanctuaries are divided by inhospitable or impassable man made obstacles.
This strategy of conservation is also used on the conservation of feral or very extensively managed domestic populations. The conservation of Ossabaw Island hogs on Ossabaw island in the USA (Brisbin, 1985) and the primitive ancestral flock of Soay sheep on the Island of Hirta (Jewell, 1974) are good examples of this method. These strategies require larger minimum population sizes than those discussed in 4.3 above, and should only be considered for populations of many hundred individuals in situations where most of the males will be allowed to remain with the female herds. The ancient wild cattle of Chillingham have survived for seven hundred years with a very limited population size and a natural breeding structure although there is evidence from blood type examination, of considerable homozygosity within the breed (Henson, 1983).
The random mating system is designed to ensure that each adult animal has an equal chance of leaving an equal number of progeny. Statistically the progeny numbers fall into a Poisson distribution. Models can be developed to randomly select mates. This method has been used unsuccessfully for small populations of poultry where 40 pairs of parents per generation has been calculated as acceptable.
In practice, Professor Crawford in Canada has maintained 17 middle level poultry stocks using a random mating strategy for a period of up to 24 generations. The populations are 11 hens, 1 turkey, 1 guinea fowl, 1 duck, 1 Muscovey duck and 2 goose lines. No significant reduction in fertility or hatchability which might be associated with inbreeding, has been observed. All birds are bred at one year of age to minimize the effects of selective mortality.
The principal problems encountered with this system of random mating has been the variation between adults or between families in a number of survival characteristics which include;
All of these variables can quickly skew the distribution and accelerate genetic drift. This method of breed conservation is ideally suited to groups or herds of animals which can be easily identified as a group and where matings can be controlled between groups, but where individual identification of animals is not practical.
By monitoring the pedigree of animals in a conservation breeding programme it is possible to ensure that each animal or family contributes equally to the next generation. Within a population of fixed size, each male can then be sure to contribute one son to the next generation and each female, one daughter.
In wildlife conservation programmes for zoo animals an effective population size of 500 is the target for long term conservation of zoo populations. This can be achieved with an actual population of only 250 animals if the sex ratio is held constant at one to one and each animal contributes equally to the next generation (Franklin, 1980). Each animal produces at least two litters by two different mates. No more than one offspring will be selected for the breeding programme from each litter. Each male will be replaced in the breeding population by one son, each female by one daughter (Foose, 1983). If carried out correctly this strategy confers negligible genetic loss over an extended period of time. The problems are practical ones. Not all animals breed equally well in captivity and some will not mate with their selected partners. There is also a problem with the public's perceptions of conservation which is not compatible with the elimination of individuals born in a litter but not needed for the breeding programme. This is particularly difficult in zoos where the public constitutes a major source of funding. It is a less serious problem in the conservation of agricultural animals where the surplus can be used for human consumption in the normal way.
The basic structure of random mating within pedigree lines has been effectively used for domestic
breeds and in particular for poultry (see table 11).
This system involves considerably more time, more expensive equipment in terms of individual cages and more skilled technicians for artificial insemination than the basic random flock system described in 4.5.2 above. However, it should result in lower inbreeding coefficients and far less genetic drift than the basic random system.
In practice the system has shown lower fertility than the flock system due to technical and practical problems with the use of artificial insemination. It has also brought to light a serious infertility problem in one female line of one of the breeds which was masked in the random flock system (Crawford, 1989).
Pedigree systems are more effective in monitoring inbreeding and therefore loss of genetic variation over long generation intervals. They also supply important relationship information to help monitor genetic diseases or defects that may occur.
The random breeding strategy described above helps to ensure that the chance of variation at any one genetic loci being lost, is kept to a minimum. However, once pedigree records are available the optimal strategy to minimize inbreeding from one generation to the next is through the maximum avoidance method. This involves selecting mates which are the least related to one another. It is particularly effective in maintaining low inbreeding coefficients and thus high levels of genetic variation in very small populations. In such cases it should be used in conjunction with an overall strategy to increase population size as rapidly as possible.
Once groups are larger a more effective method of maintaining maximum genetic variation over a longer period of time, may be to sub-divide the population and use a cyclic breeding system.
When a population is divided into separate sublines each of the sublines will, by definition, be smaller than the original group. Each group will then have a higher rate of inbreeding, and a higher rate of loss of genetic variation due to drift, than the total population group as a whole. However, the random chance that the same genetic variation will be lost in all the sublines is very low. It is, therefore, postulated that a population can successfully be divided into sublines. A maximum avoidance breeding strategy can then be adopted within each line. Periodically there can be an exchange of males, probably in a cyclic rotation between the sublines. In this way any degree of dividing into sublines, will ultimately reduce the final rate of decline of heterozygosity.
It has been postulated that a practical method for achieving this might be to inbreed within subdivided lines for 8 to 10 generations and then outcross between populations. The important feature of all these proposals remains the actual size of the population and therefore the rate of genetic loss, and the level of inbreeding the population can withstand. There are also practical restraints associated with unequal sex ratios and the overlap of generations (see appendix 5.4).
The strategy results in the conservation of a number of inbred lines with very little genetic variation within each one, but assumes that the total variation will survive in the total population of sublines. However, inbred lines are often inferior to non-inbred ones in terms of resistance to disease, reproductive success and lifespan. The strategy has been postulated for zoo populations where the method of maintaining the entire captive population under a minimum inbreeding programme has been compared to alternating mild inbreeding within zoos, combined with outcrossing between zoo populations. The latter strategy has the advantage of less movement of animals which helps in lower costs, animal welfare and the control of disease between the various breeding groups (see appendix 5.4).
Dividing into sublines is not recommended for very small populations where the size of the sublines would be so small that the inbreeding coefficients would rise too quickly and jeopordize the survival of the sublines due to reductions in fertility and viability of offspring.
In practice sub-dividing methods have been used in the conservation of the Hungarian Grey Cattle in Hungary and the Lipizzan Horse in Austria (Bodo & Pataki,. 1984).
The most effective method of conservation is that which involves the full utilization of a breed. There are many instances where breeds have become endangered, but through proper characterization and evaluation a role has been identified. In some cases the new role has been in another country, for example the prolific Finnish Landrace sheep which are now more numerous outside Finland than within the source country. In other cases it has been a regenerated home market resulting from proper research and improvement of a locally adapted breed. For example the Criollo cattle of Bolivia which were being systematically replaced by Zebu crosses. Research into the real production characteristics and survivability of the breed has resulted in a renewed interest in the Criollo which has a long term commercial role in the cattle industry of that country (Wilkins, 1984).
Conservation and utilization programmes require careful and accurate evaluation of breeds in their own local situation as compared to exotic breeds and their crosses in the same sustainable conditions. Where breeds emerge as having a real local potential they should be incorporated into proper evaluation and improvement programmes either on station or in the field situation.
The most important factor of all improvement of conservation groups is that selection is carried out in the environment to which the breed is adapted and that it is for traditionally valued characteristics. These characteristics might be foraging ability, mothering ability, ease of parturition, draught ability which might include stamina, strength, food conservation for draught or tractability, it might be for rich milk for cheese or butter making, or fat meat to supplement a low fat vegetable diet, or for soap making. Western European selection criteria should not be imposed on populations unless they are clearly appropriate. Most important of all, selection must be under the local sustainable conditions. The herds must be managed within the natural environment for that breed and need to be exposed to the conditions prevalent in the field situation. Thus treatment for a disease or parasite to which the breed has some natural resistance should not be given unless the same treatment is freely and practically available in the field. Disease and parasite resistance is biologically expensive. Thus if these heavy selection pressures are lifted, and production pressures imposed the population will very quickly shift such that they will produce more milk or meat but will lose the genetically controlled resistance for which they were valued. The same is true of heat tolerance, drought resistance or the ability to survive on diets with a very low nutritional value.
There are a number of African programmes designed to conserve and improve disease resistant strains, including those with trypanotolerant N'Dama cattle in the Republic of Guinea (Devilliard, 1983). There are also village based projects beginning around the world including the Jamnapari goat project in India (Bhattacharya, 1990) and work with the heartwater resistant Tswana sheep, goats and cattle of Botswana (Setshwaelo, 1989) both of which are discussed in chapter 5 of this manual.
Projects for conservation, utilization and improvement of breeds should attempt to lay down periodic samples of cryogenically stored material as a long term insurance policy. They must clearly define their objective which should incorporate the conservation of those characteristics for which the breed has been traditionally valued. They may then take the form of a number of cooperating farms in a male progeny testing scheme, as with the Sahiwal cattle project in India discussed in chapter 5. More frequently they are associated with open nucleus type breeding strategies.
For a nucleus herd of 200 cows the best 20 females will be selected from the villages each year and some 20 will die or be culled from the nucleus herd. This system will produce a moderate rate of improvement although in practice the village farmers are not often willing to give up their best females to the programme and the nucleus herd does not produce any better than the average for the village herds. This is in part due to the fact that the village farmers don't give up or sell their best cows, but may also be due to a difference in management. It has been found in India that breeds developed over centuries in a very small herd situation with very close interaction between farmer and animal do not take well to a large herd situation with little or no association between individual animals and people. There was therefore an initial reduction in production level from the institutional farm over the average for the village herds.
A larger nucleus of 1,000 trypanotolerant N'Dama cattle has been proposed for the republic of Guinea (Devillard, 1983) which would produce a far greater expected rate of genetic improvement and produce superior males and females for the co-operating farmers as well as surplus stock for sale to farmers within Guinea or for export to neighbouring countries.
The problems lie in the maintenance of such a large herd under sustainable village conditions, and convincing the Fulani herdsmen to sell or lease their best cows to the project. There are also some problems associated with obtaining accurate details of the production in the field situation. For example when monitoring milk production and growth rate of calves, it is essential to know if the cow is also being milked for family milk consumption, and if so, at what level.
There is an argument to use research stations to research into possible management improvements that could be practically used by the farmers, optimal dipping period, feeding regimes, feeding supplements, weaning times and so on. Selection and monitoring of breeds can then be done using national resources and large numbers of co-operating farmers.
Similar farmer co-operative methods are being used to monitor and improve the production of indigenous stocks throughout the world. Some examples may be found in chapter 5.
7. A national conservation and improvement programme for Tswana goats, which are resistant to the tick borne heartwater disease has been established in Botswana.
8. Panteneiro cattle, adapted to the extremely hot and humid conditions of the Pantenal region of Brazil. This breed is now being researched and characterised by the National Agricultural Research Institute in Brazil.