Ian R. Franklin
CSIRO Animal Production, Blacktown, NSW 2148

In 1978, Michael Soulé asked me to consider the genetical dangers faced by species maintained in small numbers, and what minimum population size should be maintained in order to avoid these dangers. Hence, I was concerned with only one aspect of population size and structure, rather than demographic or ecological issues, either of which might demand that endangered species be maintained at larger numbers than is necessary on genetic grounds alone. I concluded that there were two primary issues; an immediate danger of extinction through inbreeding depression and, in the longer term, a loss of variability that might restrict future evolutionary change. My conclusions became embodied in some of the conservation biology literature as the "50/500 rule" and a number of biologists have criticised both the "rule" and its application to conservation biology. I review some of my original conclusions in the light of twenty years of new theory, empirical evidence and much rhetoric.

Monitoring genetic erosion in isolated populations

David S. Woodruff
Department of Biology, University of California, San Diego, La Jolla, CA 92093-0116, USA (dwoodruf@ucsd.edu)

The effects of genetic erosion (GE) on the viability of small populations are understood in theory but the process has gone largely undocumented as the changes were, until recently, difficult to detect. It is now possible to monitor genetic erosion in isolated populations in ecological time by noninvasive genotyping using hypervariable nuclear microsatellite loci as markers of variability. Three studies will be reviewed to illustrate the power of noninvasive monitoring: (1) using museum skins and feathers the endangered San Clemente island loggerhead shrike was found to have lost 20% of its variability in the last 80 years; (2) using live-trapping on island rainforest fragments isolated in a new reservoir in Thailand we were able to document GE in populations of a rat, mouse and tree shrew in years 5-8 post-fragmentation; (3) using dung as a DNA source we are beginning to monitor GE in isolated populations of elephants in parks. These studies show that rates of GE are species-specific, reflecting differences in life history and behaviour; that allelic variation is lost faster than heterozygosity; and that genetic erosion may commence before the onset of obvious demographic decline. Small, recently isolated populations lose variation faster than allowed for by current conservation practice.

Exceptionally low genetic diversity in an ancient relic, the Wollemi Pine: implications for conservation theory and practice

Rod Peakall
Division of Botany and Zoology, Australian National University, Canberra, ACT 0200

The discovery of the Wollemi Pine, Wollemia nobilis in 1994 made international headlines. The relictual conifer belongs to the ancient conifer family the Araucariaceae, and is only known from 2 small populations. I am currently conducting a study of genetic variation in this species along with representatives of the only other extant genera in the family, Agathis and Araucaria. So far, within the Wollemi Pine no allozyme variability has been found at 13 allozyme loci. Furthermore, no variability has been detected at more than 800 loci, visualised by the AFLP method. While the absence of allozyme variability is known for other rare species, the lack of AFLP variation is unexpected, since this method normally reveals polymorphic loci, even when allozyme variation is absent. This suggests exceptionally low genetic diversity in the Wollemi Pine. Long term isolation, small population size and clonality may have contributed to this pattern. Preliminary data in Agathis, while showing some variation, may indicate that genetic variability is low in the family as a whole. The Araucariaceae may thus provide an opportunity to explore the cause and evolutionary consequences of low genetic diversity, both of which remain important empirical issues in conservation biology.

The Valley Pocket Gopher: a system to study fitness correlates and genetic variability in nature

Gerard P. Zegers, M.A. Sanjayan, K.K. Moran, K. Crooks, M.E. Soulé

The Valley Pocket Gopher, a fossorial rodent, occurs in locally differentiated populations, with heterozygosities ranging from 0-0.2. We present comparisons of physiological performance and genetic variability (allozyme electrophoresis, multilocus fingerprints, MHC heterozygosity via heteroduplex analysis) within and between populations.

In juveniles collected from a variable population, genetic variability was positively associated with growth rate and resting O2 consumption and negatively associated with burrowing O2 consumption. After release in a field enclosure, juveniles with lower burrowing and higher resting metabolic rates were significantly more likely to survive until winter.

Gophers from genetically homogeneous populations had lower burrowing efficiencies and lower digestive efficiencies than those from variable populations. Within the two least variable populations reciprocal skin grafts were readily accepted. Grafts between these populations and within genetically more diverse populations were routinely rejected. Background genetic variability was significantly correlated with MHC heterozygosity among 16 populations.

We have measured individual performance in the lab, shown that it translated into differential survivorship in nature and extended this work by comparing performance and genetic correlations between populations. Performance is correlated with markers of genetic diversity readily available to population managers.

Inbreeding depression in endangered species, examples from Gila topminnows, Mexican wolves, and Speke's gazelles

Philip Hedrick and Steven Kalinowski
Arizona State University, Tempe, Arizona 85287, USA

Inbreeding depression is generally considered the most important genetic concern in conservation of endangered species. However, it is often difficult to determine the extent of inbreeding depression in endangered species. First, we discuss our experimental work with Gila topminnows in which we have examined five traits related to fitness for two generations of inbreeding. The most striking finding is a highly skewed, female-biased sex ratio in one of the populations after inbreeding. Next, we discuss how to estimate inbreeding depression for viability from pedigrees in Mexican wolves. Although we did not detect any inbreeding depression, we had low statistical power to detect inbreeding depression because of the structure of the pedigree. Finally, earlier work has suggested that inbreeding depression has been purged from the captive population of Speke's gazelles. Although we also found there was a change in the level of inbreeding depression, other explanations appear more appropriate than purging for this change.

Inbreeding in pedigreed populations of tigers & racehorses

Georgina M. Mace, Sarah Christie, Jonathan Wilcken & Stuart Williams
Institute of Zoology, Regent's Park, London, NW1 4RY, UK

The effect of inbreeding on fertility and survival was investigated in Sumatran tigers (Panthera tigris sumatrae) bred in European zoos, and in a UK data set on thoroughbred racehorses. Parentage has been recorded with high reliability over 7 generations of Sumatran tigers and 30 generations of racehorses. There was also good information in studbook records on both species to allow an analysis of the effects of inbreeding in relation to other factors influencing fitness, such as the age, identity and parity of the mother and the year and place of birth. In addition, the idea that genomes previously exposed to inbreeding might be less susceptible to the effects of inbreeding depression ('purging' of inbreeding depression) was investigated using a coefficient reflecting the inbreeding ancestry of each individual.

In tigers, the main factors contributing to variation in survival were the location of birth, the identity of the dam and the year of birth. There was a small but significant effect of inbreeding on both survival and fecundity. Purging effects were weak and inconsistent. In the racehorses, the month of conception, the year of conception, the age and parity of the dam had the greatest influence on fertility. Inbreeding effects were statistically significant, but accounted for only a small proportion of the overall variation.

Managers concerned about improved rates of survival and fecundity need to view measures of inbreeding depression alongside the impact of other environmental factors that may have a major impact on fitness.

Inbreeding and longevity: the cost of inbreeding revisited

Jonathan Wilcken
Australasian Regional Association of Zoological Parks & Aquaria (ARAZPA) and School of Biological Sciences, Macquarie University

Numerous studies, across a wide range of species, have shown that inbreeding reduces individual fitness. However, estimates of the cost of inbreeding have usually been derived from studies assessing only a single component of fitness, usually juvenile mortality. To date, published studies have not evaluated the effect of inbreeding on lifetime survivorship across species.

This study assesses the effect of inbreeding on lifetime survivorship in captive populations of 27 mammalian and avian taxa. Inbreeding coefficient had a significant impact on lifetime survivorship in 18 out of the 27 taxa, in all cases reducing overall longevity. A further 4 taxa showed a similar, but non-significant trend. In 20 out of 26 taxa, this trend was not confined solely to the early period, but reflected also reduced survivorship beyond 180 days in inbred animals. Further, in half of all cases where inbreeding did not apparently influence early survivorship, a significant reduction in survivorship later on in life was revealed. A majority of the taxa which did not conform with the general trend of increased early mortality were species for which recording neonate mortality can be problematical. In these instances, inbreeding depression may not be easily revealed through studying early mortality, but tends to become more apparent in later age classes.

Overall, increased mortality rates, both in juvenile and adult age classes, act to reduce average reproductive lifespans in inbred animals. Revised estimates of the cost of inbreeding are calculated.

Partitioning additive, dominance, epistatic and maternal effects on reproductive performance in crosses between subspecies of Peromyscus polionotus mice

Robert C. Lacy
Dept. of Conservation Biology, Chicago Zoological Society

Intercrossing populations which have been evolving independently may increase fitness due to greater heterozygosity, or decrease fitness due to disruption of coadapted gene complexes. Whether heterosis or outbreeding depression predominate is unknown, because few studies have examined effects of intercrossing natural populations. The effects of intercrossing were assessed for subspecies of Peromyscus polionotus, which include closely related populations from contiguous and similar habitats as well as highly divergent subspecies long-isolated in areas of dissimilar habitat. F1, F2, and F3 crosses were made between each pair of five subspecies, and aspects of reproductive performance recorded. Results showed strong hybrid vigor, which surprisingly increased in the F2 and F3 generations. The high fitness of F2 and F3 pairs may have resulted from positive epistasis or from parental dominance effects in which heterozygous dams and sires had high reproductive success. To determine the genetic basis of effects of intercrossing, all F1, F2, F3, backcrosses, and 3-way crosses were made for three subspecies. This allows partitioning additive effects, dominance (heterosis), and epistasis (disruption of coadapted gene complexes). There were strong benefits of heterozygosity of offspring, maternal, and paternal genotypes. However, for crosses between more divergent populations there were also negative epistatic effects, suggesting that coadapted gene combinations had been disrupted.

Phylogeographic differentiation in the mitochondrial control region in the koala, Phascolarctos cinereus (Goldfuss, 1817), reveals a species comprising a single Evolutionarily Significant Unit and multiple Management Units

B.A. Houlden1,2, B.H. Costello2, D. Sharkey2, E.V. Fowler3, W.B. Sherwin2, A. Meltzer4, W. Ellis5, F. Carrick5, P.R. Baverstock6 and M.S. Elphinstone6
1Zoological Parks Board of NSW, PO Box 20, Mosman, 2088
2School of Biological Science, University of NSW, 2052
3School of Life Science, Queensland University of Technology, Brisbane, 4001
4Central Queensland University, Rockhampton 4702
5Dept. of Zoology, University of Queensland, 4072
6Faculty of Resource Science and Management, Southern Cross University, Lismore, 2480

The koala, Phascolarctos cinereus, is a geographically widespread species, with 3 currently recognised subspecies: P.c adustus, P.c. cinereus, and P.c. victor. Infraspecific variation in the mitochondrial DNA (mtDNA) control region was examined in over 200 animals from 16 representative populations throughout the speciesí range. Eighteen different haplotypes were defined in the 860bp mtDNA control region, as determined by temperature gradient gel electrophoresis (TGGE) analysis. Any single population typically possessed only one or two haplotypes yielding an average within-population haplotypic diversity of 0.180, and nucleotide diversity of 0.16%. Overall, mtDNA control region sequence diversity between populations averaged 0.67%, and ranged from 0% to 1.56%. Nucleotide divergence between populations averaged 0.51%, and ranged from 0% to 1.53%. Evolutionary relationships between mtDNA sequence were analysed using the neighbour joining method. This revealed limited phylogenetic distinction between geographically distant populations of koalas, with support for a single evolutionarily significant unit (ESU). Based on the combined molecular analysis, we recommend that the morphological differences formalised by subspecific taxonomy be interpreted as clinal variation, probably as a result of adaptations to local climate. Significant differentiation in mtDNA-haplotype frequencies between localities suggested that little gene flow currently exists among populations. When combined with microsatellite analysis, which have revealed substantial differentiation between koala populations, we conclude that the appropriate short-term management unit (MU) for koalas is the local population.

Estimating power when using genetic data to define management units

B. Taylor, S. Chivers and A. Dizon
Southwest Fisheries Science Center, P.O. Box 271, La Jolla, CA 92038- 0271 USA (taylor@caliban.ucsd.edu)

For many species molecular genetic techniques provide the only means to estimate population structure and dispersal rates. The null hypothesis is usually that putative populations are panmictic. Rejecting this hypothesis indicates such low dispersal between adjacent areas that putative populations should be treated as different management units. However, not rejecting the null hypothesis is insufficient evidence for pooling adjacent areas. For large populations, dispersal of a few animals per year could be sufficient for the level of population differentiation to be very low and likely very hard to detect. We develop a technique to estimate statistical power through a simulation model where spatial structure and abundances are specified. Both the uncertainty because of stochastic inheritance and sample size are accounted for. Using our technique, geneticists can answer pertinent management questions like: "If the dispersal rate was 1%/year, what is the probability that we would have correctly rejected the null hypothesis of panmixia?" We develop graphic displays of results that do not require the scientist to choose significance criteria. We also show the effect of increasing sample size, the relative power of different statistics for measuring population differentiation, as well as how our results differ from analytical results.

Conservation as the preservation of coding information content: phylogeny and the preservation of genetic diversity

R.H. Crozier
Department of Genetics and Human Variation, La Trobe University

Perhaps tens of millions of species inhabit the Earth. The diversity of life may best be conserved by not regarding all species as equally valuable. The preservation of the coding information content of the Earth's genomes has been suggested as a theoretical framework for conservation biology. The information content approach implies phylogeny and potentially reduces the need for universal certainty of species status. Molecular phylogeny promises a relatively objective framework for the relative worth of individual species, and in principle is automatable. The information content approach also involves taking account of the number of genes and their evolutionary history - other things being equal the information content worth of a mammal is less than four times that of a Drosophila. The phylogenetic approach rests on predictivity and the certainty that for nearly all species most aspects of the phenotype are unknown. Species which are phylogenetically close are more likely to be similar in phenotype, including unknown aspects, than are highly divergent species. Phylogeny should assist in the assessment of species in particular groups, and with foreseeable technology to whole communities. The possibility has been suggested that, as the number of species becomes very large, phylogenetic effects may be swamped and simple species richness might suffice to rank habitats, but empirical work on subterranean bacterial communities indicates that phylogenetic assessment is important when there are systematic differences between habitats.

Genetic indicators for state of the environment reporting

W. Sherwin1, A. Brown2, A. Young2, J. Burdon2, L. Christidis3, G. Clarke4 and D. Coates5
1School of Biological Sciences, University of New South Wales
2Centre for Plant Biodiversity Research, CSIRO Plant Industry
3National Museum of Victoria
4CSIRO Entomology
5WA Herbarium, Dept. Conservation and Land Management

Seven indicators of genetic diversity for state of the environment reporting in Australia are described. These are: number of subspecific taxa, population size, numbers and isolation, environmental amplitude of populations, genetic diversity at marker loci within individual populations, quantitative genetic variation, inter-population genetic structure, and mating. A rationale, monitoring strategy and list of potential data sources are given for each indicator. The selection of target taxa against which to monitor the indicators, interpretation of the indicators, and research needs are discussed.

Genetic management of chondrodystrophy in California condors

Katherine Ralls1, Jonathan Ballou1 and Richard Frankham2
1National Zoological Park, Smithsonian Institution, Washington D.C. 20008, USA
2School of Biological Sciences, Macquarie University, North Ryde, NSW 2109

Four of 19 chicks (21%) produced by one pair of captive California condors exhibited a lethal chondrodystrophy. Two forms of lethal chondrodystrophy in turkeys are due to a single autosomal recessive allele. Assuming that chondrodystrophy in condors is inherited in a similar fashion, we analyzed the pedigree of the condor population and made recommendations for genetic management of this trait. The frequency of the putative allele in the population is about 0.09, so the homozygote condition would occur in only about 1% of fertile eggs (0.09 squared) if the population were mating at random. Seventy-one of the 134 living birds are potential carriers of the allele. Selecting against the allele by not breeding potential carriers is demographically and genetically costly: it would lead to a decrease in population size at a time when rapid population growth is desirable and a decrease in the population's limited genetic variation. Furthermore, other genetic defects will undoubtedly appear in the population and a strategy of selecting against them all will not be possible. We recommend no active selection against this putative allele, avoiding the pairing of two birds with a high probability of being carriers to minimize the production of homozygote chicks, and re-pairing each member of any pair that produces an affected chick with another bird that has a low probability of being a carrier. Many other captive populations descended from a small number of founders may exhibit a variety of genetic defects and require a similar management strategy.

Can captive breeding and reintroduction rescue wild gene pools? A pedigree analysis of the Eastern Barred Bandicoot

Matthew Macdonald1, Peter Myroniuk2, John Seebeck3 and Neil Murray1
1Dept. of Genetics and Human Variation, La Trobe University
2Melbourne Zoo
3Dept. of Natural Resources and Environment, Victoria

In the light of the current rate of species extinctions, captive breeding and reintroduction programs are being attempted for a considerable number of threatened populations. The genetic diversity of such populations is thought to be an important determinant of the ultimate success of such programs. The Australian mainland population of Eastern Barred Bandicoot Perameles gunnii, whilst once widespread in south western Victoria, declined to several hundred individuals around the provincial city of Hamilton by the late 1980s, and has since declined to extinction. Following successful captive breeding, seven reintroduced populations have been established. Through pedigree analysis, using gene-dropping and a novel use of genetic distance metrics, heterozygosities and founder contributions for the reintroduced populations and the current captive population were estimated. Whilst the data are incomplete, it appears that the Woodlands Historic Park population is likely to be the most representative of the original gene pool. This applies to both nuclear and mitochondrial diversity. Genetic differences between the reintroduced sub-populations have arisen despite the intensive genetic management of captive breeding. The reasons for such differences are discussed, as are ways of more effectively conserving the original gene pool, and potential tests of the predictions arising from this analysis.

Mitochondrial DNA and microsatellite variability in a founder population of black rhinoceros (Diceros bicornis minor).

Sarah M. Brown1,2 and Bronwyn A. Houlden3,4
1Department of Evolutionary Biology, Australian Museum, 6 College St., Sydney, 2000
2Faculty of Science, School of Natural Resource Management, UNE, Armidale, NSW
3Zoological Parks Board of NSW, Mosman, 2088
4School of Biological Science, University of NSW, 2052

The black rhino, Diceros bicornis, is one of five extant species of rhinoceros belonging to four genera, all of which are endangered. The black rhino once ranged across much of Africa, and twenty years ago the total population was approximately 65,000. Due to severe poaching, it is now estimated that 2,500 individuals remain. Many of these exist in fragmented populations of less than 50. As these remnant populations become increasingly small and isolated, an understanding of the evolutionary histories and metapopulation structure is desirable in order to make informed decisions on aspects of population augmentation, relocation and resource allocation in conservation programmes.

Ten Diceros bicornis minor individuals originating from Zimbabwe have been imported to Western Plains Zoo, to establish an ex situ population. The viability and success of captive breeding programmes are dependent the maintenance of genetic variability and avoidance of inbreeding depression over the long-term. In practice, the management of a captive breeding population assumes that founder individuals obtained from the wild are unrelated, and this may not be the case. In order to ascertain the kinship and genetic variation of the population at Western Plains Zoo, analyses of microsatellites and mitochondrial DNA were undertaken. We found high levels of allelic diversity and heterozygosity in the population, consistent with levels typically found in pre-bottleneck populations. In addition, the phylogenetic relationship between the Diceros bicornis minor population, Diceros bicornis michaeli and Ceratotherium simum were investigated. High levels of sequence divergence were found between the two subspecies of D. bicornis and C. simum.

Distribution of genetic load among founders of captive populations

Jonathan D. Ballou
Department of Zoological Research, National Zoological Park, Smithsonian Institution, Washington, D.C.

Inbreeding depression is well documented in many captive, laboratory and some wild populations. Characterizing the distribution of genetic load among individuals, families or founders within these populations provides information on the degree and number of deleterious alleles causing inbreeding depression. In this analysis, I evaluate the individual contribution of founders and founder pairs to the overall inbreeding depression in four captive populations known to exhibit inbreeding depression for early mortality: the golden lion tamarin (Leontopithecus rosalia), Goeldi's marmoset (Callimico goeldi), pygmy hippopotamas (Choeropsis liberiensis), Przewalski horse (Equus przewalskii) and European bison (Bison bonasus). Multiple logistic regression was used to estimated the effects of founder partial inbreeding coefficients (i.e., the probability of alleles being identical-by-descent from a specific founder) on early survival in each of these populations. While analytical complications arise due to complexity of the pedigrees and the unequal representation of all founders' genomes in unique and variable inbreeding combinations, the results indicate that most of the inbreeding depression observed is due to genes from only a small proportion of the founders. Load distributed among only a few of a population's founders supports the hypothesis that load consists primarily of a few deleterious alleles with major effects as opposed to many alleles of minor effects. Since founder inbreeding depression effects can also be due to differences in sources of the founders (i.e., populations with different inbreeding histories), knowledge of founders' origins is also important in interpreting these results.

Rapid genetic deterioration in captive populations: causes and implication

R. Frankham, L.M. Woodworth, M.E. Montgomery, S. Margan and D.A. Briscoe
School of Biological Science, Macquarie University, NSW 2109

Many species require captive breeding to ensure their survival, with reintroduction into the wild being a long-term objective. However, genetic adaptation to captive conditions is likely to lead to reductions in reproductive fitness when they are reintroduced into the wild. We maintained 23 populations of Drosophila under benign captive conditions for 50 generations with effective sizes of 500 (2 replicates), 250 (3), 100 (4), 50 (6) and 25 (8). Family sizes were equalised to minimise selection. Reproductive fitness in the benign captive conditions (productivity per pair) at generation 50, indicated modest genetic adaptation in some of the larger populations and inbreeding depression in many of the smaller populations. Reproductive fitness under crowded, competitive conditions (competitive index) show a marked decline in all 23 populations. There was a curvilinear relationship between fitness and populations size, the 500s and 25s being the worst. Genetic adaptation was a major cause of decline, especially in large populations, while inbreeding depression made a major contribution in the smaller populations. No significant contribution from mutational accumulation was found. Genetic adaptation to captivity is likely to be a major problem in captive breeding of endangered species. Fragmentation of endangered species, with occasional immigration is recommended to minimise this problem. Our experimental results provide support for this strategy.

Genetic factors and the extinction proneness of small populations

Mark Eldridge1 and Juliet King2
1School of Biological Sciences, Macquarie University, NSW 2109
2Department of Zoology, University of Western Australia, Nedlands, WA 6907

It has been argued that demographic and/or environmental factors will cause small isolated populations to become extinct before genetic factors have a significant negative impact. Here we report that the Barrow Island population of the black-footed rock-wallaby Petrogale lateralis (Marsupialia: Macropodidae) has unprecedented low levels of genetic variation (He = 0.053; from 10 microsatellite loci) and suffers from inbreeding depression (reduced female fecundity, skewed sex ratio, increased levels of fluctuating asymmetry). Despite a long period of isolation (~1600 generations) and small effective population size (Ne~15), demographic and/or environmental factors have not yet driven this population to extinction. However, it has been significantly impacted by genetic factors, losing most of its genetic variation, becoming highly inbred (Fe = 0.91) and having reduced fitness through inbreeding depression. As five other island populations of P. lateralis similarly exhibit exceptionally low levels of genetic variation, this phenomenon may be widespread. Inbreeding in these populations is at a level associated with high rates of extinction in populations of domestic and laboratory species. Genetic factors can not then be excluded as contributing to the extinction proneness of small isolated populations.

Intron variation in marbled murrelets: new genetic tools for old ecological problems

B.C. Congdon1, J.F. Piatt2, K. Martin3 and V.L. Friesen1
1Queen's University, ON, Canada
2National Biological Service, Alaska
3Canadian Wildlife Service, BC, Canada

Marbled murrelets are coastal seabirds that breed predominantly in old growth forests throughout the North-west Pacific. Murrelets are in direct conflict with logging interests, and are officially listed as 'Threatened' or 'Endangered' throughout much of their southern distribution. Effective long-term management of this species requires detailed information on levels of interpopulation gene flow. Our study uses comparative data from nuclear introns and microsatellites to examine population genetic structure in marbled murrelets. Little or no differentiation was observed among populations over 2500 kms from the Alaskan peninsula to British Columbia. Substantial gene flow has occurred, or continues to occur among sampling locations in this region. Significant isolation was observed at shorter distances among populations in the central and western Aleutians. Breeding birds are naturally sparse or absent from many locations in this region. We conclude that historically, marbled murrelets along the Canadian and Alaskan coastlines have not been distributed as small isolated breeding populations. However, our results suggest that fragmentation of nesting habitat in this region has the potential to mimic natural fragmentation of murrelet populations in the Aleutians, and so isolate threatened populations at the southern end of the distribution.

Applications of genetics to the conservation of salmonid fishes

F.W. Allendorf
Division of Biological Sciences, University of Montana, Missoula, MT 59812, USA

Genetics has played an essential role in the management and conservation of salmonid fishes over the last 25 years. In fact, genetic issues have played a greater role in the conservation of salmonids than in perhaps any other taxonomic group for several reasons: their polyploid ancestry, their widespread artificial propagation, their homing to natal streams that has produced a network of local reproductive populations that are demographically and genetically isolated, their harvesting in mixed-stock fisheries, and the widespread occurrence of hybridization and introgression among salmonid species. The ability to study genetic variation at a virtually unlimited number of markers throughout the nuclear genome is likely to revolutionize our understanding of the genetics of salmonid species over the next 10 years. The genetic analysis of salmonids is extremely complicated relative to the elegant simplicity of disomic Mendelian inheritance. Residual tetrasomic inheritance has been reported for some loci in salmonids. Unlike disomic ratios, tetrasomic ratios are affected by differential pairing affinities, multivalent formation, crossovers, and the position of a locus in relation to the centromere.

Analysis of DNA from old scale samples: a new tool in conservation biology of salmonid fishes.

E.E. Nielsen, M.M. Hansen and V. Loeschcke
Danish Institute for Fisheries Research, Department of Inland Fisheries, Vejlsøvej 39, 8600 Silkeborg, Denmark
Department of Ecology and Genetics, University of Aarhus, Ny Munkegade, Building 540, 8000 Aarhus C, Denmark

Supportive breeding is a common practice in association with conservation projects for salmonid fishes. Due to the high degree of differentiation that is commonly found among populations of salmonids the need for use of local fish has been recognised in order to preserve large scale genetic diversity and possible local adaptations. Individuals to be used for supportive breeding should represent the native population, however the remaining individuals in a population may be the result of immigration from other populations or aquaculture, and therefore do not possess the genetic characteristics of the original population. We here present a study of a Danish population of Atlantic salmon (Salmo salar L.) from the Skjern River, which has decreased dramatically in size during the last century. There has even been some doubt as to whether the original population has gone extinct and that the present population has been founded by immigrants. To obtain genetic data from before any severe reduction in population size took place, we extracted DNA from scale samples collected in the 1930's. We compared the genetic composition of the population 60 years ago with recent genetic data, in order to determine the genetic relationship between the Skjern River population of the 1930's and the population in the river today and to evaluate possible changes in the levels of variability.

Population phylogeny, genetic structure and the mating system of endangered Lambertia orbifolia (Proteacea): implications for conservation

David J. Coates and Margaret Byrne
Conservation and Land Management , Locked Bag 104, Bentley Delivery Centre, WA 6983

Lambertia orbifolia is a bird pollinated woody shrub restricted to seven populations in south west Western Australia. Populations occur in two disjunct groups (Scott River Plains and Narrikup) some 200 km apart. Allozyme studies, based on 19 loci, revealed a high level of divergence between the two groups (Nei's D = 0.252, overall FST = 0.469) suggesting significant historical isolation and phylogeographic structure. This was supported by cpDNA and to some extent rDNA studies. Analysis of variation in the chloroplast genome showed two haplotypes one specific to each group. Variation in rDNA revealed different gene unit lengths in Narrikup and the two Scott River coastal populations, whilst the three inland Scott River populations were polymorphic for both gene unit lengths. Analysis of genetic variation within populations, based on allozymes, indicated that allelic diversity and heterozygosity were slightly higher in the smaller Narrikup populations. Mating system estimates for four populations, based on the mixed mating model, revealed a significantly higher level of inbreeding in the Narrikup population. This appears to be associated with increased habitat disturbance and a reduction in bird pollinators. The two Narrikup populations are critically endangered and declining. These data indicate the need to manage them as a separate conservation unit and that there is sufficient genetic variation within the Narrikup populations to support a proposed translocation program.

Metapopulation structure and conservation status in Spotted Owls

Susan M. Haig and Thomas D. Mullins
U.S. Geological Survey, Forest and Rangeland Eco-system Science Center, 3200 SW Jefferson Way, Corvallis OR 97331, USA (HaigS@FSL.ORST.EDU)

The concept of metapopulation has broadened our view for species conservation planning. However, often data are lacking that describe a metapopulation. In this study, we use RAPDs to examine various hypotheses regarding meta-population stucture in Spotted Owls (Strix occidentalis) so as to better understand relationships among breeding areas, local groups of breeding areas, regional groups and subspecies. We sampled 28 breeding areas (283 individuals) among the three subspecies of Spotted Owls. Results attained from 17 variable RAPD loci indicated a significant relationship between geographic distance and genetic distance (Mantel r = 0.24, p = 0.04) as well as significant differentiation at local and regional levels. Estimates of gene flow among breeding areas ranged from 0.6-1.0, far below recommended guidelines for healthy populations. Cluster analyses suggest regional groupings are the most defining for Spotted Owls. Few differences were found among Northern and California Spotted Owls in any analysis, whereas Mexican Spotted Owls clustered together to form a separate group. Two loci (12%) differentiated California and Northern Spotted Owls from Mexican Spotted Owls. We suggest that listing of Spotted Owls under the U.S. Endangered Species Act based on current taxonomy should be reconsidered.

An evaluation of the definitions and rationales for 'evolutionarily significant unit' concepts

Robert C. Fleischer

National Zoological Park, Smithsonian Institution, Washington, DC, 20008 USA (fleischer@nzp.si.edu)

An evolutionarily significant unit (ESU) may be considered to be the smallest taxonomic unit with an independent evolutionary trajectory. Many criteria for defining ESUs have been proposed and compared, but there has been little discussion about how important it really is to maintain ESUs. Rationales for ESU protection include avoidance of outbreeding depression and behavioral isolation, maintenance of genetic variability, and preservation of "natural" or historical patterns of variation. Some definitions require only minor divergence from other populations for a population to be classified as an ESU. How much divergence must there be among ESUs before the rationales noted above become important? Should ESUs be used to protect habitat (in the absence of geographical concordance with other taxa)? How do we avoid trivial designations of ESUs, such as anthropogenically-induced and recently formed population fragments (especially when resources for management of ESUs are limited)? Examples of genetic analyses that address these questions about ESUs will be provided, including recent population genetic studies of Asian elephants, Mariana birds, and clapper rails.

Taxonomy and populations structure in rock wallabies: conservation implications

Robert Close
University of Western Sydney Macarthur, PO Box 555, Campbelltown 2560

The macropod genus Petrogale currently comprises 15 species, 4 subspecies and two chromosomal races. The history of the taxonomy of the group is one of regular change and shifting affinities. Studies are continuing and more minor changes are expected. This change in taxonomy has proved difficult for researchers and wildlife officers alike. Most specimen boxes of Petrogale skulls in Australian museums bear several species names with succeeding layers of out-dated nomenclature struck out.

This changing taxonomy reflects a fascinating history of specimen collection and taxonomic decision making. Early collections were sporadic hunting excursions by ship's surgeons and a 2 year marathon collecting expedition by the British Museum to gather skins and skulls. More modern techniques have required 4WD forays to isolated areas seeking tissues for chromosome, allozyme and DNA analysis. Comparison of parasite species has also shed light on relationships between host species.

An unexpected finding was the extent of hybridisation that exists between species despite considerable chromosomal differences. This hybridisation has led to analysis of contact zones between species and consideration of the need to conserve such contacts. Overall the studies of this fascinating genus have led to insights to the process of speciation and the generation of chromosome changes. From a conservation perspective, it has been important to identify all the different genetic components of this diverse genus and develop an accurate taxonomy. Allotting priorities for conservation of this diversity will be a future and difficult step.