Genetics of the scandinavian brown bear (Ursus arctos) : implications for biology and conservation

摘要: This thesis deals with the application of molecular tools, combined field data, in wildlife management, conservation and understanding species' biology behavior. We used brown bear (Ursus arctos) as a model species Scandinavian population case study. The first part this is methodological part, which we developed or reviewed technical aspects parentage analysis; second devoted to genetics estimate sizes understand mating systems. Noninvasive methods are gaining widespread use genetic studies they do not require handling disturbance study animal. However, DNA recovered from noninvasive samples, such hairs feces, usually degraded and/or small quantities, leading genotyping errors resulting identification incorrect genotypes. major concern, especially for endangered populations, it can lead biases size estimates. With aim increasing quality quantity desired template, avoid need numerous replicates, devised two-step polymerase chain reaction (PCR) method. “multiplex pre-amplification method” was tested on different compared conventional PCR approach. It significantly improved microsatellite amplification decreased error rates fecal limiting conditions. To more specifically amplify samples bears, also redesigned primers one sex-specific primer semi-nested multiplex These new approaches could be transposed other where experience low success due concentration quality. Genotyping remain taboo subject studies, spite their occurrence most datasets negative consequences may cause interpretation results. considered four representing large variety investigations, track identify causes. In these estimated rate ranged 0.8% 2.6%, depending organism marker used. Main sources were allelic dropouts microsatellites differences peak intensities AFLPs (Amplified Fragment Length Polymorphism), but both cases, human factors non-negligible generators. present suggestions limit quantify at each step process recommend systematic reporting studies. Parentage analyses using multilocus genotypes widely assess reproductive success, patterns, kinship fitness natural populations. Several approaches, based maximum likelihood estimations /or Bayesian inference, have been recently developed, often theoretical difficult biologists apply. there clear lack assignment softwares that able consider several generations individuals allow determination parents without any prior assumptions. software PARENTE conduct inference data diploid codominant markers. Based principle compatibility, looks maternity, paternity simultaneously potential parents, birth death dates (if available). calculates probability successfully allocating an individual offspring its parents. Estimates density essential successful management species. few attempts made evaluate accuracy estimates obtained. Using protocols amplifying DNA, census methods. Two rarefaction indices two capture-mark-recapture (CMR) estimators. A total 1904 collected over 2 consecutive years 49,000-km² area south-central Sweden. Population 378 572 bears 2001 273 433 2002, method calculated minimum radio-telemetry concluded best program MARK, CMR estimator, accurate. included heterogeneity temporal variation detection probabilities, appeared our samples. Second, evaluated reliability three traditional comparison performing smaller (7,328 km²). All tended underestimate size; MARK estimator seemed perform best. approximately 550 (482-648) 223 (188-282) 7,328-km² during 2002. suggest has reached threshold core currently expands edge area. cost/benefit analysis showed less expensive than reliable preferable ethical point view. conclusion, methods, areas. out importance adequate well-distributed sampling effort advise calibration independent biased sampling, if possible. Future should collecting 2.5 3 times number “assumed” animals. confirmed good status. knowledge systems important evolution sexual selection. studied system, namely strategies employed by sexes relation sexually selected infanticide (SSI) female mate Infanticide, killing dependent young, adaptive males, following requirements fulfilled: i) shortens time mother's next estrus, ii) perpetrator father killed infants, iii) perpetrators sire female's litter. benefit females evolved counterstrategies order defend infants against infanticidal males. documented eight cases field. From sites observations, verified all SSI fulfilled, suggesting male strategy nonsocial carnivore. Contrary social species, mostly immigrant males kill mainly resident adult bears. implies differentiate own progeny unrelated cubs, perhaps recognizing mated with. Moreover, genetically 14.5% multiple paternities (28% litters young more). Female promiscuity confuse therefore counterstrategy SSI. Further, assessed criteria chose partner(s). hypothesized faced dilemma: either select high partner morphological criteria, suggested theories choice, rather future potentially i.e. geographically closest whether traits influenced found heterozygous, largest oldest might counter-strategy exercise post-copulatory cryptic body dominance, reflecting