Spatial Population Structure of Yellowstone Bison

Abstract: Increases in Yellowstone National Park, USA, bison (Bison bison) numbers and shifts in seasonal distribution have resulted in more frequent movements of bison beyond park boundaries and development of an interagency management plan for the Yellowstone bison population. Implementation of the plan under the adaptive management paradigm requires an understanding of the spatial and temporal structure of the population. We used polythetic agglomerative hierarchical cluster analysis of radiolocations obtained from free-ranging bison to investigate seasonal movements and aggregations. We classified radiolocations into 4 periods: annual, peak rut (15 Jul-15 Sep), extended rut (1 Jun-31 Oct), and winter (1 Nov-31 May). We documented spatial separation of Yellowstone bison into 2 segments, the northern and central herds, during all periods. The estimated year-round exchange rate (4.85-5.83%) of instrumented bison varied with the fusion strategy employed. We did not observe exchange between the 2 segments during the peak rut and it varied during the extended rut (2.15-3.23%). We estimated a winter exchange of 4.85-7.77%. The outcome and effectiveness of management actions directed at Yellowstone bison may be affected by spatial segregation and herd affinity within the population. Reductions based on total population size, but not applied to the entire population, may adversely affect one herd while having little effect on the other. Similarly, management actions targeting a segment of the population may benefit from the spatial segregation exhibited.     

Emigration and Density Dependence in Yellowstone Bison

Abstract: Understanding the relative importance of density-dependent and density-independent feedback on population growth is essential for developing management strategies to conserve wildlife. We examined a 99-year time series of annual counts and removals for 2 bison (Bison bison) herds occupying northern and central Yellowstone National Park in the western United States. Yellowstone's aggressive management intervention effectively recovered bison from 46 animals in 1902 to > 1,500 animals in 1954. Supplemental feeding of the northern herd facilitated rapid growth (r = 0.16) during 1902 to 1952. Augmentation of the central herd with 71 animals also led to rapid growth over 1936 to 1954 (r = 0.10). In 1969, manipulative management ceased in the park, and we detected evidence of density-dependent changes in population growth rates for both herds during 1970 to 2000 as numbers increased to >3,000 animals. The central herd showed evidence of a constant density-dependent response over 1970 to 2000. In contrast, density dependence had a stronger effect on the northern herd's growth rate during 1970 to 1981 than during 1982 to 2000. We found evidence to suggest that these trends resulted from pulses of emigration from the central herd to the northern range beginning in 1982 in response to resource limitation generated by an interaction between density and severe snow pack. Corroborative evidence supporting this interpretation included 1) the annual growth of the central herd was negatively correlated with snow pack but that of the northern herd was not, 2) growth rates of the central and northern herds were uncorrelated during 1970 to 1981 but significantly and negatively correlated during 1982 to 2000, and 3) the northern herd could not have sustained the high removals experienced during 1984 to 2000 without immigration. Density-related emigration from the central herd to the northern range may be fueling bison emigration onto private and public lands where large-scale removals occur, exacerbating the brucellosis controversy for natural resource managers.     

Environmental persistence of Brucella abortus in the Greater Yellowstone Area

Bison (Bison bison) and elk (Cervus elaphus) of the Greater Yellowstone Area (GYA) are the last remaining reservoirs of bovine brucellosis (Brucella abortus) in the United States. An important factor in evaluating the risk of transmission to cattle is the persistence of bacteria and infectious birth materials shed on pastures where cattle graze. We selected 2 study areas near the northern and western boundaries of Yellowstone National Park (YNP) to determine the persistence of bacteria on fetal tissue, soil, and vegetation, and scavenging on infectious materials from birth and abortion sites. We performed 3 independent field experiments to determine: 1) persistence of Brucella abortus (RB51) purposely applied to fetal tissues, 2) scavenging of fetuses by native scavengers, and 3) natural contamination of birth or abortion sites in the GYA. Results from these field experiments established that Brucella bacteria can persist on fetal tissues and soil or vegetation for 21–81 days depending on month, temperature, and exposure to sunlight. Bacteria purposely applied to fetal tissues persisted longer in February than May and did not survive on tissues beyond 10 June regardless of when they were set out. Brucella abortus field strain persisted up to 43 days on soil and vegetation at naturally contaminated bison birth or abortion sites. Fetuses were scavenged by a variety of birds and mammals in areas near YNP and more rapidly inside YNP than outside the Park boundary. Models derived from our data determined a 0.05% chance of bacterial survival beyond 26 days (95% Credible Interval of 18–30 days) for a contamination event in May. May 15 is the final date for hazing all bison into Yellowstone National Park under the current interagency bison management plan. With these data managers can predict when it is safe to graze cattle onto pastures previously occupied by bison. © 2011 The Wildlife Society.     

Genetic Analysis of the Henry Mountains Bison Herd

Wild American plains bison (Bison bison) populations virtually disappeared in the late 1800s, with some remnant animals retained in what would become Yellowstone National Park and on private ranches. Some of these private bison were intentionally crossbred with cattle for commercial purposes. This forced hybridization resulted in both mitochondrial and nuclear introgression of cattle genes into some of the extant bison genome. As the private populations grew, excess animals, along with their history of cattle genetics, provided founders for newly established public bison populations. Of the US public bison herds, only those in Yellowstone and Wind Cave National Parks (YNP and WCNP) appear to be free of detectable levels of cattle introgression. However, a small free-ranging population (~350 animals) exists on public land, along with domestic cattle, in the Henry Mountains (HM) of southern Utah. This isolated bison herd originated from a founder group translocated from YNP in the 1940s. Using genetic samples from 129 individuals, we examined the genetic status of the HM population and found no evidence of mitochondrial or nuclear introgression of cattle genes. This new information confirms it is highly unlikely for free-living bison to crossbreed with cattle, and this disease-free HM bison herd is valuable for the long-term conservation of the species. This bison herd is a subpopulation of the YNP/WCNP/HM metapopulation, within which it can contribute significantly to national efforts to restore the American plains bison to more of its native range.     

Connecting the dots: an invariant migration corridor links the Holocene to the present

Numerous species undergo impressive movements, but due to massive changes in land use, long distance migration in terrestrial vertebrates has become a highly fragile ecological phenomenon. Uncertainty about the locations of past migrations and the importance of current corridors hampers conservation planning. Using archeological data from historic kill sites and modern methods to track migration, we document an invariant, 150 km (one-way) migration corridor used for at least 6000 years by North America's sole extant endemic ungulate. Pronghorn (Antilocapra americana) from the Greater Yellowstone Ecosystem, like other long distant migrants including Serengeti wildebeest (Connochaetes taurinus) and Arctic caribou (Rangifer tarandus), move nearly 50 km d-1, but in contrast to these other species, rely on an invariant corridor averaging only 2 km wide. Because an entire population accesses a national park (Grand Teton) by passage through bottlenecks as narrow as 121 m, any blockage to movement will result in extirpation. Based on animation of real data coupled with the loss of six historic routes, alternative pathways throughout the 60,000 km2 Yellowstone ecosystem are no longer available. Our findings have implications for developing strategies to protect long distance land migrations in Africa, Asia and North America and to prevent the disappearance of ecological phenomena that have operated for millennia.     

The biogeochemistry of a north-temperate grassland with native ungulates: Nitrogen dynamics in Yellowstone National Park

Nutrient dynamics of large grassland ecosystems possessing abundant migratory grazers are poorly understood. We examined N cycling on the northern winter range of Yellowstone National Park, home for large herds of free-roaming elk (Cervus elaphus) and bison (Bison bison). Plant and soil N, net N mineralization, and the deposition of ungulate fecal-N were measured at five sites, a ridgetop, mid-slope bench, steep slope, valley-bottom bench, and riparian area, within a watershed from May, 1991 to April, 1992.Results indicated similarities between biogeochemical properties of Yellowstone grassland and other grassland ecosystems: (1) landscape position and soil water affected nutrient dynamics, (2) annual mineralization was positively related to soil N content, and (3) the proportion of soil N mineralized during the year was negatively related to soil C/N.Grazers were a particularly important component of the N budget of this grassland. Estimated rates of N flow from ungulates to the soil ranged from 8.1 to 45.6 kg/ha/yr at the sites (average = 27.0 kg/ha/yr), approximately 4.5 times the amount of N in senescent plants. Rates of nitrogen mineralization for Yellowstone northern range grassland were higher than those measured in other temperate grassland ecosystems, possibly due to grazers promoting N cycling in Yellowstone.     

Migrations and swimming capabilities of endangered pallid sturgeon (Scaphirhynchus albus) to guide passage designs in the fragmented Yellowstone River

Fragmentation of the Yellowstone River is hypothesized to preclude recruitment of endangered Scaphirhynchus albus (pallid sturgeon) by impeding upstream spawning migrations and access to upstream spawning areas, thereby limiting the length of free‐flowing river required for survival of early life stages. Building on this hypothesis, the reach of the Yellowstone River affected by Intake Diversion Dam (IDD) is targeted for modification. Structures including a rock ramp and by‐pass channel have been proposed as restoration alternatives to facilitate passage. Limited information on migrations and swimming capabilities of pallid sturgeon is available to guide engineering design specifications for the proposed structures. Migration behavior, pathways (channel routes used during migrations), and swimming capabilities of free‐ranging wild adult pallid sturgeon were examined using radiotelemetry, and complemented with hydraulic data obtained along the migration pathways. Migrations of 12–26% of the telemetered pallid sturgeon population persisted to IDD, but upstream passage over the dam was not detected. Observed migration pathways occurred primarily through main channel habitats; however, migrations through side channels up to 3.9 km in length were documented. The majority of pallid sturgeon used depths of 2.2–3.4 m and mean water velocities of 0.89–1.83 m/s while migrating. Results provide inferences on depths, velocities, and habitat heterogeneity of reaches successfully negotiated by pallid sturgeon that may be used to guide designs for structures facilitating passage at IDD. Passage will provide connectivity to potential upstream spawning areas on the Yellowstone River, thereby increasing the likelihood of recruitment for this endangered species.     

Microsatellite variation in the reintroduced Pennsylvania elk herd

Relocation programs have restored elk (Cervus elaphus) to portions of its vast historical range. We examine the consequences of these relocation programs by assessing variation at 10 microsatellite loci in three elk herds, a source herd (Yellowstone National Park), a large herd reintroduced from Yellowstone (Custer State Park) and a bottlenecked herd reintroduced from both Yellowstone and Custer (the Pennsylvania herd). Observed single locus heterozygosities ranged from 0.000 to 0.739. Multi-locus heterozygosities ranged from 0.222 to 0.589. Although significant differences were detected among all three herds, the Yellowstone National Park and Custer State Park herds possessed similar levels of variation and heterozygosity, and the genetic distance between these two herds was small. The Pennsylvania herd, on the other hand, experienced a 61.5% decrease in heterozygosity relative to its source herds, possessed no unique and few rare alleles, and the genetic distances between the Pennsylvania herd and its sources were large. Simulations were performed to identify bottleneck scenarios in agreement with levels of variation in the Pennsylvania herd. Our data confirm that the rate of population growth post-relocation may have important genetic consequences and indicate that theoretical predictions regarding the maintenance of genetic variation during relocation events must be viewed with caution when small numbers of a polygynous species are released.     

Genetic population substructure in bison at Yellowstone National Park

The Yellowstone National Park bison herd is 1 of only 2 populations known to have continually persisted on their current landscape since pre-Columbian times. Over the last century, the census size of this herd has fluctuated from around 100 individuals to over 3000 animals. Previous studies involving radiotelemetry, tooth wear, and parturition timing provide evidence of at least 2 distinct groups of bison within Yellowstone National Park. To better understand the biology of Yellowstone bison, we investigated the potential for limited gene flow across this population using multilocus Bayesian clustering analysis. Two genetically distinct and clearly defined subpopulations were identified based on both genotypic diversity and allelic distributions. Genetic cluster assignments were highly correlated with sampling locations for a subgroup of live capture individuals. Furthermore, a comparison of the cluster assignments to the 2 principle winter cull sites revealed critical differences in migration patterns across years. The 2 Yellowstone subpopulations display levels of differentiation that are only slightly less than that between populations which have been geographically and reproductively isolated for over 40 years. The identification of cryptic population subdivision and genetic differentiation of this magnitude highlights the importance of this biological phenomenon in the management of wildlife species.     

Reproduction and Survival of Yellowstone Bison

ABSTRACT The conservation of bison (Bison bison) from near extinction to >4,000 animals in Yellowstone National Park has led to conflict regarding overabundance and potential transmission of brucellosis (Brucella abortus) to cattle. We estimated survival and birth rates from 53 radiocollared adult female bison during 1995–2001, and we used calf:adult (C:A) ratios to estimate reproduction with the combined effects of pregnancy, fetal loss, and neonatal mortality during 1970–1997. Annual survival of adult females was high (0.92; 95% CI = 0.87-0.95) and constant. Birth rates differed by brucellosis status and age. Birth rates were 0.40 calves per female (95% CI = 0.15-0.65) for brucellosis-positive 3 year olds, 0.63 (95% CI = 0.39-0.87) for individuals testing negative, and 0.10 (95% CI = 0.00-0.24) for individuals contracting brucellosis that birth year (sero-converters). Birth rates were 0.64 (95% CI = 0.52-0.76) for brucellosis-positive individuals ≥4 years old, 0.81 (95% CI = 0.73-0.89) for brucellosis-negative individuals, and 0.22 (95% CI = 0.00-0.46) for sero-converters. Spring C:A ratios were negatively correlated with snow pack (β = −0.01 to −0.03, R² = 0.26-0.60, P < 0.05). Growth rate was highly elastic to adult survival (0.51), and juvenile survival (0.36) was 3 times more elastic than fecundity (0.12). Simulations suggested brucellosis eradication via vaccination would result in increased birth rates and a 29% increase in population growth (γ = 1.09), possibly leading to more bison movements outside the park. Our results will help park managers evaluate bison population dynamics and explore consequences of management actions and disease control programs.     

Yersinia enterocolitica: an unlikely cause of positive brucellosis tests in greater yellowstone ecosystem bison (Bison bison)

Yersinia enterocolitica serotype O:9 has identical O-antigens to those of Brucella abortus and has apparently caused false-positive reactions in numerous brucellosis serologic tests in elk (Cervus canadensis) from southwest Montana. We investigated whether a similar phenomenon was occurring in brucellosis antibody-positive bison (Bison bison) using Y. enterocolitica culturing techniques and multiplex PCR of four diagnostic loci. Feces from 53 Yellowstone bison culled from the population and 113 free-roaming bison from throughout the Greater Yellowstone Ecosystem (GYE) were tested. Yersinia enterocolitica O:9 was not detected in any of 53 the bison samples collected at slaughter facilities or in any of the 113 fecal samples from free-ranging bison. One other Y. enterocolitica serotype was isolated; however, it is not known to cause cross-reaction on B. abortus serologic assays because it lacks the perosamine synthetase gene and thus the O-antigens. These findings suggest that Y. enterocolitica O:9 cross-reactivity with B. abortus antigens is unlikely to have been a cause of false-positive serology tests in GYE bison and that Y. enterocolitica prevalence was low in bison in the GYE during this study.     

The Influence of Discharge on Duration,Ascent Distance,and Fidelity of the Spawning Migration for Paddlefish of the Yellowstone-Sakakawea Stock,Montana and North Dakota,USA

Paddlefish, Polyodon spathula, of the Yellowstone-Sakakawea stock, Missouri and Yellowstone Rivers, Montana and North Dakota, were radio-tagged to assess the influence of spring discharge on duration of river residency, ascent distance, and site-fidelity during spawning migrations of 1999–2002. Contrary to expectations and reported results from other paddlefish populations, fish remained in the river for similar periods of time and ascended to similar reaches in years of higher, more sustained discharge and in years of lower, more fluctuating discharge. In all years, 65 of the 74 migrants (88%) restricted their ascent to reaches below Yellowstone River kilometer (YRkm) 55; only six migrants were found to further ascend to upriver reaches within 20 river kilometers (rkm) of the Intake Diversion Dam (YRkm 114). The lack of detectable annual differences in ascent distance over the study period despite annual differences in Yellowstone River spring flow regimes may have been partially attributed to the apparent site-fidelity demonstrated by the tagged fish over the study period. Ten of the 22 paddlefish contacted in more than one spring migration repeatedly limited their upriver movement to sites that were within 10 rkm of each other. In addition, similar to the reproductive homing tendencies documented in other large-river migratory fishes, site-fidelity occurred in different reaches of the river system. Results from this study suggest that, in years of moderate discharge, site-fidelity may be as influential as the spring flow regime in determining the reaches to which migratory paddlefish ascend. Further research is needed to investigate potential differential spawning success in fish that return to different reaches of the lower Yellowstone River.     

Bison PRNP genotyping and potential association with Brucella spp. seroprevalence

The implication that host cellular prion protein (PrP(C)) may function as a cell surface receptor and/or portal protein for Brucella abortus in mice prompted an evaluation of nucleotide and amino acid variation within exon 3 of the prion protein gene (PRNP) for six US bison populations. A non-synonymous single nucleotide polymorphism (T50C), resulting in the predicted amino acid replacement M17T (Met --> Thr), was identified in each population. To date, no variation (T50; Met) has been detected at the corresponding exon 3 nucleotide and/or amino acid position for domestic cattle. Notably, 80% (20 of 25) of the Yellowstone National Park bison possessing the C/C genotype were Brucella spp. seropositive, representing a significant (P = 0.021) association between seropositivity and the C/C genotypic class. Moreover, significant differences in the distribution of PRNP exon 3 alleles and genotypes were detected between Yellowstone National Park bison and three bison populations that were either founded from seronegative stock or previously subjected to test-and-slaughter management to eradicate brucellosis. Unlike domestic cattle, no indel polymorphisms were detected within the corresponding regions of the putative bison PRNP promoter, intron 1, octapeptide repeat region or 3'-untranslated region for any population examined. This study provides the first evidence of a potential association between nucleotide variation within PRNP exon 3 and the presence of Brucella spp. antibodies in bison, implicating PrP(C) in the natural resistance of bison to brucellosis infection.     

Movements of European bison (Bison bonasus) beyond the Białowieża Forest (NE Poland): range expansion or partial migrations?

The European bison is the largest terrestrial mammal in Europe. After extinction in the wild at the beginning of the twentieth century, it was re-introduced to Bia?owie?a Forest and other woodlands in Eastern Europe. In this paper, we analysed the movements of European bison beyond the continuous woodlands of the Bia?owie?a Forest (NE Poland) between 1964 and 2010. In total, 1,117 direct observations of bison were collected. The number of males moving out of the forest fluctuated during the study period, whilst the number of females steadily increased. The number of male observations outside of the forest per annum was dependent on the population size and snow depth, whilst the number of cows in mixed groups moving outside of the forest was correlated with the population size only. Males were observed mainly alone (50 % of observations) or in small groups of two to three individuals (25 %); however, distribution of group size differed from those observed in the population. There was a significant difference between the direction of movement of males and females out of the forest—males moved mainly west and southwest, whilst females moved to the north. This was also significantly different from the expected movement direction. The mean distance of bison observations from the forest border was 1.8?±?0.13 km and did not differ significantly between sexes. After 1990, males were observed significantly farther away from the forest (2.2 km) than in the previous years (0.9 km). Most observations (94 % of bulls and 93 % of cows) were up to 5 km from the forest edge. The range of bison in the vicinity of the Bia?owie?a Forest was strictly seasonal. Most observations (78 % in males and 88 % in females) were recorded from November to April. Increasing utilisation of areas beyond the forest habitats may be driven by different factors but most probably it is related to range expansion and the bison's preference for open habitats. The strong seasonal pattern of bison movements indicate that the partial seasonal migrations were initiated in the Bia?owie?a population.     

Characterizing population and individual migration patterns among native and restored bighorn sheep (Ovis canadensis)

相似文献   

Comparative migration strategies of wild and captive‐bred Asian Houbara Chlamydotis macqueenii

For migratory species, the success of population reintroduction or reinforcement through captive‐bred released individuals depends on survivors undertaking appropriate migrations. We assess whether captive‐bred Asian Houbara Chlamydotis macqueenii from a breeding programme established with locally sourced individuals and released into suitable habitat during spring or summer undertake similar migrations to those of wild birds. Using satellite telemetry, we compare the migrations of 29 captive‐bred juveniles, 10 wild juveniles and 39 wild adults (including three birds first tracked as juveniles), examining migratory propensity (proportion migrating), timing, direction, stopover duration and frequency, efficiency (route deviation), and wintering and breeding season locations. Captive‐bred birds initiated autumn migration an average of 20.6 (±4.6 se) days later and wintered 470.8 km (±76.4) closer to the breeding grounds, mainly in Turkmenistan, northern Iran and Afghanistan, than wild birds, which migrated 1217.8 km (±76.4), predominantly wintering in southern Iran and Pakistan (juveniles and adults were similar). Wintering locations of four surviving captive‐bred birds were similar in subsequent years (median distance to first wintering site = 70.8 km, range 6.56–221.6 km), suggesting that individual captive‐bred birds (but not necessarily their progeny) remain faithful to their first wintering latitude. The migratory performance of captive‐bred birds was otherwise similar to that of wild juveniles. Although the long‐term fitness consequences for captive‐bred birds establishing wintering sites at the northern edge of those occupied by wild birds remain to be quantified, it is clear that the pattern of wild migrations established by long‐term selection is not replicated. If the shorter migration distance of young captive‐bred birds has a physiological rather than a genetic basis, then their progeny may still exhibit wild‐type migration. However, as there is a considerable genetic component to migration, captive breeding management must respect migratory population structure as well as natal and release‐site fidelity.     

Migratory herds of wildebeests and zebras indirectly affect calf survival of giraffes

相似文献   

Variable individual consistency in timing and destination of winter migrating fish

Migration is an important event in the life history of many animals, but there is considerable variation within populations in the timing and final destination. Such differential migration at the population level can be strongly determined by individuals showing different consistencies in migratory traits. By tagging individual cyprinid fish with uniquely coded electronic tags, and recording their winter migrations from lakes to streams for 6 consecutive years, we obtained highly detailed long-term information on the differential migration patterns of individuals. We found that individual migrants showed consistent site fidelities for over-wintering streams over multiple migratory seasons and that they were also consistent in their seasonal timing of migration. Our data also suggest that consistency itself can be considered as an individual trait, with migrants that exhibit consistent site fidelity also showing consistency in migratory timing. The finding of a mixture of both consistent and inconsistent individuals within a population furthers our understanding of intrapopulation variability in migration strategies, and we hypothesize that environmental variation can maintain such different strategies.     

Fecal androgens of bison bulls during the rut

The influence of sex hormones is a key proximate factor underlying male reproductive behavior in mammals. Effective conservation policies for the remaining purebred plains bison (Bison bison bison) herds require knowledge of the physiology underlying bison reproductive biology. We used fecal steroid analysis to characterize androgen levels in adult bison bulls before, during, and after the rut, and to examine androgen levels of bulls differing in reproductive status, age, and mating success. Fieldwork was carried out at the Fort Niobrara National Wildlife Refuge in north-central Nebraska. All adult bison in the herd were individually known by unique brands. Fecal samples were collected during 2003 from bulls during pre-rut (June), rut (July-August), and post-rut (September), and behavioral observations focused on reproductive status and mating success during the rut. Matched sample data indicated that androgen levels (ng/g feces) of bulls peaked during the rut, doubling from pre-rut to rut and then declining by 75% during post-rut. Dominant bulls that tended (guarded) cows maintained higher androgen levels than bulls that were not tending. There was a positive correlation between bull age (associated with mating success) and androgens, with higher androgen levels in prime-aged bulls compared with younger bulls. Nonetheless, there was no correlation between mating success (measured by number of copulations observed) and androgen level. This suggests that while androgens may provide the proximate motivation to compete for matings, other factors determine the mating success of bison bulls.     

Marker genotypes and population admixture and their association with body weight, height and relative body mass in United States federal bison herds

Elucidating genetic influences on bison growth and body composition is of interest, not only because bison are important for historical, cultural, and agricultural reasons, but also because their unusual population history makes them valuable models for finding influential loci in both domestic cattle and humans. We tested for trait loci associated with body weight, height, and bison mass index (BMI) while controlling for estimated ancestry to reduce potential confounding effects due to population admixture in 1316 bison sampled from four U.S. herds. We used 60 microsatellite markers to model each phenotype as a function of herd, sex, age, marker genotypes, and individual ancestry estimates. Statistical significance for genotype and its interaction with ancestry was evaluated using the adaptive false discovery rate. Of the four herds, two appeared to be admixed and two were nonadmixed. Although none of the main effects of the loci were significant, estimated ancestry and its interaction with marker loci were significantly associated with the phenotypes, illustrating the importance of including ancestry in the models and the dependence of genotype-phenotype associations on background ancestry. Individual loci contributed approximately 2.0% of variation in weight, height, and BMI, which confirms the utility and potential importance of adjusting for population stratification.