High connectivity and minimal genetic structure among ......ing connectivity. Additionally, some authors have suggested that habitat specialists are more sensitive to connectivity

690

HIGH CONNECTIVITY AND MINIMAL GENETIC STRUCTURE AMONG NORTH AMERICAN BOREAL OWL (AEGOLIUS FUNEREUS)

POPULATIONS, REGARDLESS OF HABITAT MATRIXM rni E. Koopman,1 Gregory D. Hayward, and David B. McDonald

Department of Zoology and Physiology, University of Wyoming, Laramie, Wyoming 82071, USA

Abstract.—Habitat connectivity and corridors are o en assumed to be critical forthe persistence of patchily distributed populations, but empirical evidence for thisassumption is scarce. We assessed the importance of connectivity among habitatpatches for dispersal by a mature-forest obligate, the Boreal Owl (Aegolius funereus).Boreal Owls demonstrated a lack of genetic structure ( = 0.004 ± 0.002 [SE]) amongsubpopulations, regardless of matrix type and extent, which indicates that unfor-ested matrix does not act as a barrier to dispersal for this vagile species. We foundonly slightly higher genetic distances (Cavalli-Sforza chord distances ranged from0.015 to 0.025) among patchily distributed Rocky Mountain subpopulations as com-pared with largely contiguous boreal-forest subpopulations (0.013 to 0.019) and noevidence of a genetic split across the expansive high plains of Wyoming. Even themost isolated subalpine patches are connected via gene flow. As northern borealforests continue to experience intensive harvest of mature stands, geographic dis-persion of Boreal Owl habitat may begin to more closely resemble that found inthe Rocky Mountains. We suggest that decreased connectivity poses much less ofa threat to continued abundance of this mature-forest obligate than overall loss ofnesting and foraging habitat. Assessment of the importance of corridors and con-nectivity should be conducted on a species-by-species basis, given the variation inresponse of species to discontinuity of habitat, even among closely related taxa orguilds. Received 5 October 2005, accepted 22 June 2006.

Key words: Aegolius funereus, Boreal Owl, connectivity, corridors, dispersal, geneflow, genetic structure, microsatellites.

Alta Conectividad y Estructura Genética Mínima entre Poblaciones Norteamericanas deAegolius funereus, Independientemente de la Matriz del Hábitat

Res men.—Frecuentemente, se supone que la conectividad del hábitat y loscorredores son críticos para la persistencia de poblaciones distribuidas en parches,pero la evidencia empírica sobre esto es escasa. Evaluamos la importancia de laconectividad entre parches de hábitat para la dispersión en Aegolius funereus, unaespecie restringida a bosques maduros. Encontramos una ausencia de estructuragenética entre subpoblaciones ( = 0.004 ± 0.002 [EE]), independientemente deltipo de matriz y de su extensión, lo que sugiere que las matrices no boscosas noactúan como una barrera para la dispersión en esta especie de amplia movilidad.Sólo encontramos distancias genéticas ligeramente mayores (las distancias cuerdade Cavalli-Sforza estuvieron entre 0.015 y 0.025) entre subpoblaciones de lasMontañas Rocallosas distribuidas en parches en comparación con subpoblaciones

1Present address: Rocky Mountain Research Station, 240 West Prospect, Fort Collins, Colorado 80526, USA. E-mail: [email protected]

The Auk 124(2):690–704, 2007© The American Ornithologists’ Union, 2007. Printed in USA.

High Connectivity among Boreal Owl PopulationsApril 2007] 691

Population biologists have become increas-ingly interested in the spatial ecology of popu-lations, with particular focus on dispersal asone of the fundamental processes infl uencingpopulation dynamics (Walters 2000). Becauseof discontinuity of suitable habitat, most spe-cies exist in a patchy geographic distribution inall or part of their range, with dispersal amongpatches acting to connect the population as awhole. Dispersal infl uences species ranges, thesynchrony of population fl uctuations (Huitu etal. 2003), and long-term persistence of popu-lations locally and range-wide (Levins 1969, Stacey and Taper 1992, Martin et al. 2000). Populations may experience less variation inabundance and higher persistence because ofthe exchange of individuals among patchesthat vary in productivity (Lande 1988). Theidea that movement among subpopulationsa ects the persistence and dynamics of thebroader population is central to the concept ofmetapopulations (Hanski 1999), but dispersalmay be equally important in species that are notstructured as a metapopulation.

Models and theoretical understandingindicate that the nature of the matrix (i.e.,nonhabitat) between habitat patches can havefar-reaching e ects on populations (Gardneret al. 1991). Lower matrix quality and increasedresistance may decrease the likelihood of popu-lation persistence (Fahrig 2001, Vandermeerand Carvajal 2001). Increased theoreticalunderstanding of the e ects of patchy spatialstructure on such features as genetic diversity,dispersal rates, and extinction probabilities hasled to management that o en incorporates con-nectivity and corridors (Beier 1995, Dunninget al. 1995, Donnelly and Marzlu 2004), but

their e cacy su ers from a lack of empiricalstudy (Simberlo et al. 1992, Rosenberg et al.1997, Beier and Noss 1998, Berry et al. 2005), with the exception of a few well-documentedcases (Beier 1993, Dunning et al. 1995, Mechand Halle 2001). When dispersal through thematrix is su ciently high, increased habitatconnectivity may not increase population per-sistence or abundance (Hudgens and Haddad2003), and limited conservation resources maybe be er spent preserving or improving avail-able habitat rather than improving or maintain-ing connectivity. Additionally, some authorshave suggested that habitat specialists aremore sensitive to connectivity among habitatpatches than habitat generalists (Rosenberg etal. 1997, Haddad 1999a), especially when largedistances separate patches (Haddad 1999b).Understanding commonalities among guilds ortaxa, levels of connectivity among subpopula-tions, and resulting relationships is critical to athorough understanding of population dynam-ics with implications for management and con-servation (Kareiva 1990).

Here, we explore the e ects of habitat con-nectivity on movement among subpopulationsof a mobile mature-forest obligate, the BorealOwl (Aegolius funereus). Because of the geo-graphic dispersion of suitable forest habitat,Boreal Owls exhibit two distinctive distribu-tion pa erns, which makes them a ractive forinvestigating how connectivity a ects dispersalrates. In northern boreal forests, Boreal Owlsoccur throughout highly connected habitat,but in subalpine forest farther south, BorealOwls exist in isolated high-elevation patchesseparated by variable expanses of unsuitablematrix (Fig. 1; Hayward and Hayward 1993),

de bosques boreales contiguos (0.013 a 0.019), y no observamos evidencia deuna diferenciación genética a través de las amplias planicies altas de Wyoming.Incluso los parches subalpinos más aislados están conectados por fl ujo genético.A medida que los bosques boreales del norte continúen siendo sometidos aextracción intensiva de rodales maduros, la dispersión geográfica del hábitat deA. funereus en la región podría comenzar a semejarse más a la del hábitat de lasMontañas Rocallosas. Sugerimos que la conectividad reducida representa unaamenaza mucho menor sobre la abundancia de esta especie restringida a losbosques maduros que la pérdida general de hábitat de nidificación y forrajeo.Las evaluaciones de la importancia de los corredores y la conectividad debenrealizarse especie por especie, dada la variación en la respuesta de las especies ala discontinuidad del hábitat, aún entre taxones estrechamente emparentados opertenecientes al mismo gremio.

Koopman, Hayward, and McDonald692 [Auk, Vol. 124

including lower-elevation forest, prairie, desert,and urban development.

In addition to the contrasting pa erns of dis-persion, Boreal Owls in northern boreal forestsand subalpine forests di er in ecology, behav-ior, and even basic life-history traits (Hayward1997). Given their long dispersal distances andirruptive behavior in northern boreal forests(Löfgren et al. 1986, Korpimäki et al. 1987,Sonerud et al. 1988), Boreal Owls can be consid-ered highly vagile. However, Boreal Owls havea Holarctic distribution (Hayward and Hayward1993), and data on dispersal come from north-ern boreal forests in Fennoscandia. Subalpinepopulations in the Rocky Mountains of theUnited States are patchy, and barriers couldprevent significant exchange among patches.

For example, the Northern Spo ed Owl (Strixoccidentalis caurina) requires corridors of matureforest to facilitate dispersal from one habitatpatch to another (Miller 1989, Forsman et al.2002). If Boreal Owls, which are also mature-forest obligates (Hayward 1997), required suchdispersal corridors, we would expect a negativerelationship between the extent of unforestedmatrix and gene flow.

Although dispersal is crucial for a variety ofecological functions, the di culty associatedwith estimation of dispersal rates has contrib-uted to poor understanding of this vital compo-nent of population ecology (Koenig et al. 1996, Walters 2000). We used microsatellite DNA markers to address the question of how matrixcomposition, including both type and extent,

Fig. 1. The North American range of Boreal Owls, based on the distribution of boreal and subal-pine forest. Sample sizes and expected and observed heterozygosities are listed for each samplinglocality. Despite a potential metapopulation structure based on patchily distributed habitat in theRocky Mountains of the United States, long-distance dispersal limits genetic differentiation amongsubpopulations.


a ects movement among habitat patches. Incontrast to direct methods such as radiotelem-etry or satellite telemetry, genetic markers allowinvestigation into movement pa erns of smallanimals at a continental scale. Furthermore,genetic approaches facilitate detection ofrelatively low movement rates that may not bedetected by radiotelemetry or banding e orts.Molecular methods are not only becoming moreeconomical than traditional field methods,they also may provide more informative dataon dispersal, because only successful dispers-ers incorporate their genetic signature into thepopulation (Koenig et al. 1996). With a varietyof new statistical methods available to assessdispersal rates and pa erns using molecularmarkers (Paetkau et al. 1995, Cornuet et al. 1999, Pritchard et al. 2000, Beerli and Felsenstein 2001, Goudet et al. 2002), biologists are discoveringpa erns of dispersal that were unexpected onthe basis of field data alone (Scribner et al. 2001, Kerth et al. 2002, Korfanta et al. 2005).

We studied gene flow among subpopula-tions of Boreal Owls separated by a spectrum ofmatrix extent and type to determine the limitsto dispersal by a vagile mature-forest obligate.Specifically, we hypothesized that patchy sub-populations in the Rocky Mountains, thoughgeographically proximate, would show moregenetic structure (higher FST values and geneticdistances) than northern boreal-forest sub-populations that are geographically distant buthighly connected. Because dispersal is a func-tion of the extent and resistance of the matrixas well as the vagility of a species, we expectedto find that forested matrix allowed more geneflow than unforested matrix. We hypothesizedthat the large expanse of high plains acrossWyoming would act as a barrier to dispersal forBoreal Owls, as it has for a number of speciesof montane mammals (Findley and Anderson1956). The genetic signature of such a barrierwould appear as a departure from the nullmodel of isolation-by-distance (genetic distanceamong subpopulations increasing linearly withgeographic distance), resulting in a dramaticincrease in genetic distance in the presence ofinhospitable matrix (Paetkau et al. 1997).

We also expected that subpopulations ofBoreal Owls in the Rocky Mountains wouldexhibit classical metapopulation structure. Theyfit many assumptions of the metapopulationconcept, including discrete local populations

separated by inhospitable matrix and seem-ingly independent population dynamics. Oneassumption of the original metapopulationconcept (Levins 1969) is that “the exchangerate of individuals among local populations isso low that migration has no real e ect on localdynamics in the existing populations” (Hanskiand Simberlo 1997:9). Given the extensive tree-less matrix separating Rocky Mountain subpop-ulations, it seemed likely that dispersal (geneflow) would be limited, allowing subpopulationdi erentiation. Because much of metapopula-tion theory has developed with li le empiricalsupport from vertebrate studies, the potential totest metapopulation theory in such a system isvery a ractive, especially given that connectedsubpopulations in northern boreal forests allowfor comparison to a baseline level of gene flowacross largely continuous suitable breedinghabitat.

Methods

Sample collection and molecular genetic methods.—We sampled Boreal Owls in sub-populations separated by a spectrum of dis-tances and matrix types (Table 1). By samplingthe range of connectivity available to NorthAmerican Boreal Owls, we were able to assesshow matrix composition a ects gene flowamong subpopulations. We sampled birds atsites separated by habitat (boreal or subalpineforest), and by unsuitable matrix, includinglower-elevation forest (e.g., ponderosa pine orDouglas fir), urban development, and treelessexpanse (e.g., prairie, desert, shrublands). Here,we use the term “subpopulation” loosely, torefer to an area where we sampled Boreal Owls,rather than to a biological subpopulation.

Boreal Owls were captured primarily at nestboxes along logging roads on several nationalforests where an extensive system of nest boxeswas established beginning in 1987 (Haywardet al. 1992). More than 2,000 nest boxes werechecked each year, from 1998 to 2002. Samplesfrom Idaho were collected between 1995 and2002. On average, nest box use by Boreal Owlswas only ~1%. Adult females were trappedwhilebrooding or incubating, and males were trappedas they brought food to the chicks. We collectedblood samples from all individuals captured. If we were unable to trap adults at a nest, we col-lected a blood sample from one of the nestlings.


Boreal Owls were tagged with a federal band sothat we could recognize family members andrecaptures. Tissue samples obtained from BorealOwl specimens from Manitoba and Minnesotaconsisted of heart or muscle tissue. Additionaltissue samples collected near Fairbanks, Alaska,were obtained from the University of Alaskamuseum. Blood was stored in Longmire’sSolution (Longmire et al. 1988), and most tissuewas stored in 100% ethanol.

To isolate DNA from samples, we used aSigma GenElute mammalian DNA extraction kit(Sigma-Aldrich, St. Louis, Missouri). We geno-typed 275 unrelated individuals using sevenpolymorphic microsatellite loci, following theprotocol described by Koopman et al. (2004).

Genetic structure.—We assessed genetic struc-ture of subpopulations of Boreal Owls in NorthAmerica. Because microsatellite mutation pro-cesses are not fully understood at this time andthere is no consensus on the most appropriatemeasurements to use (Goldsteinand Pollock 1994, Ruzzante 1998, Foulley and Hill 1999, Kalinowski2002), the use of a variety of measurements, withdi erent underlying assumptions, can increaseconfidence in the results, especially if they agree(Neigel 2002). Some tests or measures assess dif-ferences among user-defined “subpopulations”(FST, measures of genetic distance), whereas

others assess di erences (or similarities) amongindividuals, thereby allowing for identificationof subpopulations based on genetic substruc-ture (assignment tests, allele-sharing distances,model-based clustering method of the Bayesianprogram STRUCTURE).

We tested for departure from Hardy-Weinberg equilibrium, within and amongeach pair of loci, using GENEPOP, version 3.3 (Raymond and Rousset 1995) and, for genotypiclinkage disequilibrium, FSTAT, version 2.9.3.2 (Goudet 1995). We used sequential Bonferroniprocedures to adjust for multiple comparisons(overall = 0.05).

We estimated genetic di erentiation amongsubpopulations, (equivalent to FST) using themeasure of Weir and Cockerham (1984), whichis weighted by sample size, in FSTAT. Ninety-five percent confidence intervals around Weirand Cockerham’s FST (herea er referred tosimply as FST) were estimated with 10,000 boot-strap replicates. In GENEPOP, we ran a G-likeexact test (Goudet et al. 1996) to assess di er-ences among subpopulations in overall allelicdistributions.

To evaluate the appropriate geographic scalefor population-di erentiation analysis, weused hierarchical F statistics (Weir 1996), whichinvolve grouping of individual subpopulations,

Table 1. Pairwise comparisons of subpopulations of Boreal Owls in the boreal forest and in the RockyMountains showing the range of matrix types and geographic distances between subpopulations.We sampled subpopulations with varying types of dominant matrix between them, includingsuitable breeding habitat (boreal forest), montane forest (largely connected lower-elevation forest),patchy forest (disconnected subalpine and lower-elevation forest with interspersed grasslands),high plains, and urban development. Sample locations are shown in Figure 1.

Distance betweenSubpopulations subpopulationsbeing compared (km) Dominant matrix typesW. CO / S. CO 166.6 Patchy forestID / MT 175.9 Montane forestWY / W. CO 207.0 Patchy forest, urban developmentWY / S. CO 368.6 Patchy forest, urban developmentFAIR /ANCH 415.2 Boreal forestMT / WY 781.6 High plains, patchy forestMT / W. CO 881.3 High plains, patchy forestID / WY 893.7 High plains, patchy forestID / W. CO 961.8 High plains, patchy forestMT / S. CO 1,020.2 High plains, patchy forestID / S. CO 1,083.8 High plains, patchy forestFAIR / CAN 3,338.4 Boreal forestANCH / CAN 3,407.9 Boreal forest


on the basis of geographic proximity, until overallgenetic di erentiation is maximized. Using thegrouping with highest genetic structure (great-est FST), we tested for genetic subdivision on aregional scale using a likelihood-based assign-ment test (Paetkau et al. 1995) in DOH (seeAcknowledgments) and a Bayesian assignmenttest (Cornuet et al. 1999) in GENECLASS (seeAcknowledgments). When populations show suf-ficient genetic di erentiation (FST > 0.05; Cornuetet al. 1999), this procedure allows identificationof individuals that may have dispersed betweenpopulations (Rannala and Mountain 1997). Wealso assessed structure among subpopulationsor geographic regions using STRUCTURE (seeAcknowledgments), which determines whethersampled genotypes are substructured into mul-tiple (K > 1) clusters or whether they constitute asingle genetically homogeneous population (K =1) in Hardy-Weinberg equilibrium. We tested forone to eight separate subpopulations withoutprior information on capture location of individu-als (Pritchard et al. 2000). Burn-in and replicationvalues were set at 25,000 and 1,025,000.

Phylogenetic trees.—We calculated pairwiseCavalli-Sforza chord distances (Cavalli-Sforzaand Edwards 1967) among subpopulationsand generated a rooted neighbor-joining treein the NEIGHBOR subroutine of PHYLIP (seeAcknowledgments), with Norwegian Teng-malm’s Owls (A. f. funereus) as an outgroup.Under microsatellite locus and sample numberconditions similar to ours, chord distances showgreater success in generating the correct treetopology than other distance measurements(Takezaki and Nei 1996). One-thousand boot-strap replications were performed to calculatepercentage of support for individual nodes. A maximum-likelihood (ML) tree was also con-structed using the ML subroutine (CONTML) in PHYLIP and bootstrapped 1,000 times, withTengmalm’s Owls as an outgroup.

We calculated allele-sharing distances(Bowcock et al. 1994) using the MS TOOLS add-in for EXCEL (see Acknowledgments). Pairwiseallele-sharing distances are calculated as oneminus half the average number of shared allelesper locus. Finally, we constructed a neighbor-joining tree (Saitou and Nei 1987) in PHYLIP, using unrelated North American individuals asthe operational taxonomic units.

Isolation-by-distance.—We compared matri-ces of pairwise genetic distance and pairwise

geographic distance using a Mantel test (Mantel1967) and 5,000 permutations in MCMANTEL (McDonald et al. 1999). We calculated straight-line distances between subpopulations inARCVIEW (ESRI, Redlands, California), usingthe center of the area where the most BorealOwls were captured as the end points of eachline connecting two subpopulations. Becausemost individuals were captured near Winnipegfor our samples from Canada, we used Winnipegas the endpoint for our Canadian comparisons.

We hypothesized that, if matrix type regu-lates gene flow among subpopulations, weshould see a departure from the linear isolation-by-distance model (Paetkau et al. 1997), suchthat subpopulations separated by inhospitablematrix would have higher genetic distance thanexpected from geographic distance alone.

We used three di erent pairwise measuresof genetic distance. First, we calculated pair-wise Cavalli-Sforza and Edwards (1967) chorddistance because of its superior performancein phylogenetic tree-building (Takezaki andNei 1996) and its linear nature over large dis-tances, and because it makes no assumptionsabout mutation models. We also calculatedthe ratio FST/(1 – FST) (Rousset 1997) and Nei’sstandard distance (Nei 1972), both of whichhave been shown to accurately reflect isolation-by-distance (Takezaki and Nei 1996, Paetkauet al. 1997, Rousset 1997), though FST/(1 – FST)may be accurate only over small geographicdistances (Rousset 1997). We decided not to usemicrosatellite-specific measurements becauseof departures from a strict stepwise mutationmodel (SMM) apparent in some of our loci andbecause other studies have found that the highvariance associated with these measurementsobscures pa erns (Paetkau et al. 1997), espe-cially with high levels of gene flow (Balloux andGoudet 2002).

Results

The 275 genotyped individuals were fromeight subpopulations (Fig. 1). The number ofalleles per locus ranged from 3 to 11 amongseven microsatellite loci. We found evidence forheterozygote deficiency at one locus in the sub-population from Canada (P < 0.001). Otherwise,all subpopulations were in Hardy-Weinbergequilibrium for all loci. We found no evidencefor genotypic disequilibrium among paired loci


(P values ranged from 0.01 to 0.96; adjusted 5% nominal level for multiple comparisons = 0.0025).

Genetic structure.—Genetic di erentiationamong subpopulations of Boreal Owls through-out North America was extremely small, asdemonstrated by estimates of global FST ( = 0.004 ± 0.002 [SE]; 95% CI: 0.000 to 0.008). Wefound no significant overall di erence in allelicdistribution among subpopulations (G-liketest, P = 0.794, df = 14). Using STRUCTURE, we consistently found a higher log likelihoodof one subpopulation (–3,801.3) rather than twoto eight subpopulations (–4,004.2 to –6,699.1), which indicates that, throughout the NorthAmerican range, Boreal Owls occur as a singlegenetically homogeneous population.

When we grouped all northern (Alaskaand Canada) and southern (Colorado andWyoming) subpopulations while excludingcentral subpopulations (Montana and Idaho),slightly more genetic structure was apparent.The magnitude of this structure, however, wasstill very small ( = 0.012 ± 0.003 [SE]; 95% CI: 0.007 to 0.017). We conducted the assignmenttest by reclassifying our samples into twogroups: north and south, excluding Idaho andMontana. However, regardless of assignmentalgorithm, few individuals were correctlyassigned to their origin of capture, because oflow levels of genetic di erentiation betweennorthern and southern subpopulations (Fig. 2). When we used the assignment test of Paetkauet al. (1995), only 65% of individuals (133 of206) were correctly assigned to their sampledsubpopulation. On the basis of chance alone,one would expect 50% of individuals to be cor-rectly assigned. Under the Bayesian method ofCornuet et al. (1999), only one individual of 206 had a significantly higher likelihood of havingoriginated in the subpopulation from which itwas sampled than in the other subpopulation.

Phylogenetic trees.—A phylogenetic tree forthe eight North American subpopulationsrecovered the geographic split between north-ern and southern subpopulations of BorealOwls (Fig. 3), but bootstrap support for mostclades was extremely low. Rocky Mountainsubdivisions had higher bootstrap support thanother clades. Results from the ML tree corrobo-rated those from the neighbor-joining tree. Ourneighbor-joining tree of allele-sharing distancesbetween individuals showed no clusteringbased on capture location (Fig. 4).

Isolation-by-distance.—Cavalli-Sforza chorddistances were extremely small, ranging from0.013 to 0.019 among boreal forest subpopula-tions and 0.015 to 0.025 among Rocky Mountainsubpopulations. Similarly, Nei’s standarddistances ranged from 0.006 to 0.020 amongboreal forest subpopulations and from 0.006 to0.030 among Rocky Mountain subpopulations.Pairwise comparisons of geographic distanceamong all subpopulations were significantlycorrelated with Cavalli-Sforza chord distances(R = 0.559, P = 0.001), but not with FST/(1 – FST)(R = 0.143, P = 0.238) or Nei’s standard distance(R = 0.116, P = 0.280). By contrast, all threegenetic distances were significantly correlatedwith geographic distances when we assessedonly the Rocky Mountain subpopulations (Rranged from 0.752 to 0.847; P < 0.008; Fig. 5). Nei’s standard genetic distance had the best fit,but the relationship was confounded by the factthat larger distances between subpopulations inthe Rockies were correlated with treeless matrix(Fig. 5).

Fig. 2. Assignment likelihoods for 206 individ-uals captured in northern (Alaska and Canada)and southern (Colorado and Wyoming) sub-populations of Boreal Owls. If genotype were agood indicator of origin, most individuals fromthe north would have fallen well above the line,whereas those from the south would have fallenwell below the line. The obvious lack of patternand proximity to the line of equal likelihood bothsuggest a lack of clear genetic differentiation.


Fig. 5. Pairwise comparisons of geneticdistance and geographic distance. Althoughgenetic distances were very small, we found aslight linear increase in genetic distance withgeographic distance among Rocky Mountainsubpopulations (filled symbols) but no increaseamong subpopulations separated by boreal-forest habitat (open squares). Although theslope of the line appears to be quite steep, thetotal change in genetic distance is only 0.03; forother published studies, Nei’s distances may bean order of magnitude higher (McDonald etal. 1999). The relationship between geographicdistance and genetic distance was confoundedby matrix type in the Rocky Mountains, wheretreeless matrix was associated with longer dis-tances; but even across treeless matrix, geneticdistances were very small.

Fig. 3. Neighbor-joining tree of Cavalli-Sforza chord distances among subpopulations of NorthAmerican Boreal Owls, with the Norwegian subspecies included as an outgroup. Percentageof support for each node was calculated from 1,000 trees built from bootstrapped data. Nodeswith >50% support (marked with an asterisk) include the node joining Canada, Anchorage, andFairbanks (54% support); that joining western Colorado, southern Colorado, and Wyoming (66%); and that joining western Colorado and Wyoming (69%). Higher bootstrap support for southernRocky Mountain clades may indicate slightly less gene flow because of patchy habitat.

Fig. 4. Unrooted neighbor-joining tree ofallele-sharing distances among 250 unrelatedBoreal Owls captured in North America. If subpopulations were genetically well differenti-ated, individuals sampled from the same localitywould cluster together. The obvious lack of clus-tering indicates a lack of genetic structure amongBoreal Owl subpopulations in North America. Bycontrast, including Old World subpopulationsproduces nearly total reciprocal monophylyamong Old World and New World populations,even when using individuals as the operationaltaxonomic units (Koopman et al. 2005).


Discussion

Many species exhibit naturally patchy distri-butions, and even more are becoming patchilydistributed because of habitat loss. Resourcemanagers require improved understanding ofdispersal, resulting spatial pa erns, and relation-ships to population persistence and demography.Empirical evidence for the e cacy of connectiv-ity and corridors is scarce, but a few cases clearlydemonstrate the importance of connectivity andfavorable matrix in facilitating dispersal amonghabitat patches (Beier 1993, Dunning et al. 1995, Berry et al. 2005). Boreal Owls, on the other hand,exhibit only a modest increase in genetic struc-ture when habitat patches are more isolated (Fig.3). Dispersal rates are high evenwhen patches areseparated by inhospitable matrix, long distances,or both. Given their dependence on mature for-est for foraging and nesting (Hayward 1997), weexpected that Boreal Owls would require conti-nuity of forested habitat to traverse the matrix.Our results reveal that generalizations aboutmovement rates based on closely related taxa(e.g., Spo ed Owls) or guilds (e.g., mature-forestobligates) are not reliable, and that assessment ofthe benefits of connectivity needs to be done on aspecies-by-species basis. If dispersal rates amongsubpopulations are high, managing matrix com-position for connectivity among subpopulationsmay waste scarce conservation resources.

Are Boreal Owls a metapopulation?—With anFST of 0.004, Boreal Owls in North America arenot partitioned into distinct subpopulations.Despite the patchy distribution of spruce–firforest throughout the Rocky Mountains anddespite strong dependence of Boreal Owls onmature forest, our genetic analysis indicates thatBoreal Owl subpopulations do not constitute ametapopulation. High rates of gene flow amongsubpopulations of Boreal Owls make it unlikelythat local populations exhibit independent pop-ulation dynamics or that recruitment is almostinvariably local, some of the fundamental tenetsof classical metapopulation theory (Harrisonand Taylor 1997). Various measurements (bothclassic ones, such as FST and Nei’s standarddistance, as well as more recently developedBayesian and likelihood-based measurements)painted similar pictures of lack of distinctgenetic structure among subpopulations. A lack of genetic structure among subpopulationsindicates that this resident but highly vagile owl

frequently disperses long distances over inhos-pitable habitat, even >200 km across the highplains of Wyoming. We caution against apply-ing the term “metapopulation” on the basis ofphysical patchiness of habitat only (Major et al.1999, Martin et al. 2000, Sweanor et al. 2000).

Boreal Owls have extremely large homeranges for birds their size (Hayward et al. 1993), and in boreal forest, they are known to movegreat distances during natal dispersal and win-ter irruptions (Löfgren et al. 1986, Korpimäki etal. 1987). In addition, females find new mateseach year, sometimes in new home ranges(Hayward et al. 1993). Over their lifetimes, totalarea traversed may span hundreds to thousandsof square kilometers.

Habitat connectivity and Boreal Owls.— Geneticdistances among sites with forested and treelessmatrix di ered minimally, leading us to con-clude that dispersal rates are high, regardless ofmatrix type and extent. In the Rocky Mountains,all three measures of genetic distance increasedlinearly, if slightly, with geographic distance,as expected under the island model with nobarriers to dispersal (Fig. 5). However, greaterdistances between Rocky Mountain subpopula-tions were correlated with treeless matrix, mak-ing it di cult to determine whether matrix typeor distance was the primary factor involvedin the significant relationship. In either case,gene flow was su ciently high, even betweenthe most distant and disconnected patches, tohomogenize subpopulations genetically. Forexample, the assignment test failed to assignindividuals to their population of origin,STRUCTURE indicated that Boreal Owls con-stitute a single population, and the individual-based neighbor-joining tree (Fig. 4) showedtotal lack of clustering of individuals sampledat the same locale.

Our results demonstrate that Boreal Owlsdisperse across large areas of unsuitable habi-tat and that no North American subpopulationis genetically isolated from the others. Thesehigh rates of gene flow make it unlikely thatsubpopulations are demographically indepen-dent. High dispersal rates likely act to dampenpopulation fl uctuations and boost breedingsuccess in population sinks (the “rescue e ect”;Brown and Kodric-Brown 1977), synchronizedemographic pa erns among subpopulations(Huitu et al. 2003), and overwhelm adapta-tions to local conditions. Some subpopulations


in the Rocky Mountains are less productivethan others (M. E. Koopman unpubl. data) andmay depend on high levels of immigration forlong-term persistence. Nevertheless, the time-scale of genetic homogenization (on the orderof several to many generations) may overlookdemographic pa erns that occur on a muchshorter scale of a few years. Although geneticsubstructuring clearly suggests demographicindependence among subpopulations, a lack ofgenetic structure does not necessarily precludea degree of demographic independence. Thus,the relationship between gene flow and demo-graphic pa erns invites further investigation.

Boreal vs. subalpine subpopulations.—Di erences in climate, habitat structure, preycycles, and prey composition between northernboreal forests and more southerly subalpineforests appear to drive behavioral and ecologi-cal di erences between northern and southernsubpopulations of Boreal Owls (Hayward 1997). Judging from these di erences in broad-scaledynamics, we expected high connectivity in thenorth and classical metapopulation structurein the southern parts of the range. We foundslightly lower values of genetic distance amongwidely separated boreal-forest subpopulationsthan among proximate Rocky Mountain sub-populations (Fig. 5), which indicates that patch-iness of habitat may slow movement. Similarly,we found higher bootstrap support for subdivi-sions among Rocky Mountain subpopulationsthan for subdivisions among northern subpop-ulations in our neighbor-joining and ML trees.The di erences in genetic distance were slight,however, and were overshadowed by a consis-tent lack of genetic structure under global FST,assignment tests, a G-like test, STRUCTURE, and allele-sharing distances.

Our boreal-forest samples were from twobreeding subpopulations in Alaska, plus irrup-tive individuals in northern Minnesota andsouthern Manitoba (labeled “CAN” in Fig. 1). Our samples from Canada represented indi-viduals that moved south from a wider breed-ing range farther north because of severe winterconditions. No genetic di erentiation existedbetween our samples from Alaska and Canada,and we feel confident that this is representativeacross the boreal forest of North America.Similarly, no genetic di erentiation existedamong Boreal Owls sampled in far eastern andfar western locations in the Eurasian boreal

forest (Koopman et al. 2005). By contrast, OldWorld and New World populations showed ahigh degree of di erentiation ( = 0.37; Koopmanet al. 2005), demonstrating that a threshold levelexists, at least at intercontinental scales.

Genetic di erentiation.—The level of geneticdi erentiation in our study ( = 0.004) waslower than that found, using microsatellitemarkers, among subpopulations of otheravian species (0.027 for Greater Sage-Grouse[Centrocercus urophasianus], Oyler-McCance etal. 1999; 0.014 for Yellow Warbler [Dendroica petechia], Gibbs et al. 2000; 0.014 for BurrowingOwl [Athene cunicularia], Korfanta 2001; 0.02 for Song Sparrow [Melospiza melodia], Chanand Arcese 2002), especially consideringthat our study was conducted over a largerspatial extent than most of the others. Eventhough genetic distances were small, wefound evidence of limited genetic subdivisionin the neighbor-joining and ML trees, whichrevealed two clades among North AmericanBoreal Owls. The southern clade showed >50% support for subdivisions in Colorado andWyoming. Hierarchical F statistics lent supportto the north versus south split, with Montanaand Idaho acting as middle ground betweenthe two clades. The fact that slight genetic sub-division is apparent only at a continent-widescale does not support our hypothesis thatthe treeless sagebrush steppe of the WyomingBasin acts as a physical barrier to dispersal forBoreal Owls. By contrast, many mammalianspruce–fir forest obligates have distinct north-ern and southeastern subspecies or are limitedin their range by the Wyoming Basin (Findleyand Anderson 1956).

Boreal-forest subpopulations of Boreal Owlsundergo irruptions, or mass southward move-ments of individuals, during extreme condi-tions (Hayward and Hayward 1993). Subalpinesubpopulations in the Rocky Mountains donot, possibly explaining their slightly greatergenetic di erentiation. Winter irruptions maydrive waves of immigrants from northern borealforests into southern subalpine forests, therebylargely overwhelming any local adaptations orgenetic structuring.

Implications for Boreal Owl management in sub-alpine and boreal forests.—Breeding populationsof Boreal Owls throughout inland mountainranges of the western United States were notdetected until the mid- to late 1980s (Hayward


et al. 1987). Since then, Boreal Owls have beenregarded as isolated mountain-top dwellersthat are rarely heard or seen. Subpopulationsare o en managed at the geographic scale ofindividual national forests, and local subpopu-lations likely remain undiscovered in certainregions. The response of Boreal Owls to forest-management practices and large-scale habitatalterations is, therefore, virtually unknown.

In light of earlier understanding, the pres-ent study provides managers with a moreoptimistic scenario for long-term persis-tence of Boreal Owls, especially in the RockyMountains, where individual subpopulationsare smaller and potentially more vulnerable toextinction. Because Boreal Owls appear to bestructured not as a metapopulation, but insteadas a well-connected yet patchily distributedpopulation, consideration of connectivity andmatrix composition is not as critical for man-agement as it would be under a classic meta-population structure. Additionally, temporaryextinctions of local populations resulting fromlarge-scale natural disturbances and extensivetimber harvest, both of which we have observed(M. E. Koopman et al. unpubl. data), are likelyto be followed by recolonization when maturespruce–fir habitat is restored, even when thenearest extant subpopulation is distant or sepa-rated by treeless matrix. With this knowledge,managers can focus conservation resources onother aspects of Boreal Owl life history, suchas managing large tracts of mature spruce–firforest habitat to sustain foraging and nestingrequirements (Hayward 1997).

We have referred to the northern boreal forestas a large swath of connected habitat hospitableto Boreal Owls. Indeed, Boreal Owls are quitenumerous throughout the region. Because ofintensive logging pressure in the boreal forest,however, Boreal Owls in this region may beginto more closely resemble Rocky Mountain sub-populations as large tracts of habitat are lostand remaining tracts become disconnected. Ona positive note, discontinuity of boreal forestmay not significantly increase extinction proba-bilities because of high rates of dispersal amongsuitable habitat patches, as long as su cienttracts of mature forest continue to persist onthe landscape. We acknowledge, however, thata decrease in connectivity is only one of manydeleterious e ects of habitat fragmentation.Infl ux of invasive species, disease, competitors,

and predators (Bri ingham and Temple 1983, Wilcove 1985, Burke and Nol 1998, Brown andSullivan 2005) can act to degrade remaininghabitat patches and negatively a ect mature-forest obligates such as Boreal Owls. AlthoughBoreal Owls are currently numerous acrossmuch of their range, and their ability to disperseacross inhospitable matrix increases their prob-ability of persistence, their close ties to a quicklyvanishing habitat type continues to represent asignificant threat to the future abundance of thisspecies.

Ac nowledgments

More U.S. Forest Service (USFS) biologiststhan we can name here provided logistical sup-port in the field—thanks to all of them, especiallyF. Gordon, C. Hescock, S. Jacobson, J. Ormiston,and R. Skorkowsky. We are grateful to T. Bodreaux, B. Di rick, T. Holland, L. Moorehead,T. Swem, E. Taylor, and especially C. Schultz, forgenerously sending samples from Boreal Owls.Additional Boreal Owl genetic samples wereprovided by the Bell Museum, the AgriculturalUniversity of Norway, Manitoba Conservation,University of Alaska Museum, and the BurkeMuseum. We had excellent field assistance bymany dedicated biologists, including L. Ayers,J. Bassinger, J. Benne , J. Carpenedo, S. Dubay,T. Hampton, the Hayward family, T. Heekin,K. Ke er, S. Koopman, S. Mullins, K. O , M. Suedkamp,and P. Sutherland. J. Benne providedGIS support. This project was funded by GlobalForest (GF-18-2000-132), USFS Rocky MountainResearch Station, the Nansen Endowment(grant to G. A. Sonerud), and awards from theAmerican Museum of Natural History, Sigma Xi,and the Department of Zoology and Physiologyand the Institute of Environment and NaturalResources at the University of Wyoming. Insightand ideas provided by S. Jackson, J. Lovvorn,and S. Anderson are greatly appreciated. Thecomments of G. A. Sonerud and an anonymousreviewer greatly enhanced this manuscript. TheDOH assignment test calculator is available atwww2.biology.ualberta.ca/jbrzusto/Doh.php,GENECLASS at www.montpellier.inra.fr/URLB/geneclass/geneclass.html, PHYLIP atevolution.genetics.washington.edu/phylip.html,the MS TOOLS add-in for EXCEL atanimalgenomics.ucd.ie/sdepark/ms-toolkit/, andSTRUCTURE at pritch.bsd.uchicago.edu/.


Literature Cited

Balloux, F., and J. Goudet. 2002. Statisticalproperties of population di erentiationestimators under stepwise mutation in afinite island model. Molecular Ecology 11:771–783.

Beerli, P., and J. Felsenstein. 2001. Maximumlikelihood estimation of a migration matrixand e ective population sizes in n sub-populations by using a coalescent approach.Proceedings of the National Academy ofSciences USA 98:4563–4568.

Beier, P. 1993. Determining minimum habitatareas and habitat corridors for cougars.Conservation Biology 7:94–108.

Beier, P. 1995. Dispersal of juvenile cougarsin fragmented habitat. Journal of WildlifeManagement 59:228–237.

Beier, P., and R. F. Noss. 1998. Do habitat cor-ridors provide connectivity? ConservationBiology 12:1241–1252.

Berry, O., M. D. Tocher, D. M. Gleeson, and S. D. Sarre. 2005. E ect of vegetationmatrix on animal dispersal: Genetic evi-dence from a study of endangered skinks.Conservation Biology 19:855–864.

Bowcock, A. M., A. Ruiz-Linares, J. Tomfohrde,E. Minch, J. R. Kidd, and L. L. Cavalli-Sforza. 1994. High resolution of humanevolutionary trees with polymorphic micro-satellites. Nature 368:455–457.

Brittingham, M. C., and S. A. Temple. 1983. Have cowbirds caused forest songbirds todecline? BioScience 33:31–35.

Brown, J. H., and A. Kodric-Brown. 1977. Turnover rates in insular biogeography:E ect of immigration on extinction. Ecology58:445–449.

Brown, W. P., and P. J. Sullivan. 2005. Avian com-munity composition in isolated forest frag-ments: A conceptual revision. Oikos 111:1–8.

Burke, D. M., and E. Nol. 1998. Infl uence offood abundance, nest-site habitat, and forestfragmentation on breeding Ovenbirds. Auk 115:96–104.

Cavalli-Sforza, L. L., and A. W. F. Edwards. 1967. Phylogenetic analysis: Models andestimation procedures. American Journal ofHuman Genetics 19:233–257.

Chan, Y., and P. Arcese. 2002. Subspecificdi erentiation and conservation of SongSparrows (Melospiza melodia) in the San

Francisco Bay region inferred by microsatel-lite loci analysis. Auk 119:641–657.

Cornuet, J.-M., S. Piry, G. Luikart, A. Estoup, and M. Solignac. 1999. New methodsemploying multilocus genotypes to select orexclude populations as origins of individu-als. Genetics 153:1989–2000.

Donnelly, R., and J. M. Marzluff. 2004. Importance of reserve size and land-scape context to urban bird conservation.Conservation Biology 18:733–745.

Dunning, J. B., Jr., R. Borgella, Jr., K. Clements, and G. K. Meffe. 1995. Patch isolation, cor-ridor e ects, and colonization by a residentsparrow in a managed pine woodland.Conservation Biology 9:542–550.

Fahrig, L. 2001. How much habitat is enough?Biological Conservation 100:65–74.

Findley, J. S., and S. Anderson. 1956. Zoogeography of the montane mammals ofColorado. Journal of Mammalogy 37:80–82.

Forsman, E. D., R. G. Anthony, J. A. Reid, P. J. Loschl, S. G. Sovern, M. Taylor, B. L. Biswell, A. Ellingson, E. C. Meslow, G. S. Miller, and others. 2002. Natal and breed-ing dispersal of Northern Spo ed Owls.Wildlife Monographs, no. 149.

Foulley, J.-L., and W. G. Hill. 1999. On theprecision of estimation of genetic distance.Genetics, Selection, Evolution 31:457–464.

Gardner, R. H., M. G. Turner, R. V. O’Neill, and S. Lavorel. 1991. Simulation of thescale-dependent e ects of landscapeboundaries on species persistence and dis-persal. Pages 76–89 in Ecotones: The Role ofLandscape Boundaries in the Managementand Restoration of Changing Environments(M. M. Holland, P. G. Risser, and R. J. Naiman,Eds.). Chapman and Hall, New York.

Gibbs, H. L., R. J. G. Dawson, and K. A. Hobson. 2000. Limited di erentiation in microsatel-lite DNA variation among northern popula-tions of the Yellow Warbler: Evidence formale-biased gene flow? Molecular Ecology9:2137–2147.

Goldstein, D. B., and D. D. Pollock. 1994. Least squares estimation of moleculardistance—Noise abatement in phylogeneticreconstruction. Theoretical PopulationBiology 45:219–226.

Goudet, J. 1995. FSTAT (version 1.2): A com-puter program to calculate F-statistics.Journal of Heredity 86:485–486.


Goudet, J., N. Perrin, and P. Waser. 2002. Testsfor sex-biased dispersal using bi-paren-tally inherited genetic markers. MolecularEcology 11:1103–1114.

Goudet, J., M. Raymond, T. de Meeüs, and F. Rousset. 1996. Testing di erentiation in dip-loid populations. Genetics 144:1933–1940.

Haddad, N. M. 1999a. Corridor use predictedfrom behaviors at habitat boundaries.American Naturalist 153:215–227.

Haddad, N. M. 1999b. Corridor and distancee ects on interpatch movements: A land-scape experimentwith bu erflies. EcologicalApplications 9:612–622.

Hanski, I. 1999. Metapopulation Ecology.Oxford University Press, United Kingdom.

Hanski, I. A., and D. Simberloff. 1997. Themetapopulation approach, its history, con-ceptual domain, and application to con-servation. Pages 5–26 in MetapopulationBiology: Ecology, Genetics, and Evolution(I. A. Hanski and M. E. Gilpin, Eds.).Academic Press, San Diego, California.

Harrison, S., and A. D. Taylor. 1997. Empiricalevidence for metapopulation dynamics.Pages 27–42 in Metapopulation Biology:Ecology, Genetics, and Evolution (I. A. Hanski and M. E. Gilpin, Eds.). AcademicPress, San Diego, California.

Hayward, G. D. 1997. Forest management andconservationof Boreal Owls in North America.Journal of Raptor Research 31:114–124.

Hayward, G. D., and P. H. Hayward. 1993. BorealOwl (Aegolius funereus). In The Birds of NorthAmerica, no. 63 (A. Poole, P. Ste enheim, andF. Gill, Eds.). Academy of Natural Sciences,Philadelphia, and American Ornithologists’Union, Washington, D.C.

Hayward, G. D., P. H. Hayward, and E. O. Garton. 1993. Ecology of Boreal Owls in thenorthern Rocky Mountains, USA. WildlifeMonographs, no. 124.

Hayward, G. D., P. H. Hayward, E. O. Garton, and R. Escano. 1987. Revised breeding dis-tribution of the Boreal Owl in the northernRocky Mountains. Condor 89:431–432.

Hayward, G. D., R. K. Steinhorst, and P. H. Hayward. 1992. Monitoring Boreal Owlpopulations with nest boxes: Sample sizeand cost. Journal of Wildlife Management56:777–785.

Hudgens, B. R., and N. M. Haddad. 2003. Predicting which species will benefit from

corridors in fragmented landscapes frompopulation growth models. AmericanNaturalist 161:808–820.

Huitu, O., K. Norrdahl, and E. Korpim ki.2003. Landscape e ects on temporal andspatial properties of vole population fl uc-tuations. Oecologia 135:209–220.

Kalinowski, S. T. 2002. Evolutionary and statis-tical properties of three genetic distances.Molecular Ecology 11:1263–1273.

Kareiva, P. 1990. Population dynamics in spa-tially complex environments: Theory anddata. Philosophical Transactions of the RoyalSociety of London, Series B 330:175–190.

Kerth, G., F. Mayer, and E. Petit. 2002. Extremesex-biased dispersal in the communallybreeding, nonmigratory Bechstein’s bat(Myotis bechsteinii). Molecular Ecology 11:1491–1498.

Koenig, W. D., D. van Vuren, and P. N. Hooge. 1996. Detectability, philopatry, and thedistribution of dispersal distances in verte-brates. Trends in Ecology and Evolution 11:514–517.

Koopman, M. E. 2003. Genetic structure ofBoreal Owls. Ph.D. dissertation, Universityof Wyoming, Laramie.

Koopman, M. E., D. B. McDonald, G. D. Hayward, K. Eldegard, G. Sonerud, and S. G. Sermach. 2005. Genetic similarityamong Eurasian subspecies of Boreal Owls.Journal of Avian Biology 36:179–183.

Koopman, M. E., N. A. Schable, and T. C. Glenn. 2004. Development and optimiza-tion of microsatellite DNA primers forBoreal Owls (Aegolius funereus). MolecularEcology Notes 4:376–378.

Korfanta, N. M., D. B. McDonald, and T. C. Glenn. 2005. Burrowing Owl (Athene cunic-ularia) population genetics: A comparison ofNorth American forms and migratory hab-its. Auk 122:464–478.

Korpim ki, E., M. Lagerstr m, and P. Saurola.1987. Field evidence for nomadism inTengmalm’s Owl Aegolius funereus. OrnisScandinavica 18:1–4.

Lande, R. 1988. Genetics and demographyin biological conservation. Science 241:1455–1460.

Levins, R. 1969. The e ect of random variationsof di erent types on population growth.Proceedings of the National Academy ofSciences USA 62:1061–1065.


Löfgren, O., B. Hörnfeldt, and B.-G. Carlsson. 1986. Site tenacity and nomadism inTengmalm’s Owl (Aegolius funereus (L.)) inrelation to cyclic food production. Oecologia69:321–326.

Longmire, J. L., A. K. Lewis, N. C. Brown, L. M. Buckingham, L. M. Clark, M. D. Jones, L. J. Meincke, J. Meyne, R. L. Ratliff, F. A. Ray, and others. 1988. Isolation and molecularcharacterization of a highly polymorphiccentromeric tandem repeat in the familyFalconidae. Genomics 2:14–24.

Major, R. E., F. J. Christie, G. Gowing, and T. J. Ivison. 1999. Age structure and density ofRed-capped Robin populations vary withhabitat size and shape. Journal of AppliedEcology 36:901–908.

Mantel, N. 1967. The detection of diseaseclustering and a generalized regressionapproach. Cancer Research 27:209–220.

Martin, K., P. B. Stacey, and C. E. Braun. 2000. Recruitment, dispersal, and demographicrescue in spatially-structured White-tailedPtarmigan populations. Condor 102:503–516.

McDonald, D. B., W. K. Potts, J. W. Fitzpatrick,and G. E. Woolfenden. 1999. Contrastinggenetic structures in sister species of NorthAmerican scrub-jays. Proceedings of theRoyal Society of London, Series B 266:1117–-1125.

Mech, S. G., and J. G. Hallett. 2001. Evaluatingthe e ectiveness of corridors: A geneticapproach. Conservation Biology 15:467–474.

Miller, G. S. 1989. Dispersal of juvenileNorthern Spo ed Owls in western Oregon.M.S. thesis, Oregon State University,Corvallis.

Nei, M. 1972. Genetic distance between popula-tions. American Naturalist 106:283–292.

Neigel, J. E. 2002. Is FST obsolete? ConservationGenetics 3:167–173.

Oyler-McCance, S. J., N. W. Kahn, K. P. Burnham, C. E. Braun, and T. W. Quinn.1999. A population genetic comparisonof large and small-bodied Sage Grouse inColorado using microsatellite and mito-chondrial markers. Molecular Ecology 8:1457–1465.

Paetkau, D., W. Calvert, I. Stirling, and C. Strobeck. 1995. Microsatellite analysis ofpopulation structure in Canadian polarbears. Molecular Ecology 4:347–354.

Paetkau, D., L. P. Waits, P. L. Clarkson, L. Craighead, and C. Strobeck. 1997. Anempirical evaluation of genetic distancestatistics using microsatellite data frombear (Ursidae) populations. Genetics 147:1943–1957.

Pritchard, J. K., M. Stephens, and P. Donnelly. 2000. Inference of population structureusing multilocus genotype data. Genetics155:945–959.

Rannala, B., and J. L. Mountain. 1997. Detecting immigration by using multilocusgenotypes. Proceedings of the NationalAcademy of Sciences USA 94:9197–9201.

Raymond, M., and F. Rousset. 1995. An exact testfor population di erentiation. Evolution 49:1280–1283.

Rosenberg, D. K., B. R. Noon, and E. C. Meslow. 1997. Biological corridors: Form, function,and e cacy. BioScience 47:677–687.

Rousset, F. 1997. Genetic di erentiation andestimation of gene flow from F-statisticsunder isolation by distance. Genetics 145:1219–1228.

Ruzzante, D. E. 1998. A comparison of severalmeasures of genetic distance and popula-tion structure with microsatellite data: Biasand sampling variance. Canadian Journal ofFisheries and Aquatic Sciences 55:1–14.

Saitou, N., and M. Nei. 1987. The neighbor-joining method: A new method for recon-structing phylogenetic trees. MolecularBiology and Evolution 4:406–425.

Scribner, K. T., M. R. Petersen, R. L. Fields, S. L. Talbot, J. M. Pearce, and R. K. Chesser. 2001. Sex-biased gene flow in spec-tacled eiders (Anatidae): Inferences frommolecular markers with contrasting modesof inheritance. Evolution 55:2105–2115.

Simberloff, D., J. A. Farr, J. Cox, and D. W. Mehlman. 1992. Movement corridors:Conservation bargains or poor investments?Conservation Biology 6:493–504.

Sonerud, G. A., R. Solheim, and K. Prestud.1988. Dispersal of Tengmalm’s Owl Aegoliusfunereus in relation to prey availability andnesting success. Ornis Scandinavica 19:175–181.

Stacey, P. B., and M. Taper. 1992. Environmentalvariation and the persistence of small popu-lations. Ecological Applications 2:18–29.

Sweanor, L. L., K. A. Logan, and M. G. Hornocker. 2000. Cougar dispersal pat-


terns, metapopulation dynamics, and con-servation. Conservation Biology 14:798–808.

Takezaki, N., and M. Nei. 1996. Genetic dis-tances and reconstruction of phylogenetictrees from microsatellite DNA. Genetics144:389–-399.

Vandermeer, J., and R. Carvajal. 2001. Metapopulation dynamics and the qual-ity of the matrix. American Naturalist 158:211–220.

Walters, J. R. 2000. Dispersal behavior: An orni-thological frontier. Condor 102:479–481.

Weir, B. S. 1996. Genetic Data Analysis II: Methods for Discrete Population GeneticData. Sinauer Associates, Sunderland,Massachuse s.

Weir, B. S., and C. C. Cockerham. 1984. Estimating F-statistics for the analysis of pop-ulation structure. Evolution 38:1358–1370.

Wilcove, D. S. 1985. Nest predation in foresttracts and the decline of migratory song-birds. Ecology 66:1211–1214.

Associate Editor: K. Winker

Documents

High connectivity and minimal genetic structure among ......ing connectivity. Additionally, some authors have suggested that habitat specialists are more sensitive to connectivity