Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d 778 The Condor 112(4):778?788 ? the cooper ornithological Society 2010 The Condor, Vol. 112, number 4, pages 778?788. iSSn 0010-5422, electronic iSSn 1938-5422. ? 2010 by the cooper ornithological Society. all rights reserved. please direct all requests for permission to photocopy or reproduce article content through the university of california press?s Rights and permissions website, http://www.ucpressjournals.com/ reprintinfo.asp. doi: 10.1525/cond.2010.100146 Resumen. las poblaciones de Euphagus carolinus han disminuido dram?ticamente en toda su distribuci?n en am?rica del norte al menos desde la d?cada de los sesenta, pero las causas de esta disminuci?n se desconocen. Medimos los cocientes de is?topos estables de hidr?geno (?d) en plumas recolectadas entre 2005 y 2009 en el valle aluvial del Mississippi (n = 255 aves) y en las planicies costeras de carolina del Sur y Virginia (n = 281 aves) para estimar la regi?n de origen de las aves que pasaban el invierno al oeste y al este de las apalaches, respectivamente. tambi?n medimos los valores de ?d en plumas de todos los espec?menes coleccionados en esas dos regiones entre 1879 y 1990 que estaban disponibles (n = 190 aves). los valores isot?picos demuestran conectividad migratoria en esta especie: las poblaciones que se reproducen en el oeste y en bosques boreales centrales migran a lo largo de un corredor central o del Mississippi, mientras que las que se reproducen en los bosques boreales del este migran a un ?rea de invernada al este de los apalaches. Sin embargo, encontramos variabilidad temporal de corto plazo en el origen reproductivo de las aves que pasaron el invierno en las planicies costeras entre 2007 y 2009. la divisi?n entre dos rutas migratorias sugiere que los esfuerzos de manejo deben ajustarse al menos para las subpoblaciones del este y del oeste tanto en las ?reas de cr?a como en las de invernada. nuestro enfoque puede aplicarse a un rango amplio de especies migratorias en am?rica del norte y en otros continentes. MigRatoRy connectiVity in the RuSty BlackBiRd: iSotopic eVidence fRoM featheRS of hiStoRical and conteMpoRaRy SpeciMenS conectividad Migratoria en Euphagus carolinus: evidencia isot?pica de plumas de espec?menes hist?ricos y contempor?neos Migratory connectivity in the Rusty Blackbird keith a. hobson et al. Abstract. the Rusty Blackbird (Euphagus carolinus) has declined dramatically across its range in north america since at least the 1960s, but the causes for this decline are unknown. we measured ratios of stable hydro- gen isotopes (?d) in feathers collected from Rusty Blackbirds wintering in the Mississippi alluvial Valley (n = 255 birds) and the coastal plain of South carolina and Virginia (n = 281 birds), 2005?2009, to estimate the region of origin of birds wintering west and east of the appalachians, respectively. we also measured ?d values in feath- ers from all available museum specimens collected from 1879 to 1990 in these same two regions (n = 190 birds). isotopic values support migratory connectivity in this species with breeding populations in the western and cen- tral boreal forest migrating through a central or Mississippi flyway and those breeding in the eastern boreal forest migrating to a winter range east of the appalachians. we detected little long-term change in the breeding origins of modern and historical populations wintering east and west of the appalachians. however, we found short-term temporal variability in the breeding origins of birds wintering on the coastal plain from 2007 to 2009. the migra- tory divide suggests that efforts at management should be tailored to at least eastern and western subpopulations on both the breeding and wintering grounds. our approach can be applied to a broad range of migratory species in north america and on other continents. 4e-mail: keith.hobson@ec.gc.ca 5current address: School of natural Sciences and psychology, liverpool John Moores university, liverpool l3 3af, united kingdom. 0 0 0 0 ? the cooper ornithological Society, 2010 2010 Keith a. hobson1,4, Russell GReenbeRG2, steven l. van WilGenbuRG3, and Claudia MettKe-hofMann2,5 1Environment Canada, 11 Innovation Boulevard, Saskatoon, Saskatchewan S7N 3H5, Canada 2Smithsonian Migratory Bird Center, National Zoological Park, Washington, DC 20008 3Environment Canada, 115 Perimeter Road, Saskatoon, Saskatchewan S7N 0X4, Canada Key words: deuterium, euphagus carolinus, migratory connectivity, Rusty Blackbird, stable isotopes. intRoduction the effective conservation of migratory birds that travel an- nually between distant breeding, stopover, and wintering sites is made difficult by a general lack of knowledge of the connectivity between these sites for specific populations of interest (webster et al. 2002). this task is made even more difficult for those species that are distributed broadly but in low numbers over vast and remote regions. Species with broad distributions also present considerable challenges to the interpretation of global declines in populations. in such cases, knowledge of historic patterns of abundance and whether declines are disproportionate in one region versus another would assist in testing hypotheses regarding which 17_MS100146.indd 778 11/29/10 10:58:16 AM MigRatoRy connectiVity in the RuSty BlackBiRd 779 Spe c ia l Se c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d factors may have caused the decline. for broadly distrib- uted species, however, population censuses are often limited, nonexistent, or spatially biased. furthermore, conventional mark?recapture approaches are often poorly suited to estab- lishing patterns of migratory connectivity and, of course, cannot be used in retrospective studies unless the appropriate data were collected in the past. fortunately, the development of the use of endogenous markers to infer origins of migra- tory wildlife has provided extremely valuable tools to assist in current and retrospective studies. in particular, the mea- surement of the abundance of stable hydrogen isotopes (mea- sured as ?d, see Methods) in feathers of migratory birds and tissues of other wildlife has proven useful in making spatial connections among components of the annual cycle (hobson 2005, hobson and wassenaar 2008). in north america the Rusty Blackbird (Euphagus caroli- nus) is a classic example of a species whose population has de- clined catastrophically (greenberg and droege 1999) and for which interpretations of the causes of the decline are difficult because of a paucity of data from conventional mark?recapture approaches. this species has a large breeding range extending across the north american boreal forest from Maine to alaska, where it nests in wetlands that are usually difficult to access (avery 1995). the winter range is smaller, covering much of the eastern u.S. (fig. 1). within this outlined area, populations tend to be unstable, particularly in terms of how long they re- main in more northern areas and how they respond to year-to- year variation in weather (hamel and ozdenerol 2009). the Rusty Blackbird is the passerine most closely tied to boreal forest wetlands for breeding (erskine 1977, Spindler figuRe 1. Breeding and wintering ranges of the Rusty Blackbird (Euphagus carolinus). dark line indicates the approximate position of the divide between the atlantic coastal plain and Mississippi alluvial Valley drainages along the appalachian Mountains. 17_MS100146.indd 779 11/29/10 10:58:47 AM 780 keith a. hoBSon et al. Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d and kessel 1980). it nests near open water (kennard 1920) and feeds primarily on the adults and aquatic larvae of wet- land insects (ellison 1990). Several changes in the availability and quality of boreal wetlands may be contributing to the spe- cies? population decline (greenberg and droege 1999). dis- turbance regimes of wetlands and rates of population decline of blackbirds, however, do not appear to be uniform across the vast boreal region of north america. in the northwestern por- tion of the breeding range in alaska, boreal wetlands are dry- ing widely from climate warming (klein et al. 2005, Riordan et al. 2006) with shrinkage of wetlands resulting in changes in water chemistry, concentration of nutrients, and decreases in the abundance of macroinvertebrates (corcoran et al. 2009). Resource development is widespread or increasing along the southern edge of the boreal forest, and acid rain and mercury contamination have affected wetlands in the eastern part of the boreal forest (greenberg et al. 2011). anthropogenic disturbances in the boreal forest may be having broad effects, as several species breeding only in bo- real wetlands are among the most rapidly declining birds in the continent. in addition to the Rusty Blackbird, severe long- term declines have been documented across all or major por- tions of the breeding ranges of the co-occurring lesser Scaup (Aythya affinis), white-winged (Melanitta fusca) and Surf (M. perspicillata) Scoters, horned grebe (Podiceps auritus), lesser yellowlegs (Tringa flavipes), and Solitary Sandpiper (T. solitaria) (north american waterfowl Management plan 2004, Sauer et al. 2007, u.S. fish and wildlife Service 2009). these species migrate and winter in different habitats and ar- eas, so parsimony argues that events on the breeding grounds are at least in part responsible for their declines. if a similar suite of factors operating on the breeding grounds is contrib- uting to the declines of these species, the Rusty Blackbird may be particularly sensitive to its effects, as among these birds it has declined most precipitously (?12.5% per year range-wide, Sauer et al. 2007). on the wintering grounds, Rusty Blackbirds are strongly associated with wooded wetlands but use other forested hab- itats with mast (such as pecan orchards), partially flooded impoundments, and feedlots. although the species winters broadly throughout the eastern u.S., the greatest concentra- tions are in the Mississippi alluvial Valley and adjacent ar- eas to the north and east (niven et al. 2004), as well on as the atlantic coastal plain (particularly South carolina and geor- gia; hamel and ozdenerol 2009). Rusty Blackbirds appear to need relatively warm conditions for feeding on arthropods in the sod in and around temporary pools and continue to move south in november and well into december, probably to avoid frozen ground. threats to winter habitat include loss of for- ested wetlands to agriculture and perhaps loss of areas with intermittent flooding to water management (greenberg et al. 2011). at this point, it is not clear if habitat loss and degrada- tion have been more severe in the Mississippi alluvial Valley than along the coastal plain (hamel et al. 2009). Most of the hypothetical causes of decline should result in a distinct geographic pattern in the decline. But establish- ing these predicted patterns first requires an understanding of the connectivity of breeding and wintering populations. in this study we estimated the breeding origins of Rusty Black- birds wintering across the southeastern u.S. to (1) identify connectivity between breeding and wintering populations and (2) determine if the distribution of breeding origins may have changed historically and through the period of major decline. we did this by comparing measurements of stable hydrogen isotope ratios in feathers of museum specimens (1879?1990) and of wintering Rusty Blackbirds recently (2005?2009) cap- tured live. our intention was to identify regions where the change in breeding distribution is greatest in order to gain in- sight into possible mechanisms causing the species? decline. MethodS featheR SaMpling we obtained historical feather samples from Rusty Black- birds collected 1879?1990 in several museums (see acknowl- edgments). these feathers were removed from study skins by museum staff, and all collection information was noted (in- formation available from the senior author). detailed exami- nation of live birds captured in the field and specimens held in museums reveals that the species undergoes a pre-alternate molt restricted to feathers around the head and nape (Mettke- hofmann et al. 2010). our feather samples avoided these areas and instead included either flight feathers or buff or rusty-tipped contour feathers (feathers molted in winter lack these) from the flank or belly. we obtained modern samples by trapping birds in mist nets on their wintering grounds in (1) the Mississippi alluvial Valley near greenville, Mississippi (33? 29? 40? n, 90? 5? 5? w), near Johnson, arkansas (36? 8? 2? n, 94? 9? 57? w), and at tensas River national wildlife Refuge (nwR), louisiana (32? 9? 48? n, 91? 26? 51? w) be- tween 2005 and 2009 and (2), the atlantic coastal plain near donnelly wildlife Management area, South carolina (32? 40? 30? n, 80? 37? 54? w), conestee, South carolina (34? 46? 3? n, 82? 20? 52? w), and williamsburg, Virginia (37?16? 14? n, 76? 43? 2? w) between 2007 and 2009. we also sampled feathers from available museum specimens of birds collected on the breeding grounds. these were augmented with feath- ers from breeding Rusty Blackbirds captured at nest sites in anchorage, alaska (61? 17? 14? n, 149? 44? 25? w) in 2007 (Matsuoka et al. 2010). we obtained these samples of known origin opportunistically to test how well they corresponded to values of ?d in feathers that we predicted from values of ?d in precipitation (see below). StaBle-iSotope analySiS we cleaned all feathers of surface oils in a solvent consisting of a 2:1 ratio of chloroform to methanol and, following drying under a fume hood, prepared them for analysis of stable hydrogen 17_MS100146.indd 780 11/29/10 10:58:50 AM MigRatoRy connectiVity in the RuSty BlackBiRd 781 Spe c ia l Se c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d isotopes at environment canada?s stable-isotope laboratory in Saskatoon, Saskatchewan. analyses of stable hydrogen iso- topes in feathers used calibrated keratin isotope reference ma- terials and the comparative equilibration method described in detail by wassenaar and hobson (2003). we measured ratios of stable hydrogen isotopes on h2 derived from high- temperature flash pyrolysis of feathers by using continuous-flow isotope-ratio mass spectrometry. all ?d values are expressed in the typical delta notation, in units of per mil (?) and normalized on the Vienna Standard Mean ocean water?Standard light antarctic precipitation scale. Repeated analyses of hydrogen isotope inter-comparison material iaea-ch-7 (?100 ?) and our calibrated keratin reference materials yielded an external repeatability of better than ? 3.2 ?, based on the distribution of within-autorun residuals of accepted versus measured values for the reference materials. StatiStical analySiS we first examined factors influencing ?d values of feather samples. we considered a potential migratory divide be- tween origins of birds wintering on the atlantic coastal plain versus those wintering in the Mississippi alluvial Val- ley since the appalachians could represent a natural barrier. furthermore, the subspecies E. c. nigrans, which is restricted to breeding in atlantic canada, is thought to winter in north carolina and georgia (avery 1995). we therefore tested for mean differences in feather ?d values by wintering popula- tion (Mississippi alluvial Valley versus coastal plain) and time period (historic versus modern) with a full-factorial, two-way analysis of variance (anoVa). we tested the as- sumption of homogeneity of variance with levene?s test and examined the data and residuals for normality both graphi- cally and by the kolmogorov?Smirnov test. our data showed only a slight departure from normality; however, variance was greater in the Mississippi alluvial Valley than in the coastal plain (F3,724 = 11.3, P < 0.001). Because variances were heterogeneous, we checked our results with a kruskal? wallis test. we used SpSS version 17.0 (SpSS 2008) for all analyses. Establishing blackbird origins. in order to assign birds to geographic origins, we first converted the ?d values in feathers (hereafter ?df) to corresponding values of expected growing-season ?d in precipitation (hereafter ?dp; Bowen et al. 2005). we treated values of ?df from passerines and waterfowl of known source in north america as predictors regressed against ?dp estimated from a giS model of ?dp (Bowen et al. 2005, clark et al. 2006). Since there is error associated with both the analytical measurement of isotope values and the geospatial models of ?dp, we used reduced major-axis regression (Bowen et al. 2005). we ran the regres- sion on the mean values of ?df at each site to control for both pseudoreplication among individuals sampled at same site and differences in sampling effort among sites. parameter estimates for this regression were similar to those reported for other passerines (see references in hobson 2008), with the relationship defined as ?dp = 24.92 + 1.04?df (r2 = 0.95, df = 18, P < 0.001). we also examined this relationship from the feathers of breeding Rusty Blackbirds that we sampled from museum specimens (19 specimens, 1896?1988) and a recent field study (17 nesting individuals from anchorage, alaska, 2007). using these birds of known source, we used the re- gression equation above to convert the observed ?df values to ?dp values and then contrasted these with the values of growing-season ?dp predicted from the giS model by ordi- nary least-squares regression. assigning birds to their geographic origins on the basis of ?df is complicated by differences in physiology among birds growing feathers at the same site, analytical error, and errors associated with the isotope surface to which birds are being assigned (wunder and norris 2008, wunder 2010). therefore, it is not reasonable to assume that measured values of ?df can be mapped directly to a giS-based model of ?dp. therefore, prior to assignment of probable origins, we propagated er- rors associated with both ?df and ?dp. using data reported by clark et al. (2006), we calculated the standard deviation (Sd) of ?df values among wetland- and upland-associated birds at each of the 20 sites across north america ranging from 31.0? to 67.4? n latitude. we used maximum likelihood and fit a gamma distribution to the 20 observed Sds within each pop- ulation to estimate the nature of the distribution (Venables and Ripley 2002). using this distribution, we then propagated errors by simulating 1000 new expected ?dp values for each measured ?df value. each corresponding ?dp value was sim- ulated with a mean equal to the calculated ?dp value, and a standard deviation was drawn at random from the derived gamma distribution with a shape parameter of 7.156 and scale of 1.322. for each blackbird population of interest (Mississippi alluvial Valley historical and recent; coastal plain historical, winter 2007, and winter 2008?2009), we used the simulated ?dp dataset to calculate a probability density function with the MaSS library in R version 2.6.2 (R development core team 2008). the probability densities were normalized to the largest observed probability density within each blackbird population of interest. to map the probability of breeding or natal origins of wintering Rusty Blackbirds, we used arcgiS version 9.1 and Spatial analyst (eSRi 2005), to reclassify the giS-based model of ?dp (Bowen et al. 2005) into a probability surface. we accomplished this reclassification by (1) multiplying the probability-density values for a given giS grid cell (0.33?, Bo- wen et al. 2005) by 107 to provide the integer values required by arcgiS, (2) reclassifying these transformed values into the ?dp surface with Spatial analyst, (3) converting the reclassi- fied grid to a floating-point grid, and (4) dividing the values in the floating-point grid by 107 to obtain a probability grid with values ranging from 0 to 1. 17_MS100146.indd 781 11/29/10 10:58:51 AM 782 keith a. hoBSon et al. Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d ReSultS the anoVa examining factors influencing ?df explained 30% of the variance in the data. our results support the hy- pothesis of a migratory divide between the populations win- tering in the Mississippi alluvial Valley and atlantic coastal plain because there was a significant effect of wintering popu- lation on ?df (F1,724 = 216.8, P < 0.001). a kruskall?wallis test yielded similar results (kruskal?wallis ?2 = 223.4, df = 1, P < 0.001). Values of ?df did not differ by time period (F1,724 = 0.2, P > 0.1) nor with the interaction between time period and win- tering population (F1,724 = 1.9, P > 0.1). however, within the figuRe 2. temporal patterns of ?d values in feathers of Rusty Blackbirds captured in their winter range in the atlantic coastal plain and Mississippi valley, 1879?2009. Boxes represent upper and lower quartiles, heavy horizontal lines represent medians, whiskers represent ranges of nonoutlier values, and open circles indicate outliers. Sample sizes are given below the whiskers. More negative values represent origins at higher latitudes. coastal-plain sample, there was substantial short-term tempo- ral variation in the ?df values (table 1), which remained con- stant from 1879 to 1990, increased in 2007, and then shifted back to historical values in 2008?2009 (fig. 2). the 2007 coastal-plain sample came exclusively from South carolina; ?df values for this sample were not consistent with those for other samples collected in South carolina in other years or with samples from birds in neighboring states (fig. 3). our derived probability-density function for the population wintering on the coastal plain suggested that in 2007 the breed- ing origins of that population shifted south and possibly east of taBle 1. Summary statistics for ?d values (?) in feathers of Rusty Blackbirds captured on their wintering grounds in the Mississippi alluvial Valley and east of the appalachians on the atlantic coastal plain. historic samples (1879?1990) were obtained from museum collections and current samples were obtained from birds captured in the winter from 2000 to 2009. historic 2000?2007a 2008?2009b population n Mean Sd n Mean Sd n Mean Sd Mississippi valley 101 ?131.1 24.7 169 ?133.1 20.4 86 ?135.9 19.4 coastal plain 92 ?109.1 14.0 67 ?84.2 16.5 213 ?114.7 12.3 aMississippi valley sampled 2000?2007, coastal plain sampled only in winter of 2007. bBoth populations sampled during this period. 17_MS100146.indd 782 11/29/10 10:58:53 AM MigRatoRy connectiVity in the RuSty BlackBiRd 783 Spe c ia l Se c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d its historic area (fig. 4a and fig. 5a vs. 5b). the isotopic values of the historic population of birds wintering in the coastal plain corresponded to a band stretching from the mid-boreal zone of Manitoba through the hudson Bay lowlands of ontario and northeast toward the boreal shield of labrador (fig. 5a). the winter 2007 sample likely originated from the extreme south- eastern portion of the breeding range (fig. 5b), but the 2008? 2009 sample was consistent with the likely origins of the more broadly distributed historical population (fig. 5c vs. fig. 5a). in contrast, the historic and current Mississippi alluvial Val- ley populations had similar probability-density functions, so the current population likely originated from areas similar to those of the historic population (fig. 4b). Birds wintering in the Mississippi alluvial Valley potentially originated from a broad region from labrador west to alaska (fig. 5d?f), an area largely northwest of the probable current geographic origins of the coastal-plain population (figure 5a-c). finally, we examined the relationship between ?df and ?dp for Rusty Blackbirds of known origin. we considered this relationship by using all specimens except for three outliers. the relationship without outliers (observed ?dp = ?9.5 + 0.9 ? predicted ?dp; r2 = 0.41, n = 33, P < 0.001) provides some con- fidence in our overall approach and supports the contention that this relationship did not change during the last 100 years. diScuSSion population StRuctuRe our isotopic analysis of feathers from modern and histori- cal wintering populations of the Rusty Blackbird revealed a strong difference in the breeding origins of birds wintering on the east and west sides of the appalachians. on the ba- sis of both historic and modern samples, birds wintering to the west of the appalachians, in the Mississippi alluvial Val- ley, originated across a broad span of the boreal forest from alaska to labrador. the origins of birds wintering on the coastal plain, however, appeared to be much more restricted to eastern regions of the breeding range. while by no means definitive, our results support the hypothesis of a migratory divide in breeding origins of birds wintering to the east and west of the appalachians. Migratory divides have been doc- umented in other boreal-forest songbirds, such as wilson?s warbler (kelly et al. 2002) and Swainson?s thrush (Ruegg 2008), but these are located at the interface of the cordille- ran forests of the Rockies and the low-elevation boreal belt. if a migratory divide exists in the Rusty Blackbird, its loca- tion in the center of the boreal belt is a region lacking any ob- vious underlying physiographic division. Barriers to mixing on the wintering grounds might be a more likely mechanism figuRe 3. geographic variation of ?d values in feathers of Rusty Blackbirds captured on the atlantic coastal plain, 1879?2009. Boxes represent upper and lower quartiles, heavy horizontal lines represent medians, whiskers represent ranges of nonoutlier values, and open cir- cles indicate outliers. Sample sizes are given below the whiskers. More negative values represent origins at higher latitudes. 17_MS100146.indd 783 11/29/10 10:58:55 AM 784 keith a. hoBSon et al. Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d figuRe 4. probability-density distributions of Rusty Blackbird populations wintering in the southeastern u.S. in the (a) atlantic coastal plain and (B) Mississippi alluvial Valley based on ?d analysis of feathers. the coastal-plain sample consisted of feathers from 91 museum specimens and birds captured in the field in 2007 (n = 68) and 2008?2009 (n = 213). the Mississippi alluvial Valley sample consisted of feathers from 99 museum specimens and birds captured in the field in 2007 (n = 169) and 2008?2009 (n = 86). probability distributions result from fitting probability-density functions to propagated errors associated with isotope values. feather isotope values were converted to expected ?d precipitation values with a reduced major-axis regression of birds of known source (see Methods). for the creation of such a migratory divide. Recent analyses of christmas Bird counts by hamel and ozdenerol (2009) suggest that the appalachians may create a gap in the Rusty Blackbirds? winter distribution. Similarly, band recoveries suggest that only a small proportion of birds crosses the ap- palachians (hamel et al. 2009). Regardless, the strong migra- tory connectivity between breeding and wintering regions to the east and west of the appalachians has important implica- tions for management. the eastern wintering population appeared to shift breed- ing origins to more southeastern locations in 2007. however, the 2008?2009 sample was indistinguishable from the his- torical distribution. currently, we do not know if this change represented a shift in the breeding distribution that year or 17_MS100146.indd 784 11/29/10 10:59:09 AM MigRatoRy connectiVity in the RuSty BlackBiRd 785 Spe c ia l Se c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d figuRe 5. probable breeding origins of Rusty Blackbirds winter- ing in the southeastern u.S. along the atlantic coastal plain based on (a) museum specimens collected 1879?1990 and from live captures from (B) winter 2007 and (c) winter 2008?2009; probable breeding origins of Rusty Blackbirds wintering in the Mississippi alluvial Valley from (d) museum specimens and live captures from (e) win- ter 2007 and (f) winter 2008?2009. probable origins are inferred from probability densities fit to ?d values in feathers and incorporat- ing propagated errors associated with isotope signatures (see Meth- ods). probability densities were relativized to maximum values and then mapped to a giS-based model of ?d in precipitation (Bowen et al. 2005), after feather values were converted into expected pre- cipitation values on the basis of a reduced major-axis regression of birds of known source (see Methods). Maps depict the full isotopic solution space encompassing the Rusty Blackbird?s entire breeding range; however, birds wintering in the atlantic coastal plain more likely originate from eastern portions of the breeding range. if it was an artifact of sampling. in winter, Rusty Blackbirds are difficult to trap, particularly where they occur in low den- sities. wintering blackbirds undoubtedly change locations from year to year (hamel and ozdenerol 2009), and it is cur- rently unclear how populations may be further structured on the wintering grounds with respect to breeding origin. our results underline the need to sample as broadly as possible when isotopic methods are used in an attempt to infer dis- tributional changes in a broadly distributed species like the Rusty Blackbird. alternatively, our data may indicate higher variability in recruitment from source breeding populations to the wintering grounds in the east. although surveys in the southeastern breeding ground indicate a range contraction (hodgman and hermann 2003, hodgman and yates 2007) and a steep decline in breeding pairs (Sauer et al. 2007; Mari- time Breeding Bird atlas, unpubl. data), no equivalent data are available from the northernmost part of the breeding range in the east. Several studies suggest that habitat condi- tions in the northern boreal regions have changed (klein et al. 2005, Riordan et al. 2006, Savignac 2006). a fruitful avenue of research would be to identify possible historical changes in wetland health such as wetland size, water quality (degree of acidification, contamination), and food-web structure be- tween these different source areas. the more restricted area of the atlantic coastal plain sug- gests that attention be paid to the conservation of wintering habitat in that region since Rusty Blackbirds wintering there likely represent the small remaining population breeding in new england and the Maritime provinces. Stronger connec- tivity between eastern breeding and wintering populations also allows more direct testing of factors affecting population regulation there. for example, if wetland conditions or other components of suitability of breeding habitat in the east could be improved, we would expect to see a measurable response in populations wintering in the coastal plain. a lack of response would point more to population limitation on the wintering grounds or at stopover sites en route. the western wintering population almost certainly de- rives from a much larger and broader-based breeding popula- tion. our data suggest that little has changed historically in the distribution of origins of these birds. anecdotally, the southern limit of the Rusty Blackbird?s range appears to be retracting across the continent, yet populations have been more stable in the northwestern segment of the range in canada and alaska (Machtans et al. 2007, greenberg et al. 2011). however, the geo- graphic broadness of the isotopic contours related to growing- season precipitation, particularly across the east-to-west bands of the boreal forest, may have obscured changes across the central and western portions of the breeding range. More re- search is now required to improve our understanding of con- nectivity between breeding and wintering populations within this region. consideration of differences in wetland conditions in the western and eastern sections of the boreal forest and between those on the boreal shield in the east and those to the south appears to be the most parsimonious direction for future research. one additional consideration is that the populations wintering in the Mississippi alluvial Valley consist of birds that migrate, on average, much farther than those wintering along the atlantic coast, which should have implications for both demographic constraints and conservation strategies. the broad decline of the species on both sides of the mi- gratory divide (greenberg and droege 1999) indicates that 17_MS100146.indd 785 11/29/10 10:59:42 AM 786 keith a. hoBSon et al. Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d several factors are likely responsible. in the western breeding range, a likely candidate is wetland drying, which needs to be studied in more detail (greenberg et al. 2011). even though destruction and alteration of winter habitat are a likely cause for the historic decline of the species, it is less likely that these are the cause for the continuing decline as afforestation pro- grams, particularly in the Mississippi alluvial Valley, have increased the wooded area available to the species (hamel et al. 2009, greenberg et al. 2011). nonetheless, attention should be paid to possible changes in food availability or increased contamination in this region. in the eastern breeding range, acid rain in combination with methylmercury contamination (edmonds et al. 2010) and logging (powell et al. 2010) may be suppressing the population. data about habitat availability, habitat quality, and food availability are urgently needed there. finally, almost no information is available about resource re- quirements and limiting factors during migration. habitat de- struction and alteration may have a significant effect on both western and eastern populations of the Rusty Blackbird. StaBle iSotopeS, connectiVity, and conSeRVation the stable-isotope approach we have demonstrated here pro- vides an advance in the way we can examine population con- nectivity and limitation in a migratory species that is dispersed broadly and in low numbers across its breeding range. Several authors are now applying this approach to examine pressing conservation problems (hobson 2005, greenberg et al. 2007, perez and hobson 2007, wunder and norris 2008, hobson et al. 2009). this approach is especially useful for generating new and testable hypotheses that would be difficult or impos- sible through any other means. nonetheless, it is productive in all such applications to consider the assumptions inherent in this approach and to examine possible alternative hypotheses to explain the isotopic results (hobson and wassenaar 2008, wunder et al. 2009). we adopted a statistical approach to assigning breeding origins of migratory birds that (1) accounts for errors in both ?df and ?dp on the basis of reduced major-axis regression between ?df and ?dp for birds from known breeding loca- tions (clark et al. 2006) and (2) propagates population-level errors via simulation. however, our method did not account for errors associated with the long-term global network of isotopes in precipitation (gnip) dataset used to generate the mean surface of ?dp averaged over the growing season (Bo- wen et al. 2005). currently, an error surface is not available for the gnip dataset (g. J. Bowen, pers. comm.). another ba- sic assumption in the analysis of stable-isotope data for tem- poral trends is that the baseline isotopic surface of ?dp has not changed appreciably over the period of retrospective compari- son, about 100 years in our case. we assumed all changes in feathers? isotope values to be associated with changes in the distribution of origin of individuals and not to changes in the base map of precipitation. anecdotally, we note that there was no long-term trend in ?df values in Rusty Blackbirds wintering in either the Mississippi alluvial Valley or atlantic coastal plain. in addition, other authors who have used museum spec- imens of birds in north america report no effect of date on ?df values (kelly et al. 2002, lott and Smith 2006). So, our assumption of no major change in the ?dp surface is reason- able. in addition, our birds of known origin collected on the breeding grounds provided a reasonable fit to our algorithm linking values of ?df and ?dp. Several authors have emphasized the critical importance of establishing population connectivity between breeding, stopover, and wintering sites of migratory birds and other ani- mals (webster et al. 2002, webster and Marra 2005). knowl- edge of migratory connectivity will allow a more informed analysis of population demography and trends and the effects of seasonal interactions (Marra et al. 1998, Sillett et al. 2000, gunnarsson et al. 2005, holmes 2007). while the concept of protecting populations throughout their annual cycle seems obvious, our ability to make connections among populations within a species has been restricted overwhelmingly to game species that are amenable to band recoveries and satellite te- lemetry (hobson 2003). our study has emphasized the po- tential of the stable-isotope approach as a means of quickly and efficiently establishing patterns of migratory connectiv- ity in a declining but broadly distributed species. we also have shown the value of museum specimens for isotopic analysis as a means of retrospectively examining factors that may be linked to long-term declines in a species (Smith et al. 2003). importantly, our approach has served to quickly narrow down research questions associated with the potential recovery of the Rusty Blackbird. for example, we now suggest a focus on long-term population trends for the populations wintering in the Mississippi alluvial Valley and atlantic coastal plain. we also suggest that more scrutiny be directed toward under- standing Rusty Blackbirds breeding in the southern boreal forest of central and western canada since the christmas Bird count shows population trends for the Mississippi alluvial Valley and atlantic coastal plain to be equally negative de- spite the migratory divide (niven et al. 2004). finally, an ex- amination of historical landscape change and wetland quality east and west of the appalachians is now warranted. acknowledgMentS feathers were provided by Steve Matsuoka from anchorage, alaska, dan twedt from tensas River nwR, louisiana, and Jason d. luscier from arkansas. Samples were provided by Jacob armiger from Vir- ginia and by patty newell from the atlantic coastal plain. we thank the curators of all the museums who provided us with historic sam- ples: national Museum of natural history, Smithsonian institution; louisiana State university Museum of natural Science; american Museum of natural history; Museum of Vertebrate Zoology, univer- sity of california, Berkeley; georgia Museum of natural history; academy of natural Sciences, philadelphia; university of kansas 17_MS100146.indd 786 11/29/10 10:59:43 AM MigRatoRy connectiVity in the RuSty BlackBiRd 787 Spe c ia l Se c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d natural history Museum; north carolina Museum of natural Sci- ences, Burke Museum of natural history and culture, university of washington, Seattle; Mississippi Museum of natural Science. we thank the u.S. forest Service, center for Bottomland hardwoods, and the theodore Roosevelt nwR complex for providing hous- ing, and logistical support. the yazoo nwR complex, leroy percy State park, delta national forest, delta Research and extension center, frederick Ballard, and other farmers kindly permitted us to work on their refuges and lands. we very much thank our field assis- tants kathryn hayden, catherine Ricketts, and gerhard hofmann. Blanca Mora alvarez prepared feathers for analysis. len wassenaar provided expert assistance with the isotopic measurement of feath- ers. funding was provided by an operating grant to kah from envi- ronment canada and the western Boreal conservation initiative. Research, assistants, travel and material were supported by the u.S. fish and wildlife Service, the u.S. geological Survey, the cana- dian wildlife Service, friends of the national Zoo, the Max planck institute for ornithology, andechs, the german ethological Society, the Society for tropical ornithology, and the arthur-von-gwinner foundation. the paper benefited from the constructive comments of Steve Matsuoka and two anonymous reviewers. liteRatuRe cited aveRy, M. l. 1995. Rusty Blackbird (Euphagus carolinus), no. 200. In a. poole and f. gill, [eds.], the birds of north america. acad- emy of natural Sciences, philadelphia. boWen, G. J., l. i. WassenaaR, and K. a. hobson. 2005. global application of stable hydrogen and oxygen isotopes to wildlife forensics. oecologia 143:337?348. ClaRK, R. G., K. a. hobson, and l. i. WassenaaR. 2006. geographic variation in the isotopic (?d, ?13c, ?15n, ?34S) composition of feathers and claws from lesser Scaup and northern pintail: impli- cations for studies of migratory connectivity. canadian Journal of Zoology 84:1395?1401. CoRCoRan, R. M., J. R. lovvoRn, and P. J. heGlund. 2009. long- term changes in limnology and invertebrates in alaskan boreal wetlands. hydrobiologia 620:77?89. edMonds, s. t., d. C. eveRs, d. a. CRistol, C. MettKe-hofMann, l. l. PoWell, a. J. MCGann, J. W. aRMiGeR, o. P. lane, d. f. tessleR, P. neWell, K. heyden, and n. J. o?dRisColl. 2010. geographic and seasonal variation in mercury exposure of the declining Rusty Blackbird. condor 112:789?799. ellison, W. G. 1990. the status and habitat of the Rusty Blackbird in caledonia and essex counties. Vermont fish and wildlife department, woodstock, Vt. eRsKine, a. 1977. Birds in boreal canada. canadian wildlife Ser- vice Report Series 41. eSRi. 2005. arcgiS version 9.1. earth Systems Research institute, Redlands, ca. GReenbeRG, R., d. W. deMaRest, s. M. MatsuoKa, C. MettKe- hofMann, M. l. aveRy, P. J. blanCheR, d. eveRs, P. b. haMel, K. a. hobson, J. lusCieR, d. K. niven, l. l. PoWell, and d. shaW. 2011. understanding declines in Rusty Blackbirds. Studies in avian Biology, in press. GReenbeRG, R., and s. dRoeGe. 1999. on the decline of the Rusty Blackbird and the use of ornithological literature to document long-term population trends. conservation Biology 13:553?559. GReenbeRG R., P. P. MaRRa, and M. J. WoolleR. 2007. Stable- isotope (c, n, h) analyses help locate the winter range of coastal plain Swamp Sparrow (Melospiza georgiana nigrescens). auk 124:1137?1148. GunnaRsson, t. G., J. a. Gill, J. neWton, P. M. Potts, and W. J. sutheRland. 2005. Seasonal matching of habitat quality and fitness in a migratory bird. proceedings of the Royal Society B 272:2319?2323. haMel, P. b., d. de steven, t. leininGeR, and R. Wilson. 2009. historical trends in Rusty Blackbird nonbreeding habitat in forested wetlands, p. 341?353. In t. d. Rich, c. armizmendia, d. demarest, and c. thompson [eds.], tundra to tropics: con- necting birds, habitats and people. proceedings of the 4th inter- national partners in flight conference, 13?16 february 2008, Mcallen, tX. haMel, P. b., and e. ozdeneRol. 2009. using the spatial filtering process to evaluate the nonbreeding range of Rusty Blackbird, p. 334?340. In t. d. Rich, c. armizmendia, d. demarest, and c. thompson [eds.], tundra to tropics: connecting birds, habi- tats and people. proceedings of the 4th international partners in flight conference, 13?16 february 2008, Mcallen, tX. hobson, K. a. 2003. Making migratory connections with stable iso- topes, p. 379?391. In p. Berthold, e. gwinner, and e. Sonnen- schein [eds.], avian migration. Springer-Verlag, Berlin. hobson, K. a. 2005. Stable isotopes and the determination of avian migratory connectivity and seasonal interactions. auk 122:1037? 1048. hobson, K. a. 2008. applying isotopic methods to tracking ani- mal movements, p. 45?78. In k. a. hobson and l. i. wassenaar [eds.], tracking animal migration using stable isotopes. terres- trial ecology Series, vol. 2. academic press, london. hobson, K. a., and l. i. WassenaaR [eds.]. 2008. tracking animal migration using stable isotopes. terrestrial ecology Series vol. 2. academic press, london. hobson, K. a., h. loRM?e, s. l. van WilGenbuRG, l. i. Wasse- naaR, and J. M. boutin. 2009. Stable isotopes (?d) delineate the origins and migratory connectivity of harvested animals: the case of european woodpigeons. Journal of applied ecology 46:572?581. hodGMan, t. P., and h. l. heRMann. 2003. Rusty Blackbird, p. 65? 72. In h. l. hermann, t. p. hodgman, and p. deMaynadier [eds.], a survey of rare, threatened, and endangered fauna in Maine: St. John uplands and Boundary plateau (2001?2002). Maine department of inland fisheries and wildlife, Bangor, Me. hodGMan, t. P., and d. yates. 2007. Rusty Blackbird, p. 71?74. In h. l. hermann, t. p. hodgman, and p. deMaynadier [eds.], a survey of rare, threatened, and endangered fauna in Maine: aroostook hills and lowlands (2003?2005). Maine department of inland fisheries and wildlife, Bangor, Me. holMes, R. t. 2007. understanding population change in migratory songbirds: long-term and experimental studies of neotropical migrants in breeding and wintering areas. ibis 149:2?12. Kelly, J. f., v. atudoRei, z. d. shaRP, and d. M. finCh. 2002. insights into wilson?s warbler migration from analyses of stable hydrogen isotope ratios. oecologia 130:216?221. KennaRd, f. h. 1920. notes on the breeding habits of the Rusty Blackbird in northern new england. auk 37:412?422. Klein, e., e. e. beRG, and R. dial. 2005. wetland drying and suc- cession across the kenai peninsula lowlands, south-central alaska. canadian Journal of forest Research 35:1931?1941. lott, C. a., and J. P. sMith. 2006. a geographic-information-sys- tem approach to estimating the origin of migratory raptors in north america using hydrogen stable isotope ratios in feathers. auk 123:822?835. MaChtans, C. s., s. l. van WilGenbuRG, l. a. aRMeR and K. a. hobson [online]. 2007. Retrospective comparison of the occurrence and abundance of Rusty Blackbird in the Mac- kenzie Valley, northwest territories. avian conservation and ecology 2(1):3. (14 december 2009). 17_MS100146.indd 787 11/29/10 10:59:45 AM 788 keith a. hoBSon et al. Sp e c ia l S e c t io n : R a n g e w id e e c o l o g y o f t h e d e c l in in g R u St y B l a c k B iR d MaRRa, P. P., K. a. hobson, and R. t. holMes. 1998. linking win- ter and summer events in a migratory bird using stable carbon isotopes. Science 282:1884?1886. MatsuoKa, s. M., d. shaW, P. h. sinClaiR, J. a. Johnson, R. M. CoRCoRan, n. C. dau, J. a. Johnson, P. M. MeyeRs, and n. a. RoJeK. 2010. nesting ecology of the Rusty Blackbird in alaska and canada. condor 112:810?824. MettKe-hofMann, C., P. h. sinClaiR, P. b. haMel, and R. GReen- beRG. 2010. implications of prebasic and a previously undescribed prealternate molt for aging Rusty Blackbirds. condor 112:855? 862. niven, d. K., J. R. saueR, G. s. butCheR, and W. a, linK. 2004. christmas bird count provides insights into population change in land birds that breed in the boreal forest. american Birds 58:10?20. noRth aMeRiCan WateRfoWl ManaGeMent Plan. [online]. 2004. north american waterfowl Management plan 2004. Strategic guidance: strengthening the biological foundation. canadian wildlife Service, u.S. fish and wildlife Service, Secretar?a de Medio ambiente y Recursos naturales. (14 december 2009). PeRez, G., and K. a. hobson. 2007. feather deuterium measure- ments reveal origins of migratory western loggerhead Shrikes (Lanius ludovicianus excubitorides) wintering in Mexico. diver- sity and distributions 13:166?171. PoWell, l. l., t. P. hodGMan, W. e. Glanz, J. osenton, and C. fisheR. 2010. nest-site selection and nest survival of the Rusty Blackbird: does timber management adjacent to wetlands create ecological traps? condor 112:800?809. R develoPMent CoRe teaM. 2006. R: a language and environment for statistical computing. R foundation for Statistical computing, Vienna, austria. RioRdan, b., d. veRbyla, and a. d. MCGuiRe. 2006. Shrinking ponds in subarctic alaska based on 1950?2002 remotely sensed images. Journal of geophysical Research 111, g04002. RueGG, K. 2008. genetic, morphological, and ecological character- ization of a hybrid zone that spans a migratory divide. evolution 62:452?466. saueR, J. R., J. e. hines, and J. fallon [online] 2007. the north american Breeding Bird Survey, results and analysis 1966?2006. Version 10.13.2007. uSgS patuxent wildlife Research center, laurel, Md. (14 december 2009). saviGnaC, C. 2006. coSewic assessment and status report on the Rusty Blackbird (Euphagus carolinus) in canada. committee on the Status of endangered wildlife in canada, ottawa, on. sillett, t. s., R. t. holMes, and t. W. sheRRy. 2000. impacts of a global climate cycle on population dynamics of a migratory songbird. Science 288:2040?2042. sMith, t. b., P. MaRRa, M. s. WebsteR, i. lovette, l. Gibbs, R. t. holMes, K. a. hobson, and s. RohWeR. 2003. a call for feather sampling. auk 120:218?221. sPindleR, M. a., and b. Kessel. 1980. avian populations and habi- tat use in interior alaska taiga. Syesis 13:61?104. SpSS. 2008. SpSS Statistics Base 17.0 user?s guide. SpSS, chicago. u. s. fish and Wildlife seRviCe [online]. 2009. waterfowl popu- lation status, 2009. u. S. department of the interior, washing- ton, dc. (05 January 2010). venables, W. n., and b. d. RiPley. 2002. Modern applied statistics with S, 4th edition. Springer, new york. WassenaaR, l. i., and K. a. hobson. 2003. comparative equilibra- tion and online technique for determination of non-exchangeable hydrogen of keratins for use in animal migration studies. iso- topes in environmental and health Studies 39:1?7. WebsteR, M. s., P. P. MaRRa, s. M. haiG, s. bensCh, and R. t. holMes. 2002. links between worlds: unraveling migratory connectivity. trends in ecology and evolution 17:76?83. WebsteR, M. s., and P. P. MaRRa. 2005. the importance of under- standing migratory connectivity and seasonal interactions, p. 199?209. In R. greenberg and p. p. Marra [eds.], Birds of two worlds. Johns hopkins university press, Baltimore. WundeR, M. b. 2010. using isoscapes to model probability surfaces for determining geographic origins, p. 251?270. In J. B. west, g. J. Bowen, t. e. dawson, and k. p. tu [eds.], isoscapes: under- standing movement, pattern, and process on earth through iso- tope mapping. Springer, new york. WundeR, M. C., and d. R. noRRis. 2008. improved estimates of certainty in stable isotope-based methods for tracking migratory animals. ecological applications 18:549?559. WundeR, M., K. a. hobson, J. Kelly, P. MaRRa, l. i. WassenaaR, C. stRiCKeR, and R. douCette. 2009. does a lack of design and repeatability compromise scientific criticism? a response to Smith et al. (2009). auk 126:922?926. 17_MS100146.indd 788 11/29/10 10:59:46 AM