Translator Disclaimer
26 July 2017 The first 50 years of the North American Breeding Bird Survey
Author Affiliations +

The vision of Chandler (Chan) S. Robbins for a continental-scale omnibus survey of breeding birds led to the development of the North American Breeding Bird Survey (BBS). Chan was uniquely suited to develop the BBS. His position as a government scientist had given him experience with designing and implementing continental-scale surveys, his research background made him an effective advocate of the need for a survey to monitor pesticide effects on birds, and his prominence in the birding community gave him connections to infrastructure—a network of qualified volunteer birders who could conduct roadside surveys with standardized point counts. Having started in the eastern United States and the Atlantic provinces of Canada in 1966, the BBS now provides population change information for ∼546 species in the continental United States and Canada, and recently initiated routes in Mexico promise to greatly expand the areas and species covered by the survey. Although survey protocols have remained unchanged for 50 years, the BBS remains relevant in a changing world. Several papers that follow in this Special Section of The Condor: Ornithological Advances review how the BBS has been applied to conservation assessments, especially in combination with other large-scale survey data. A critical feature of the BBS program is an active research program into field and analytical methods to enhance the quality of the count data and to control for factors that influence detectability. Papers in the Special Section also present advances in BBS analyses that improve the utility of this expanding and sometimes controversial survey. In this Perspective, we introduce the Special Section by reviewing the history of the BBS, describing current analyses, and providing summary trend results for all species, highlighting 3 groups of conservation concern: grassland-breeding birds, aridland-breeding birds, and aerial insectivorous birds.

An Influential and Evolving Survey

The North American Breeding Bird Survey (BBS) was initiated in 1966 with a goal of monitoring change in North American breeding bird populations (Robbins et al. 1986). It now provides long-term population change data for ∼424 species over most of North America, with more limited data for an additional ∼122 species. BBS data inform virtually all geographic studies of North American birds; analyses show us which species are increasing and decreasing, and by how much (Sauer et al. 2017a). Its comprehensive nature and the ready availability of its results via the Internet have contributed to the perception of the BBS as a “one-stop shop” for population change data (Pardieck et al. 2016, Environment and Climate Change Canada 2017, Sauer et al. 2017a). The BBS is the premier source of bird population status and change data for conservation activities and scientific studies, as reviewed in two papers in this Special Section (Hudson et al. 2017, Rosenberg et al. 2017). Nevertheless, even after 50 years of data collection, the BBS is still a work-in-progress; the scope of the survey continues to expand while ongoing work seeks to strengthen BBS methods and analyses.

Origins of the BBS. The BBS was Chandler (Chan) S. Robbins's idea. As a biologist working for the U.S. Fish and Wildlife Service (USFWS), he had three research themes that came together to form the BBS (Robbins 2016, Sauer 2016). First, he had worked with DDT and other pesticides that affected birds, starting with field experiments at the Patuxent Wildlife Research Center in the 1940s (Linduska and Surber 1948). Rachel Carson edited Chan's reports on the consequences of DDT on birds, and he credited Carson with creating the public interest in bird populations that ultimately persuaded USFWS administrators to let him start the BBS (Chandler S. Robbins, personal communication). By the early 1960s, Chan was frequently being asked by the public about the effects of pesticides on bird populations, and he was keenly aware of the reports attributing avian mortality to pesticide exposure. Chan used Carson's (1962) Silent Spring as the basis of his lobbying within the USFWS for a continental-scale survey that would help us understand whether regional populations were declining and better evaluate pesticide effects on bird populations.

Second, Chan had been developing and implementing roadside surveys for American Woodcock (Scolopax minor), Mourning Dove (Zenaida macroura), and Wilson's Snipe (Gallinago delicata). He had been tasked with developing approaches for surveying these harvested species that would permit estimation of population change. He realized that these roadside survey methods could be easily modified to collect data on all species encountered along roads, as long as a corps of observers could be found to survey them.

Third, by the 1960s, Chan had several decades of experience working with citizen science projects, in particular the Christmas Bird Count, hawk watches, and breeding-bird censuses (Sauer and Droege 1990), and he had an extensive network of birding contacts across North America (Robbins 2016). He knew the value of networking and collaboration, and from these contacts he recruited observers for the survey and set up a network of state and provincial coordinators who could tend to the ongoing task of matching local birders to nearby routes. At the 2016 symposium at the North American Ornithological Conference celebrating the BBS's 50th anniversary, Chan related the story that the same day he received permission to start the breeding bird survey, he called Anthony (Tony) Erskine from the Canadian Wildlife Service and asked if Canada would be interested in participating. Tony took the proposal to his superiors, and he almost immediately called Chan back to say that Canada “was in.” Tony, and thus Canada, was a partner from the very start of the BBS. Chan capitalized on the pesticide concerns as a rationale for the survey, drew upon his prior experience in surveys to design the program, and was able to convince his birding and other professional connections to implement the program.

Silk purses and sows' ears. Chan was apparently a firm believer in the maxim “The perfect is the enemy of the good.” The BBS is (we would argue) “good,” and perhaps even unique and unparalleled as a coherent, continental-scale monitoring program. However, from the start, Chan endured aggressive criticism that the BBS's design had fatal flaws. Some of his colleagues in the USFWS asked pointed questions along these lines: How can you consider developing a monitoring program with no means of estimating detection rates of birds, and along roadsides where bird populations may not represent the broader landscape? Fifty years into the program, we are still asking these questions, and critics still point to these concerns with the BBS. However, the risks that Chan took in starting the BBS appear to have been justified; even though many alternatives to point counts now exist, research has not yet produced an alternative approach to data collection that is clearly superior to point counts and feasible to implement along BBS routes. Additionally, although research on deficiencies in BBS sampling has documented the need for ongoing vigilance in BBS analyses (e.g., Griffith et al. 2010), the research has not demonstrated fatal flaws in the BBS methods.

Consequently, the BBS's design and field protocols have remained the same over 50 years of surveying. Surveyors from 1966 could run a BBS route today and feel completely comfortable, although they might be a bit inconvenienced by safety straps, alarm chimes, and odd buzzing noises, or distracted by the built-in GPS units and media centers of modern vehicles. Once escaping the vehicle, however, the survey proceeds as it did in 1966. This is remarkable, considering how much the world has changed around the BBS. Sauer et al. (2013) describe some of these changes: (1) In addition to changes in car technology, there are more cars on the roads, and their presence influences counts; (2) climate is changing, as evidenced by earlier springs and differing seasonal patterns of bird activity; (3) roadside habitats have changed, with more houses and fewer natural habitats along BBS routes; and (4) small roads that host BBS routes have become larger roads with more cars and more disturbance. The survey has also expanded, from the original survey area in the eastern United States in 1966 to the contiguous United States and southern Canada by 1968. Additional expansion has occurred almost every year of the BBS, and recent expansion has taken the survey into northern Mexico.

Along with the environment in which counts are conducted, our notions of appropriate ways of counting birds have also changed. Simple point counts such as those collected by the BBS have been shown to be subject to a variety of environmental factors that influence detection of birds (Nichols et al. 2009), and the analysis of a survey that “encounters an unknown proportion of birds in an undefined area” (Link and Sauer 1998a) has its complications. In the years following the implementation of the BBS, myriad quantitative approaches were developed for obtaining reliable estimates of bird population size or density from counts (Nichols et al. 2009). Maintenance of the simple survey design in the face of these methodological developments is not due to apathy or a lack of inspiration; the BBS programs in Canada, Mexico, and the United States have encouraged these developments by sponsoring many research programs designed to test new counting methods (e.g., Farnsworth et al. 2005) and assess consequences of roadside sampling (Sauer et al. 2013, Veech et al. 2017). Rather, it is due to the fact that no method yet suggested has the flexibility to be implemented on roadside surveys conducted by thousands of observers. There is also a scale issue, as the current analyses focus on estimation of change at the route level, rather than at the scale of individual counting locations (stops) along the route (Sauer 2016). Many of the factors that influence detectability, such as habitat, operate at the scale of individual stops, but it is only in recent years that the BBS offices have begun to curate bird and location data at the stop level. Full investigation of detection, as it relates to the BBS, must wait until reliable information exists as to where stops actually occur along BBS routes. In the absence of stop-level information provided by the BBS, researchers have used remote sensing to determine this information for individual projects, as in Niemuth et al. (2017) in this Special Section.

The BBS has maintained credibility in the face of changing environments and developed a reputation for robustness due to innovations in analyses. Development of Bayesian approaches for fitting hierarchical models have allowed us to overcome scale-specific limitations that made early analyses of BBS data cumbersome exercises in approximation (Sauer 2016). Implementation of these model-based approaches has also allowed us to address the fundamental criticisms of the BBS (e.g., Link and Sauer 1998b) by providing the means for evaluating effects of the changing world on BBS results and by controlling for environmental changes such as vehicle disturbance and phenology (as indexed by counting day). Expansion of the survey is accommodated in this model-based framework by imposing hierarchical structure among regions to enhance estimation in strata with limited data. Modeling can be extended to accommodate off-road expansions of the survey, as is reported for Alaska in this Special Section (Handel and Sauer 2017). In our view, BBS analyses require statistical controls for the effects of routes and observers; we can think of no inference based on BBS data that would provide reliable results without these controls, and hierarchical models are an essential component of the ongoing exploration of how the changing environment along BBS routes influences counts (e.g., Griffith et al. 2010).

One great benefit of hierarchical models is that they have changed our perspective on detectability modeling; advances in hierarchical modeling have led to a confluence of approaches for estimating both population change and detectability. In each, the underlying population size (at stops or routes, depending on the analysis) is viewed as a latent parameter, and modeled connections of the counts to the underlying population sizes form the basis of inference (Kéry et al. 2009). However, the goal of most BBS analyses is unbiased estimation of change over time, while most detectability analyses focus on directly estimating the local population size. For the goal of estimation of population change over time, current analysis methods control for observer differences at the scale of routes and also allow for controlling for additional features such as vehicle-related disturbance or phenology (Sauer et al. 2013), features long thought to possibly bias estimation of change over time. Identifying factors that influence detection, and determining their importance for inclusion in the BBS analysis, is our primary tool for addressing concerns about the counting process of the BBS (e.g., Sauer et al. 1994); ongoing assessments include modeling the effects of experimental protocol changes (e.g., collecting time-distance information; Twedt 2015) and phenology change (Sauer et al. 2013).

One key consequence of the BBS design and analysis is that population size is not easily estimated; the model-based controls for detectability allow for estimation of population change but do not provide the information needed to scale the relative population indices produced in BBS analyses to an absolute population size. Although changes in field protocols have been suggested for the BBS to better inform population estimation (e.g., Farnsworth et al. 2005, Twedt 2015), analyses using these approaches have not yet proved effective for estimating detectability at critical scales needed for analysis as they have been applied only to estimate species-level detection rates. These species-level detectability adjustments provide no information relevant for BBS population change analyses. However, population estimates are often required for management needs such as setting population goals (Rosenberg and Blancher 2005) or estimating allowable take (Runge et al. 2009). Researchers have used additional data to scale BBS results to actual population sizes through (1) applying a series of adjustments that collectively estimate actual detection rates (Rosenberg and Blancher 2005, Runge et al. 2009), (2) using data from other surveys to scale BBS data to produce an unbiased population estimate (e.g., Zimmerman et al. 2015, 2017), or (3) modeling on-road vs. off-road populations using population and habitat data collected on and off roads (Sauer et al. 2013, Sauer 2016).

This discussion emphasizes an essential attribute of the BBS, and of any other omnibus, continental-scale survey: Wise use and interpretation of the survey involves an ongoing process of exploring how the counts relate to actual populations, in terms of both detectability and how sampling varies across space and time, and in developing appropriate models that adequately represent these relationships.

BBS Results

Hierarchical models for BBS analyses. Here, we provide a brief summary of BBS results from 50 years of surveying. Unfortunately, Mexican results do not yet provide sufficient information for analyses of population change. We provide results for 424 species from a “core” area that includes data extending back to 1966, as well as results from the period 1993–2015 for 546 species in an expanded survey area. The core area is the contiguous United States and southern portions of Canada (Sauer and Link 2011). The expanded area adds 7 additional strata (defined by Bird Conservation Regions within states and provinces): Western Alaska, Alaska Arctic Plains and Mountains, Alaska Northern Pacific Rainforest, Alaska Northwestern Interior Forest, Yukon Territory Northwestern Interior Forest, Northwest Territories Boreal Taiga Plains, and Newfoundland Boreal Softwood Shield. Prior to 1993, these 7 strata had very limited coverage. See Sauer et al. (2017b) for details of the core and expanded survey areas and strata.

The summary results we present here are based on a log-linear hierarchical model in which the log of the expected counts is a linear function of stratum (S), slope (β), year (γ), observer/route (ω), first year (η), and overdispersion (ϵ) effects, that is:


Counts are assumed to be distributed as Poisson, i, j, and t index stratum, route/observer, and year, respectively, and t* is a fixed year (1986) that centers the regression. Descriptions of the distributions of these parameters are provided in Sauer et al. (2013); the analysis presented here differs slightly from earlier implementations, in that stratum and slope effects are hierarchical, governed by mean and variance hyperparameters that have diffuse normal and gamma distributions, respectively (Sauer et al. 2017b).

This model contains parameters related to population change (i.e. β, γ) that are indexed at the stratum scale. Summary of population change is accomplished by first estimating a time series of annual indices that are functions of stratum abundance, slope and year effects, and variance components that are added to accommodate asymmetries in the log normal distribution:

where zi is a scaling factor (proportion of routes in which the species was encountered in the region). Indices for groups of strata are area-weighted (among regions) yearly indices. Trend is defined as the ratio of annual indices (for region i) for the first year (ta) and last year (tb) in the period of interest, taken to the appropriate power:

For regions composed of several strata, trend was defined as the ratio of the regional annual indices. Trend is presented as percent change per year (i.e. (Bi − 1)*100%). Models were fit using Bayesian methods, via the program JAGS (, and inference was based on medians and credible intervals computed from the posterior distributions of parameters and derived statistics. To accommodate the differences in estimated precision in comparing species trend results, we employed the hierarchical model approach described by Sauer and Link (2002) for ranking and displaying summary trend results. Each species' estimated trend is considered to be normally distributed, with a mean and variance that represent the trend parameter and variance for the species. These trend parameters are defined as normally distributed, with a common overall mean and variance (hierarchical parameters). Modeling the distribution of trends across all species allowed us to estimate the number of increasing species (species with trend >0) and provided a better ranking of the actual trend parameters. We also implemented a State of the Birds summary (e.g., North American Bird Conservation Initiative, U.S. Committee [NABCI] 2014) of composite trajectories for selected species groups. These summaries apply an analysis similar to that in Sauer and Link (2002) to estimate yearly composite mean change, by applying a hierarchical model to annual estimates of change from an initial base year for each subsequent year in the time series (Sauer and Link 2011). The yearly hierarchical models differ from those in Sauer and Link (2002) in that the log means were modeled, leading to a geometric mean summary of trajectories over time (Sauer and Link 2011).

We note that prior BBS analyses (e.g., Sauer et al. 2017a) did not include hierarchical structure in β and S (i.e. these parameters were assumed to be independently distributed as normal random variables with mean 0 and variance 1 × 10−6). The present analysis also included strata with smaller sample sizes than were used in prior analyses (≥3 routes; Sauer et al. 2017b). We thus computed trends for the 424 species in the core area from 1966 to 2015 using Sauer et al.'s (2017a) model, and we provide occasional comparisons with results to reassure readers of continuity with prior analyses. In recent years, BBS data have been used to document consistent declines in several groups of birds, particularly in grassland-obligate and aridland-obligate breeding bird species (NABCI 2014) and in aerial insectivore species (Nebel et al. 2010, Smith et al. 2015). Because these groups are experiencing the largest declines of any group of species in North America, we highlight their trends in our summary analyses and use Sauer and Link's (2002) method to estimate the proportion of those species with positive trends for the periods 1966–2015 and 1993–2015. We also computed composite population change graphs (i.e. State of the Birds summaries) for these groups.

Fifty-year trends. Over the long term (1966–2015), significantly more species are declining than increasing in the core area. Of the 424 species we analyzed, 195 (95% credible interval: 186, 205) species, or 46% (43.8, 48.2), had positive trends as estimated using Sauer and Link's (2002) hierarchical model (Figure 1 and Appendix; for scientific names of species, see Appendix). Extreme declines occurred in Black Swift (−7.5% yr−1; −9.1, −4.3), Bank Swallow (−4.9% yr−1; −6.0, −3.9), Evening Grosbeak (−5.0% yr−1; −6.4, −3.9), Chestnut-collared Longspur (−4.1% yr−1; −5.1, −3.3), and Blackpoll Warbler (−4.3% yr−1; −8.2, −1.7). However, other species are experiencing extreme population increases. Top increasers include Eurasian Collared-Dove (32.2% yr−1; 27.6, 35.4), Cave Swallow (22.5% yr−1; 18.1, 26.7), Wild Turkey (8.0% yr−1; 7.1, 8.8), Couch's Kingbird (9.0% yr−1; 8.0, 11.4), and Swallow-tailed Kite (6.5% yr−1; 5.1, 7.3). Extreme increasing and declining species are listed in the rank order estimated by Sauer and Link's (2002) hierarchical model. The analysis using nonhierarchical β and S indicated 182 (171, 194) species, or 43% (40.2, 45.6) species with positive trends (Figure 1; Sauer et al. 2017a).


Ranked trends (1966–2015) for 424 species of North American birds, as estimated by the North American Breeding Bird Survey. The upper panel shows the results from the log-linear model with hierarchical β and S components, and the lower panel shows results from Sauer et al. (2017a) for the model with nonhierarchical β and S. Species trend data are presented in the Appendix. In each panel, the horizontal line indicates the zero trend, and species are ranked by trend magnitude on the x-axis. Each species trend is indicated by the median (circle) and the 95% credible interval of the posterior distribution of the trend parameter estimated using Sauer and Link's (2002) model. The vertical line indicates the rank order of the species with positive trends (i.e. species to the right of the line have positive trends as identified by the hierarchical model). Red circles indicate ranks of grassland-breeding bird species, and black circles indicate ranks of aridland-breeding bird species.


One of the most obvious generalizations to be made from BBS data is that “big-picture” views of bird populations are not particularly informative. To make sense of a collection of population change estimates from an omnibus survey such as the BBS, we must either consider the individual characteristics of the species (e.g., the extreme increases shown by the invasive Eurasian Collared-Dove) or look for patterns of change among species sharing common habitat or other life-history attributes. Sauer et al. (2013) provide some discussion of species and group patterns of population change. Here, we update the status assessment of our 3 groups of management interest. Of the 24 grassland bird species, 8 (5, 10) species, or 32% (20, 40), were increasing. Of the 22 aridland bird species, 7 (5, 10) species, or 31.8% (22.7, 45.4), were increasing. Of the 31 aerial insectivores, 8 (6, 10) species, or 25.8% (19.3, 32.3), were increasing. State of the Bird summaries for the 3 species groups (Figure 2) show similar patterns in the context of time series of composite change for the groups. The model with hierarchical β and S indicates slightly more positive trajectories than the model with nonhierarchical β and S, with very similar patterns of year-to-year change.


State of the Birds composite summaries of population change for 3 groups of management interest: (A) grassland-breeding birds, (B) aridland-breeding birds, and (C) aerial insectivorous species, as defined by Nebel et al. (2010). Index is total proportional change (median and 95% credible interval of the posterior distribution) from the 1968 base year.


Recent changes in the expanded survey area (1993–2015). Over the short term (1993–2015), bird species tend to have more positive population trajectories. Of the 546 species included in the expanded area analysis, 306 (294, 318) species, or 56% (53.8, 58.2), had positive trends (Figure 3 and Appendix). Core area results based on the 424 species for which long-term trends were computed had similar proportions of increasers to the larger species collection, with 54% (51.8, 56.5) of species increasing. The declining species groups, although still declining, show more positive trends compared to long-term results in composite analyses. Of the 24 grassland bird species, 10 (8, 12) species, or 41% (32, 48), were increasing. Of the 22 aridland bird species, 10 (7, 12) species, or 45.4% (31.8, 54.6), were increasing. And of the 31 aerial insectivorous species, 11 (8, 13) species, or 35.4% (25.8, 41.9), were increasing. This pattern of less-severe declines after 1993 contrasts with the group trajectories for aerial insectivores estimated in Smith et al. (2015), which generally showed that more recent trends were more severe than earlier trends.


Ranked trends (1993–2015) for 546 species of North American birds, as estimated from the North American Breeding Bird Survey. Shown are the results from the log-linear model with hierarchical β and S components for the expanded survey area. Species trend data are presented in the Appendix. The horizontal line indicates the zero trend, and species are ranked by trend magnitude on the x-axis. Each species trend is indicated by the median (circle) and the 95% credible interval of the posterior distribution of the trend parameter estimated using Sauer and Link's (2002) model. The vertical line indicates the rank order of the species with positive trends (i.e. species to the right of the line have positive trends as identified by the hierarchical model). Red circles indicate ranks of grassland-breeding bird species, and black circles indicate ranks of aridland-breeding bird species.


Changing Bird Populations, Changing Analyses

The 50 years of BBS population change results provide the fundamental information base for bird conservation in North America (Hudson et al. 2017, Rosenberg et al. 2017). Identification of species-level patterns of population change and identifying commonalities in trends among species that share breeding habitats or migration status have proved to be effective approaches for defining groups of species meriting conservation action (NABCI 2014). As evidenced by recent population increases, period-specific patterns of change are also of conservation interest and provide important insights into population change associated with temporal variation in weather and other environmental features (Huang et al. 2016). In addition to describing patterns of population change, modern BBS analyses offer new opportunities for testing hypotheses regarding factors that influence population change. With data available at scales ranging from an individual stop to Bird Conservation Regions and even larger geographic scales, the BBS can be used to model spatial as well as temporal associations of bird abundance and change (Niemuth et al. 2017). Hierarchical models also permit aggregation of information among species and can be configured as full life-cycle models that integrate BBS data with banding and other information (Hudson et al. 2017). Model-based BBS analysis thus provides a framework both for controlling for structural limitations such as detectability and for development of models that allow us to predict environmental influences on bird populations. Integrated population models such as that employed for Wood Ducks in this Special Section (Zimmerman et al. 2017) illustrate how hierarchical models allow us to combine BBS results with other datasets to enhance the use of BBS data in population management.

Although the hierarchical models we use for BBS analyses offer many possibilities for analysis, it is difficult to avoid getting bogged down in details of the many models that could be applied to the BBS (Link and Sauer 2016). Even among the national agencies that administer the surveys, we choose slightly different model structures and spatial structuring for summary analyses (e.g., Environment and Climate Change Canada 2017, Sauer et al. 2017a). One of the perennial to-do-list items for administration of the BBS is to tighten collaboration between the national BBS offices, as well as among other groups doing BBS analyses, to ensure authoritative presentation of results. At the moment, achieving this goal is complicated by two issues: (1) uncertainty about details of model structure (Link and Sauer 2016, Link et al. 2017) and (2) expansion of the survey into new regions. Both of these are topics of active research (e.g., Link and Sauer 2016, Sauer et al. 2017b). Link et al. (2017) used cross-validation methods to compare 4 alternative models for 20 species from BBS data. Given the complexity of the modeling, the lack of temporal and spatial balance in the data due to the expansion of the survey over time (Sauer et al. 2013), and the regions of analysis, our perceptions of the best analysis are certain to be evolving. Although the timely incorporation of improved analyses can be helpful in terms of providing the best available information to users, we strongly advocate peer review of new methods and comparative analyses that ensure credibility and consistency in results over time (e.g., Smith et al. 2014, Sauer et al. 2017b).APPENDIX.


The real heroes of the BBS program are the volunteer observers, who get up early, travel long distances, provide their expertise to monitor North American bird populations, and follow through with data entry and verification. We thank them, and we also thank BBS staff from 3 countries who shuffle the papers and crunch the numbers, providing the crucial link from project to product. Finally, we thank Chan Robbins, Tony Erskine, and the other founders of the BBS, who had the fortitude to establish and maintain the BBS in the early years when its value was not always evident to administrators. This Perspective and other papers in this Special Section had their impetus in a symposium on the BBS at the North American Ornithological Conference in Washington, D.C., in 2016. J. Hatfield provided useful comments on the manuscript. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. or Canadian Government.

Author contributions: J.R.S., K.L.P., D.J.Z., A.C.S., M.-A.R.H., V.R., and H.B. organized the symposium. J.R.S., W.A.L., and D.K.N. conducted analyses. J.R.S. wrote the original draft and all authors participated in editing the manuscript.



Carson, R. (1962). Silent Spring. Houghton Mifflin, Boston, MA, USA. Google Scholar


Environment and Climate Change Canada (2017). North American Breeding Bird Survey: Canadian Trends Website, Data, version 2015. Environment and Climate Change Canada, Gatineau, Quebec, Canada. Google Scholar


Farnsworth, G. L., J. D. Nichols, J. R. Sauer, S. G. Fancy, K. H. Pollock, S. A. Shriner, and T. R. Simons (2005). Statistical approaches to the analysis of point count data: A little extra information can go a long way. InBird Conservation Implementation and Integration in the Americas: Proceedings of the Third International Partners in Flight Conference ( C. J. Ralphand T. D. Rich, Editors). USDA Forest Service General Technical Report PSW-GTR-191.pp. 736–743. Google Scholar


Griffith, E. H., J. R. Sauer, and J. A. Royle (2010). Traffic effects on bird counts on North American Breeding Bird Survey routes. The Auk 127:387–393. Google Scholar


Handel, C. M., and J. R. Sauer (2017). Combined analysis of roadside and off-road breeding bird survey data to assess population change in Alaska. The Condor: Ornithological Applications 119:557–575. Google Scholar


Huang, Q., J. R. Sauer, A. Swatantran, and R. Dubayah (2016). A centroid model of species Distribution with applications to the Carolina Wren Thryothorus ludovicianus and House Finch Haemorhous mexicanus in the United States. Ecography 39:54–66. Google Scholar


Hudson, M.-A. R., C. M. Francis, K. J. Campbell, C. M. Downes, A. C. Smith, and K. L. Pardiek (2017). The role of the North American Breeding Bird Survey in conservation. The Condor: Ornithological Applications 119:526–545. Google Scholar


Kéry, M., R. M. Dorazio, L. Soldaat, A. van Strien, A. Zuiderwijk, and J. A. Royle (2009). Trend estimation in populations with imperfect detection. Journal of Applied Ecology 46:1163–1172. Google Scholar


Linduska, J. P., and E. W. Surber (1948). Effects of DDT and other insecticides on fish and wildlife: Summary of investigations during 1947. U.S. Fish and Wildlife Service Circular 15. Google Scholar


Link, W. A., and J. R. Sauer (1998a). Estimating population change from count data: Application to the North American Breeding Bird Survey. Ecological Applications 8:258–268. Google Scholar


Link, W. A., and J. R. Sauer (1998b). Estimating relative abundance from count data. Austrian Journal of Statistics 27:83–97. Google Scholar


Link, W. A., and J. R. Sauer (2016). Bayesian cross-validation for model evaluation and selection, with application to the North American Breeding Survey. Ecology 97:1746–1758. Google Scholar


Link, W. A., J. R. Sauer, and D. K. Niven (2017). Model selection for the North American Breeding Bird Survey: A comparison of methods. The Condor: Ornithological Applications 119:546–556. Google Scholar


Nebel, S., A. Mills, J. D. McCracken, and P. D. Taylor (2010). Declines of aerial insectivores in North America follow a geographic gradient. Avian Conservation and Ecology 5(2):1. Google Scholar


Nichols, J. D., L. L. Thomas, and P. B. Conn (2009). Inferences about landbird abundance from count data: Recent advances and future directions. InEnvironmental and Ecological Statistics, vol. 3: Modeling Demographic Processes in Marked Populations ( D. L. Thomson, E. G. Cooch, and M. J. Conroy, Editors). Springer, New York, NY, USA. pp. 201–235. Google Scholar


Niemuth, N. D., M. E. Estey, S. P. Fields, B. Wangler, A. A. Bishop, P. J. Moore, R. C. Grosse, and A. J. Ryba (2017). Developing spatial models to guide conservation of grassland birds in the U.S. Northern Great Plains. The Condor: Ornithological Applications 119:506–525. Google Scholar


North American Bird Conservation Initiative, U.S. Committee (2014). The State of the Birds 2014 Report. U.S. Department of Interior, Washington, DC, USA. Google Scholar


Pardieck, K. L., D. J. Ziolkowski, Jr., M.-A. R. Hudson, and K. Campbell (2016). North American Breeding Bird Survey Dataset 1966–2015. Version 2015.1. U.S. Geological Survey, Patuxent Wildlife Research Center, Laurel, MD, USA. Google Scholar


Robbins, C. S. (2016). Early avian studies at Patuxent. InThe History of Patuxent—America's Wildlife Research Story ( M. C. Perry, Editor). U.S. Geological Survey Circular 1422. Google Scholar


Robbins, C. S., D. Bystrak, and P. H. Geissler (1986). The Breeding Bird Survey: Its first fifteen years, 1965–1979. U.S. Fish and Wildlife Service Resource Publication 157. Google Scholar


Rosenberg, K. V., and P. J. Blancher (2005). Setting numerical population objectives for priority landbird species. InBird Conservation Implementation and Integration in the Americas: Proceedings of the Third International Partners in Flight Conference ( C. J. Ralphand T. D. Rich, Editors). USDA Forest Service General Technical Report PSW-GTR-191.pp. 57–67. Google Scholar


Rosenberg, K. V., P. J. Blancher, J. C. Stanton, and A. O. Panjabi (2017). Use of North American Breeding Bird Survey Data in avian conservation assessments. The Condor: Ornithological Applications119:594–606. Google Scholar


Runge, M. C., J. R. Sauer, M. L. Avery, B. F. Blackwell, and M. D. Koneff (2009). Assessing allowable take of migratory birds: Black Vultures in Virginia. The Journal of Wildlife Management 73:556–565. Google Scholar


Sauer, J. R. (2016). Patuxent's role in the development of the Breeding Bird Survey. InThe History of Patuxent—America's Wildlife Research Story ( M. C. Perry, Editor). U.S. Geological Survey Circular 1422. Google Scholar


Sauer, J. R., and S. Droege (Editors) (1990). Survey designs and statistical methods for the estimation of avian population trends. U.S. Fish and Wildlife Service Biological Report90(1). Google Scholar


Sauer, J. R., and W. A. Link (2002). Hierarchical modeling of population stability and species group attributes using Markov chain Monte Carlo methods. Ecology 83:1743–1751. Google Scholar


Sauer, J. R., and W. A. Link (2011). Analysis of the North American Breeding Bird Survey using hierarchical models. The Auk 128:87–98. Google Scholar


Sauer, J. R., W. A. Link, J. E. Fallon, K. L. Pardieck, and D. J. Ziolkowski, Jr. (2013). The North American Breeding Bird Survey 1966–2011: Summary analysis and species accounts. North American Fauna79. Google Scholar


Sauer, J. R., D. K. Niven, J. E. Hines, D. J. Ziolkowski, Jr., K. L. Pardieck, J. E. Fallon, and W. A. Link (2017a). The North American Breeding Bird Survey, Results and Analysis 1966–2015. Version 12.23.2015. USGS Patuxent Wildlife Research Center, Laurel, MD, USA. Google Scholar


Sauer, J. R., D. K. Niven, K. L. Pardieck, D. J. Ziolkowski, and W. A. Link (2017b). Expanding the North American Breeding Bird Survey analysis to include additional species and regions. Journal of Fish and Wildlife Management 8:154–172. Google Scholar


Sauer, J. R., B. G. Peterjohn, and W. A. Link (1994). Observer differences in the North American Breeding Bird Survey. The Auk 111:50–62. Google Scholar


Smith, A. C., M.-A. R. Hudson, C. M. Downes, and C. M. Francis (2014). Estimating breeding bird survey trends and annual indices for Canada: How do the new hierarchical Bayesian estimates differ from previous estimates?Canadian Field-Naturalist 128:119–134. Google Scholar


Smith, A. C., M.-A. R. Hudson, C. M. Downes, and C. M. Francis (2015). Change points in the population trends of aerial-insectivorous birds in North America: Synchronized in time across species and regions. PLoS ONE 10:e0130768. Google Scholar


Twedt, D. J. (2015). Estimating regional landbird populations from enhanced North American Breeding Bird Surveys. Journal of Field Ornithology 86:352–358. Google Scholar


Veech, J., K. Pardieck, and D. Ziolkowski, Jr. (2017). How well do route survey areas represent landscapes at larger spatial extents? An analysis of land cover composition along Breeding Bird Survey routes. The Condor: Ornithological Applications 119:607–615. Google Scholar


Zimmerman, G. S., J. R. Sauer, G. S. Boomer, P. K. Devers, and P. R. Garrettson (2017). Integrating Breeding Bird Survey and demographic data to estimate Wood Duck population size in the Atlantic Flyway. The Condor: Ornithological Applications 119:616–628. Google Scholar


Zimmerman, G. S., J. R. Sauer, K. Fleming, W. A. Link, and P. R. Garrettson (2015). Combining waterfowl and breeding bird survey data to estimate Wood Duck breeding population size in the Atlantic Flyway. The Journal of Wildlife Management 79:1051–1061. Google Scholar



Population trends for 546 species of North American breeding birds during the periods 1966–2015 and 1993–2015, as documented by the North American Breeding Bird Survey, with lower (2.5%) and upper (97.5%) limits of 95% credible intervals. N is the total number of routes used in the analysis for each species.

© 2017 Cooper Ornithological Society.
John R. Sauer, Keith L. Pardieck, David J. Ziolkowski, Adam C. Smith, Marie-Anne R. Hudson, Vicente Rodriguez, Humberto Berlanga, Daniel K. Niven, and William A. Link "The first 50 years of the North American Breeding Bird Survey," The Condor 119(3), 576-593, (26 July 2017).
Received: 26 April 2017; Accepted: 1 May 2017; Published: 26 July 2017

Back to Top