AudioDateDownLeftRightUpIconCloseDownload iconfacebookReportGallerySettingsGiftLanguageGridLanguage iconListMapMenunoAudionoPhotoPhotoPlayPlusSearchStartwitterUserVideo

General

What are eBird Science Products?

The eBird Status and Trend products provide basic ecological information for over 100 species in North America, describing their ranges, abundances, environmental associations, and population trends. To generate these products, we use statistical and machine learning analyses designed to combine eBird data with environmental data from NASA. The analyses are used to predict the occurrence and abundance of species across North America at weekly intervals. These predictions are the building block of the Status and Trends products and are summarized in several ways to produce the different products.

There are five major products for eBird Status and Trends:

  1. Abundance animations represent weekly relative abundances, revealing movements of a population throughout the year.
  2. Seasonal abundance maps indicate the average relative abundance of a species in each season of their annual cycle.
  3. Trend maps show where there is evidence of population change between 2007 and 2016 across the species range.
  4. Habitat association regional charts show seasonal patterns of habitat association and avoidance within regions.
  5. Range maps show species seasonal range boundaries, similar to traditional range maps.

Different products require different volumes of data and the most data-intensive products such as trend maps are only available for some of the species and seasons. We only present products for species and seasons that have passed analytical and expert quality review tests.

What eBird data were used to generate the Status and Trends Products?

The eBird checklists that were used to generate the Status and Trends Products are referred to as “qualifying checklists.” Checklists need to meet a number of conditions to be qualifying checklists:

  • Submitted as of 31 January 2017
  • Observation dates from 1 January 2004 through 31 December 2016
  • In the Western Hemisphere and north of the equator
  • Complete checklists (all bird species detected and identified were included)
  • The primary checklist in a shared checklist
  • Checklists that used the generic traveling, stationary, or area protocols (i.e., not incidental or historical protocol)
  • If traveling checklists, were not longer than 15 kilometers
  • If area count checklists, did not cover more than 5626 hectares
  • Not longer in duration than 24 hours
  • Contained information on: start time, duration, protocol, number of observers, and distance traveled.
  • Counts of species were available (i.e., not just ‘present’)

See this link for more information about eBird checklists.

Can I contribute data for the Status and Trends Analysis?

Yes, any eBirder can contribute! These analyses were only possible thanks to the eBird submissions of hundreds of thousands of eBird users. Our ability to update and improve the Status and Trends products in the future continues to depend on the contributions of eBirders like you!

If you submitted checklists that meet all the requirements above, then you have already contributed data to the Status and Trends! Any future checklists you submit that meet these requirements will automatically be included in analyses for the updated Status and Trends Products.

Remember, any observation is useful, whether it is from today or your field notebooks from 15 years ago. Whether it is from a hotspot with amazing birds, or a place with few species – all checklists are valuable. To ensure your eBirding checklists are most useful to scientific efforts like this, you can make your checklists:

  • Complete checklists (i.e., record all species you were able to identify),
  • Provide a count or estimate of the number of individuals for each species
  • Use one of these protocols: traveling or stationary count.
  • Provide information on the start time, duration, number of observers, and distance traveled (The eBird mobile App now does many of these automatically with the new tracks)
  • Provide documentation of unusual sightings with descriptions or photos.

See our article on how to make your eBird checklists more valuable.

What environmental data were used for the Status and Trend products?

The analyses used to produce the Status and Trend products rely on matching up bird observations with certain environmental characteristics. For example, some bird species, such as Mourning Doves, are often seen in low elevation croplands, whereas other species may often be seen in higher elevation forests, such as Clark’s Nutcracker. To match up species to certain habitats, it is important to include good information on habitats in the analysis.

The analyses for the Status and Trend products use data on elevation, topography, and habitat to describe the local landscapes where eBirders searched for birds. Each checklist location is matched to the environmental data within approximately a 1.5km radius around the location. Elevation data described the elevation within the checklist radius, topography the aspect within the checklist radius (Amatulli et al. 2017), and habitat the proportion of the checklist radius that is comprised of 19 different land cover categories. These land cover categories were based on NASA satellite imagery (Friedl et al. 2010) and combined into cover categories by the University of Maryland (Hansen et al. 2000).

How are seasons defined for each species? Why are there gaps between seasons?

Breeding and non-breeding season dates are defined for each species as the weeks when the species’ population does not move. For this reason, these seasons are also described as stationary periods. The dates were defined by experts in the status and distribution of North American birds based on the weekly abundance maps. The selected dates were then checked to make sure that they generally matched expected patterns of phenology for the species.

Migration periods are defined as the periods of movement between the stationary non-breeding and breeding seasons. Note that for many species these migratory periods include not only movement from breeding grounds to non-breeding grounds, but also post-breeding dispersal, molt migration, and other movements. For some species, the transition between stationary and migratory seasons is not clear. Both breeding and non-breeding ranges are often represented within the migratory seasons since some individuals will have arrived in those areas while other individuals of the species are still migrating. In these cases transitional weeks were excluded to provides the clearest picture of individual seasons. For some species, this resulted in seasons that appear shorter than expected, especially when considered within specific regions.

Why are pre-breeding and post-breeding migrations sometimes separated?

Some species have pre-breeding and post-breeding migration seasons combined into a single migratory season. These species (e.g., Magnolia WarblerBlack-throated Gray Warbler) use fairly similar areas for both their migrations. However, some species such as Rufous Hummingbird use different paths for their two migrations. For these species we split the map to show pre-breeding migration (green) and post-breeding migration (yellow) separately. If at least 40% of the area used for one migration season is not covered by the other migration season, then we show them as distinct colors.

What is the difference between “modeled area” and “no prediction”?

On the relative abundance and range maps there are two separate colors of gray. The paler gray refers to areas of “no predictions” where there was insufficient data to assess whether the species was present or absent.

The darker gray shows the “modeled area”, where there was sufficient data to run a model, but the species was predicted to be absent. Sufficient data required at least 50 qualifying checklists within the region and a 30-day period.

Why are models available for North America only?

The Status and Trend products were limited to North America continent because it currently has the highest density of eBird data. In future releases we will expand the Status and Trend products to include more species across more regions around the globe.

What happened to Cuba? Why is it all blank?

You may notice that Cuba is defined as an area of “no prediction” on all the maps. The main reason Cuba has been excluded from is because there was insufficient eBird data from Cuba to ensure accurate results. We expect to include Cuba in future Status and Trend releases as we receive more eBird checklists from this region.

The predictive models used to generate the Status and Trend products account for gaps in eBird data by sharing information in nearby areas. This generally works well on the mainland, where our habitat data does a reasonably good job describing the ecological similarity of nearby areas. However, where there is a distinct island like Cuba, without much data from nearby and similar areas, it is hard for the model to share information. The model can extrapolate from the nearest areas, such as Florida, however, for many species this is not biologically sensible. For example, this can lead to errors when common species in Florida or Mexico are rare or absent in Cuba (e.g., Northern Cardinal, Black Vulture).

Note that the range data for Cuba is available for download in the raw GPKG file; please use this information with caution.

What are Bird Conservation Regions and how have they been defined?

Bird Conservation Regions (BCRs) are regions with similar bird communities, environmental conditions and resource management issues (Sauer et al., 2003). We used the established Bird Conservations Regions for the United States and Canada in accord with the NABCI Bird Conservation Regions, several of which span the border into Mexico. Because many of the Mexican Bird Conservation regions are too small to support the spatial scale of the Status and Trend analysis, we aggregated the Mexican Bird Conservation regions into larger regions in consultation with partners at CONABIO. Since Bird Conservation regions have not been defined in Central America and the Caribbean we defined regions based on countries, aggregating the smallest countries or groups of countries into units that were large enough to support the spatial scale of our analysis. See the map below for the BCRs that we use, including the customized versions for Mexico, Central America, and the Caribbean.

Some of the maps have errors? Why does this happen?

Like any predictive models, the Status and Trend models make errors when predicting species ranges and abundance. When predicting ranges there are two types of errors: predicting species absence in areas that are actually occupied, and predicting species presence in areas that are actually unoccupied. There can also be errors in the estimates of relative abundance, with estimates that are higher or lower than the actual counts.

The predictive models used to generate the Status and Trend products account for gaps in eBird data by sharing information from nearby areas. This works well when:

  • There is sufficient eBird data to capture patterns of species’ occurrence and abundance, and,
  • When the environmental data used in the models does a reasonably good job describing the ecological characteristics that are important to birds.

Therefore, error rates generally increase in regions when one or both of these conditions is not met. First, in regions where the density of checklists is low (e.g. central Canada, western Alaska, Panama, the Caribbean, and portions of the Great Basin deserts near Nevada) there is little information to learn patterns of species’ occurrence and abundance . In these areas, incorrect extrapolations can be a risk. For example, Black Vulture records from Western Mexico were extrapolated to Baja California where the species has virtually never been recorded. Second, error rates also tend to be higher when species’ detection rates are low. Even if there are many checklists in a region, having very few detections of a species limits the amount of information available to characterize the environment that the species is associated with. Third, error rates increase in regions where the environmental data fail to describe important ecological features for species.

See the link What are the “Predictive performance metrics”? for more information about how we assess the quality of model estimates.

Why do some products show as “Unavailable” or cannot be clicked on?

Species products may be missing for various reasons:

  • For some species (e.g., Chimney Swift, Franklin’s Gull) the main non-breeding area is outside the North American study area. We do not show non-breeding maps for these species.
  • For other species, the absence of range or abundance maps for a season indicates that model was poor and expert review indicated that the season should be excluded from visualizations and analysis. See section Some of the maps have errors? Why does this happen? For more information about prediction errors.
  • The “Percentage of total population in region” and “Percentage of range in region” statistics were excluded for species when experts judged that more than 25% of the range occurred outside the “modeled area”.
  • Finally, trend maps are presented for only subset of species during the breeding and non-breeding seasons, due to the greater volume of data required for these analyses. We will present trend estimates for a larger group of species in future releases.

Can I download the data?

The following products are currently available for download:

  • Spatial data for range boundaries can be downloaded as Geopackage (GPKG) files. The boundaries are available both as raw (directly from the analysis) and smoothed (as seen in the visualizations) range boundaries. See the Range section below to learn more.
  • All images of the range maps, abundance maps, habitat charts, and predictive performance plots can be downloaded and used for presentations and display.
  • The complete set of regional abundance and range statistics are available as a CSV file for download.

Raster data for the weekly abundance estimates will be made available for download in 2019.

All downloads are available on the Download page.  All downloadable products can be used for research, presentations, or display on webpages provided they are properly attributed; see our Recommend Citation below or here. Note that the downloadable range information available in the raw GPKG files includes results for Colombia, Venezuela, and Cuba which are excluded from the online maps. We recommend caution when using the products in these areas.

Recommended citation: Fink, D., T. Auer, A. Johnston, M. Strimas-Mackey, M. Iliff, and S. Kelling. eBird Status and Trends. Version: November 2018. https://ebird.org/science/status-and-trends. Cornell Lab of Ornithology, Ithaca, New York.

How can I find out more about the data analysis used to generate these products?

To generate the eBird Status and Trend products, we use statistical and machine learning models that combine the bird observation data from eBird with environmental data from NASA. The analytical process is also designed to correct for biases and gaps within the eBird data. You can read more about the methods in a number of technical scientific publications. For a more comprehensive description of the eBird database from a scientific perspective see Sullivan et al. 2014. The preprint by Fink et al. 2018 provides an overview of the data and models used for the Status and Trend products. The model used to estimate species’ relative abundance is described in Johnston et al. 2015. Details about the methods used to scale the relative abundance model across large spatial and temporal extents can be found in Fink et al. 2010 and Fink et al. 2013. Finally, a technical report with additional information on the specific methods used for the Status and Trend is forthcoming, so check back for the link!

Abundance

How is relative abundance defined?

Relative abundance was defined as the count of individuals of a given species we expect a single skilled eBirder to record if they conducted a hypothetical checklist traveling one kilometer over one hour at the optimal time of day for detection of that species. For each species relative abundance was estimated for all 52 weeks of the year across a regular spatial grid with a density of one location per 2.8km × 2.8.km grid cell across North America. Estimates at each location and date were made based on the local habitat, elevation, topography at that location.

Because detecting birds in the environment can be difficult, we know that there are always some individual birds that are missed by eBirders. For this reason, we refer to the quantity estimated as a relative measure of abundance. Although the relative abundance estimates will underestimate the true abundance, they do provide a standardized index that can be used to compare abundance in different regions. For example, if relative abundance is 10 in one area and 5 in another area, then we would estimate abundance is twice as high in the first area, even if we’re not sure of the actual number of individuals in the area.

Interpreting the Abundance Maps

The Abundance Maps show the species’ seasonally averaged relative abundance with each season stacked upon one-another. The sequence of seasons from top to bottom are:

Top layer: Year-round (purple) > Breeding (red) > Non-breeding (blue) > Pre-breeding migration (green) > Post-breeding migration (yellow): Lowest layer.

When the breeding and non-breeding seasons overlap this is shown as year-round. In most species, the two migration seasons have at least 60% overlap and are combined into one “migration” season (yellow); for an exception, see the Rufous Hummingbird example under “Interpreting the Range Maps”. Below is a guide to specific aspects of this Eastern Phoebe example map.

For this Eastern Phoebe abundance map refer to the original image to see the legend. Pale gray indicates “no prediction” areas (1, 2), such as northern Canada, areas of West Mexico, and Cuba. Areas shown in dark grey indicate where abundance is predicted to be zero (3), such as Nevada or Honduras. The most intense red indicates high breeding season abundance (e.g., around Ithaca, New York) (4). Paler red in Nova Scotia (5) indicates lower breeding season abundance; areas at the edge of the range consistently show lower abundance, such as eastern Colorado (6). Core winter areas, such as Florida (7), show high non-breeding season abundance, while areas of central and southern Mexico, such as Veracruz (8), reflect lower non-breeding season abundance. Eastern Phoebe occurs year-round in much of the southern United States, occurring at high abundance in Piedmont areas such as the Atlanta area of Georgia (9) and lower abundance in extensive agricultural areas, such as the Central Mississippi River Valley (10); this avoidance of agricultural areas is a pattern shown by many species. Eastern Phoebe occurs only as a migrant, both spring and fall, in areas shown in yellow (11), such as northwest Texas and central Mexico.

How is “Percentage of total population in region” calculated?

For each species and stationary season, we summed the relative abundance estimates across the selected region and then divided it by the sum of the relative abundance estimates across the entire seasonal range. The result is presented as a percentage.

This will be a reasonable estimate if the whole population is within the “modeled area”. For this reason, the reporting of this value was contingent on the majority of a species’ known seasonal range being within the modeled area. If the ‘percentage of total population’ value is missing, but the season was mapped in other products, it is because an expert reviewer believed that a significant portion of the species’ seasonal range was not included, but that the distribution within the modeled area was correct.

Note that since the state/province and BCR boundaries do not have offshore coverage, species with significant abundance offshore (e.g., Brown Pelican) do not include the offshore portion of their population in these calculations.

Trends

How do I interpret the trend maps?

Each map shows the average annual rate of change in the species’ population in the decade from 2007-2016. Significant increases in population size are shown in blue and significant decreases are shown in red. Darker colors indicate larger trends, either positive or negative. Areas shown in grey are those where the average annual rate of change was not clearly in one direction. This may be due to a lack of a strong population trend or insufficient data to estimate a reliable trend.

Each dot on the map represents a 25km x 25km area. The size of each individual dot on the map has been scaled according to the maximum abundance at that location during the 10-year study period. By visualizing trend and abundance information together in this way, regions that have experienced the largest change in population size stand out. If important species strongholds are declining, there will be dark red colors on large dots. For example, Wood Thrush during Breeding season shows large population decreases across the core of the range, where relative abundances are high and there are medium trends. At the same time, the Wood Thrush population underwent modest population increases on low-abundance range edges in Minnesota and Wisconsin.

Trend estimates are only calculated for those areas that have been occupied for at least seven years of the 10-year study period. Because trend estimation requires more data than abundance estimation, you may notice that some trend maps cover slightly smaller spatial extents compared to the range and abundance maps.

Trend maps are presented for a selection of species during the breeding and non-breeding seasons. For each of these trend maps, an extensive simulation study was performed to assess the likely accuracy of the trend estimates. For each trend map presented, we are confident that the vast majority (95%) of the colored dots will correctly identify population changes. Therefore the colored dots can be seen as reliable indicators of changing populations. The black contour line delineates the region where the population trend estimates are most reliable. Within this contour population trends of 1% per year would be identified at least half the time.

For this Wood Thrush trend map for breeding season refer to the original image to see the legend. As with abundance maps and range maps (see above), areas outside the prediction zone are shown in pale gray (1), such as Florida and northern Canada for this Wood Thrush trend map. Areas with a predicted abundance of zero (2), such as northwest Iowa and southern Louisiana, are shown as medium gray. Areas with low abundance of Wood Thrush are represented by small dots (3) while those with higher abundance are represented by large dots (4, 5). Those dots are whitish (3) when the trend is not significantly different from zero (note that these areas could actually have a slight increasing trend or decreasing trend). Areas with a declining trend are shown in red (5); the darker the red color the stronger the decline. Areas with an increasing trend are shown in blue (6). Areas within the black contour line (7) have the most reliable trend estimates.

Range

How are species’ ranges defined?

Species’ ranges were definedas the areas where the species is expected to occur on at least 1 out of 7of the predicted hypothetical checklists in a given week. This is equivalent to a single skilled eBirder traveling one km over one hour at the optimal time of day for detection of that species for each day of the week, and detecting the species on at least one checklist within a week.

Each species’ range was estimated for all 52 weeks of the year at 2.8km × 2.8.km grid cell locations across North America. Estimates at each location and date were made based on the local habitat and elevation at that location. Under this definition, species’ range is an estimate of area of occupancy (IUCN 2001).

To create easy-to-read range boundaries, the 2.8km grid data were aggregated to an 8.4km grid and spatially smoothed. Both the smoothed, aggregated seasonal boundaries and the raw 2.8km grid cell boundaries are provided for download.

Interpreting the Range Maps

The Range Maps show the species’ seasonal range boundaries with each season stacked upon one-another. The sequence of seasons from top to bottom are:

Top layer: Year-round (purple) > Breeding (red) > Non-breeding (blue) > Pre-breeding migration (green) > Post-breeding migration (yellow): Lowest layer.

When the breeding and non-breeding seasons overlap this is shown as year-round. In most species, the two migration seasons have at least 60% overlap and are combined into one “migration” season (yellow);  but in the Rufous Hummingbird example in this section the breeding and non-breeding seasons have less than 60% overlap and have been shown separately. Below is a guide to specific aspects of this Rufous Hummingbird range map.

For this Rufous Hummingbird range map refer to the original image to see the legend. Range boundaries are defined by the transitions between areas where a skilled eBirder could expect to encounter the species at least once out of seven bird walks of one kilometer for one hour in the area. No prediction zones, shown as the palest grey, such as northern Canada (1) and Cuba (2)  are areas where there was not enough information to define the seasonal range. Areas of dark gray (3), such as Iowa and the Yucatan Peninsula, indicate areas are outside of the range of Rufous Hummingbird. Breeding range is shown in red (4), such as British Columbia; non-breeding range is shown in blue (5), such as Jalisco. Note for Rufous Hummingbird, isolated non-breeding occurrence in cities and towns along the Gulf coast may be shown (6). The pre-breeding migration range appears in green (7), such as in much of California, and the post-breeding migration range appears as yellow (8), such as in Colorado; the separation of the migratory ranges indicates that the migratory range was at least 40% distinct between seasons.

How is “Percentage of region occupied” calculated?

Percentage of region occupied is calculated as the percent of the selected region that is covered by the range of the species. Note that since the state/province and BCR boundaries do not have offshore coverage, species with significant water-based abundance values (e.g., Brown Pelican) do not include the offshore portion of their population in these calculations.

How is “Percentage of range in region” calculated?

Percentage of range in region is calculated as the fraction of a species’ total North American range that falls within the selected region.

This will be a reasonable estimate if the whole population is within the “modeled area”. For this reason, the reporting of this value was contingent the majority of a species’ known seasonal range being within the modeled area. If the ‘percentage of total population’ value is missing, but the season was mapped in other products, it is because an expert reviewer believed that a significant portion of the species’ seasonal range was not included, but that the distribution within the modeled area was correct.

Note that since the state/province and BCR boundaries do not have offshore coverage, species with significant abundance offshore (e.g., Brown Pelican) do not include the offshore portion of their population in these calculations.

How is “Days of occupation in region” calculated?

Days of occupation in region is the number of days that a species is present in the selected region. A species is defined to be present in a region when least 5% of the region was within the species range during the given season.

Habitat

How do I interpret Regional habitat association charts?

The regional habitat association charts show the relative strength of attraction and avoidance of different habitat types for the selected region for all weeks of the year. Colored areas above the zero-line show attractive habitat types (i.e. positive associations), whereas colored areas below the zero line show habitats that the species avoids (i.e. has negative associations). The width of each color represents the relative strength of the attraction or avoidance of the given habitat type. Please see link to What environmental data were used for the Status & Trend products? for more information on the habitat data.

Often, when relative abundance is low, the habitat associations and avoidances become more variable and sometimes drop out completely. This happens at edge of species’ ranges or at the beginning and end of seasons. To help understand when relative abundance is low, the black line over the habitat chart shows the trajectory of weekly mean relative abundance. The trajectory has been scaled to reach a seasonal maximum value of 1.0 and is zero when the species is absent from the region.

Hover over the chart to see the numerical values describing the relative strength of habitat attraction and avoidance for the given week. These values describe the relative importance of the habitat within the 1.5km radius around the checklist location. Habitat types without consistent associations and land cover classes with values less than 1% are not shown, resulting in total contributions that may sum to less than 100%.

For this Wood Thrush habitat chart for Georgia please refer to the original image for the interactive tool. The black line (1) shows the the relative weekly abundance in the region. Population abundance and habitat associations are zero in non-breeding season (2) when Wood Thrush is absent from Georgia, having migrated to Middle America for the boreal winter. Refer to the legend to match a color to its habitat (3). Habitats that are most attractive for Wood Thrush in Georgia are Mixed forest (5) followed by Woody savannas (4). Urban developed areas (5) are attractive habitats for Wood Thrush only during the migratory periods. There are a number of negative predictors, including several aquatic habitat classes shown in blue, as well as croplands (7) shown in red. Note that croplands are not a significant negative predictor in breeding season (8), because Wood Thrushes breed in and near areas that include a mix of crops and forest within a 2.8km x 2.8km neighborhood. 

Appendix

What are “Predictive performance metrics”?

To assess and communicate the quality of the model estimates, we evaluated several metrics that together describe the model’s ability to predict the observed patterns of species occupancy and abundance. These metrics are made available primarily for interested researchers.

To quantify the performance of the range estimates we used the Area Under the Curve (AUC) and Kappa statistics to describe the models’ ability to correctly classify occupied and unoccupied sites (shown in the first row). AUC measures a model’s ability to discriminate between locations where species was detected and locations where the species was not detected. Technically, it is the probability that the model will rank a randomly chosen positive observation (species detected) higher than a randomly chosen negative one (species not detected). Cohen’s Kappa statistic was designed to measure the same metric, but taking into account the background prevalence. To quantify the quality of the relative occupancy predictions we also evaluated AUC and Kappa (shown in the second row). To quantify the quality of the abundance estimates we computed Spearman’s Rank Correlation (SRC) and the percent Poisson Deviance Explained (P-DE) (Shown in the third row). SRC measures how well the abundance estimates rank the observed abundances and the P-DE measures the correspondence between the magnitude of the estimated counts and observed counts.

References

Amatulli, G., Domisch, S., Tuanmu, M.N., Parmentier, B., Ranipeta, A., Malczyk, J. and Jetz, W. 2018. A suite of global, cross-scale topographic variables for environmental and biodiversity modeling. Scientific data, 5, p.180040.

Fink, D., Auer, T., Ruiz-Gutierrez, V., Hochachka, W. M., Johnston, A., La Sorte, F. A., & Kelling, S. (2018). Modeling Avian Full Annual Cycle Distribution and Population Trends with Citizen Science Data. bioRxiv, 251868.

Fink, D., Damoulas, T., & Dave, J. (2013, July). Adaptive Spatio-Temporal Exploratory Models: Hemisphere-wide species distributions from massively crowdsourced eBird data. In AAAI.

Fink, D., Hochachka, W. M., Zuckerberg, B., Winkler, D. W., Shaby, B., Munson, M. A., … & Kelling, S. (2010). Spatiotemporal exploratory models for broad‐scale survey data. Ecological Applications20(8), 2131-2147.

Friedl, M.A., D. Sulla-Menashe, B. Tan, A. Schneider, N. Ramankutty, A. Sibley and X. Huang. 2010. MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets, 2001-2012, Collection 5.1 IGBP Land Cover, Boston University, Boston, MA, USA.

Hansen, M. C., DeFries, R. S., Townshend, J. R. G., & Sohlberg, R. 2000. Global land cover classification at the 1km spatial resolution using a classification tree approach. International Journal of Remote Sensing 21:1331-1364.

IUCN. 2001. IUCN Red List categories and criteria: Version 3.1. Prepared by IUCN Species Survival Commission. World Conservation Union, Gland, Switzerland and Cambridge, United Kingdom. Ii + 30 pp.

Johnston, A., Fink, D., Reynolds, M. D., Hochachka, W. M., Sullivan, B. L., Bruns, N. E., … & Kelling, S. (2015). Abundance models improve spatial and temporal prioritization of conservation resources. Ecological Applications25(7), 1749-1756.

Sauer, J.R., Fallon, J.E. & Johnson, R. 2003. Use of North American Breeding Bird Survey data to estimate population change for Bird Conservation Regions. The Journal of Wildlife Management 67:372–389.

Sullivan, B. L., Aycrigg, J. L., Barry, J. H., Bonney, R. E., Bruns, N., Cooper, C. B., … & Fink, D. (2014). The eBird enterprise: an integrated approach to development and application of citizen science. Biological Conservation169, 31-40.