How to Access the Information

To view interactive maps showing survey and detection locations, species counts, as well as density, occupancy, and trend estimates by stratum, click the Explore the Data tab at the top of the page.

Selecting Filters

We designed the Rocky Mountain Avian Data Center (RMADC) to provide information for specific questions, and therefore, it works best when users select multiple filters for a query.

Tip: Look at the annual IMBCR field season report to determine the Stratum or Superstratum relevant to you. Under the Results section, click on the relevant agency or geographic region on the left sidebar. Then use the “On this page” menu on the right side of the page to choose a management unit of interest. The relevant Stratum or Superstratum will be listed in a callout block in the reported results for that unit.

Geographic filters

Bird Conservation Region: Selecting this filter will provide you with results for all strata and superstrata within a particular Bird Conservation Region.

State: Selecting this filter will provide you with results for all strata and superstrata within a particular state.

Land Management Agency: This filter will allow you to select results for All Other Lands, Colorado State Land Board, The Nature Conservancy (TNC), US Bureau of Indian Affairs (BIA), US Bureau of Land Management (BLM), US Department of Defense (DOD), US Fish and Wildlife Service (USFWS), US Forest Service (USFS), or National Park Service (NPS).

Stratum (Management Unit): This filter allows you to select results for a geographic or management unit. Strata refer to areas of interest for funding partners or the areas for which we are estimaing bird desnity, occupancy, and trend. A National Forest or a BLM field office are common strata. In the Stratum search box, type in the name of your management unit or scroll down in the pop-up box to find your stratum (they are organized alphabetically by state. You can view an interective StoryMap of strata locations here.

Superstratum (Region): This filter allows you to select results for states or regions containing multiple strata that were analyzed jointly (e.g., the entire state of Colorado which contains 30 individual strata). Please note, some management units represent superstrata if they have been stratified into smaller areas based on elevation, BCR/state boundary, or other features relevant for management. For example, the White River National Forest in Colorado is a superstratum because it contains 3 individual strata representing low, medium, and high elevation zones within the Forest.

Other filters

Species: This filter allows you to select results for an individual or multiple species.

Year: This filter allows you to select results for a particular year(s).

Viewing Maps and Counts (Map Tab)

What is displayed

The map displays all survey locations corresponding to your set of filters. If you have filtered by species, black circles represent survey locations where that species was not detected and pink circles represent survey locations where that species was detected.

By default, the zoom capability on the map is restricted to protect the privacy of private landowners. Funding and/or implementation partners wishing for precise location information should email Jennifer Timmer. Non-IMBCR partners should complete this Data Request Form.

The table below the map displays a count of the number of individuals detected on each individual sampling unit within a stratum, along with the survey date.

Adding map layers

Selecting a Bird Conservation Region or a Stratum in the filter boxes will display the boundary for that geographic extent.

Viewing Density/Occupancy/Trend Results (Density, Occupancy, Trend Tabs)

You may view the density, occupancy, or trend estimates tables for all appropriate strata, based on your filter selection, for which we have results, by clicking on the tabs labeled “Density”, “Occupancy”, or “Trend Estimates”.

The density tables display stratum, species, year, number of birds estimated per km² (Density), percent coefficient of variation (% CV), the 95% credible interval (95% CI), the number of independent detections used in analyses (Detections), the number of surveyed grid cells within a stratum in a particular year, and the estimated total abundance of birds within a stratum. The occupancy tables display stratum, species, year, Occupancy (proportion of plots expected to be occupied by the species in the stratum), the percent coefficient of variation (% CV), the 95% credible interval (95% CI), and the number of surveyed points on which the species was detected.

The trend tables display stratum, species, the monitoring period (Years), and the median trend estimate, confidence in the direction of the trend (f), and the 95% credible interval (95% CI) each for density and occupancy.

Superstratum and Stratum Estimates

Density, occupancy, and trend estimates will always be displayed by individual strata in the top table and by superstrata in the bottom table. If, for example, you use the “Stratum” filter to select an individual management unit, estimates for all relevant superstrata will be included in the table below for regional context. If, for example, you use the “Superstratum” filter to select a region or state, estimates for all individual strata contained within that superstratum will be included in the top table.

Downloading the Information

You may click the Download button below each table to download species counts or density/occupancy/trend estimates by strata as a CSV file.

Viewing the Charts (Density and Occupancy Only)

When viewing the occupancy and density charts, the occupancy or density estimate for each year is indicated with a circle, and the error bars represent the 95% credible interval.

How to Interpret the Results

Density

Density is the number of birds per square kilometer within a stratum. Total Abundance is the total number of individuals within a stratum. It's the density estimate multiplied by the stratum area. Density is useful when you need to know how abundant a species is within a particular state, region, or management unit (stratum). The 95% CI is the upper and lower bounds containing the true population estimate with 95% certainty.

Occupancy

Occupancy is the probability that a 125-m radius plot contains at least one individual of the species. You can also think of it as the proportion of plots within a stratum that contain the species or how widespread the species is within a stratum. For ~25% of the species we include in the analysis, occupancy represents the probability that a 989-m radius plot contains at least one individual of the species (e.g., raptors, corvids).

Trend

We report trend as the percent population change per year based on all years a stratum was surveyed. We estimate population change based on both the density and occupancy estimates. Declining population trends are highlighted in red, while increasing population trends are highlighted in blue. Although density and occupancy have different units of measure, trend on either is interpreted the same.

How to Determine Informative Estimates

Density and Occupancy

The coefficient of variation (CV) is an indicator of precision for density and occupancy estimates. We classify density and occupancy estimates as informative if their CV is < 50%, which is highlighted in yellow in the tables.

Trend

The f value is the probability that the population trend is in the direction of the reported percent change per year (i.e., positive or negative). You can think of it as our confidence that a population has truly increased or decreased within a stratum. For example, an f value of 0.98 means we are 98% certain the population has increased or decreased within the stratum. We classify trend estimates as supported if they have an f value > 0.9, which are highlighted in yellow. For partners that wish to use a more stringent cutoff for even higher confidence, we also highlight f values > 0.975 in gold.

We also report the 95% credible intervals to further determine if a trend estimate is supported. Credible intervals (95% CI) are the upper and lower bounds containing the true population trend with 95% probability. 95% CIs which do not overlap zero are equivalent to f > 0.975 (highlighted in gold). Note that trend estimates with an f value between 0.9-0.975, although supported, will have 95% CIs that overlap 0.

Stable populations have a trend estimate of approximately 0, which means the population hasn’t increased or decreased each year. Stable trends will also have a lower f value.

IMBCR Estimates FAQs

Below are some frequently asked questions about the IMBCR population estimates we provide each year. You may also visit the "Tutorial" section on the Rocky Mountain Avian Data Center for an explanation on how to query and interpret density, occupancy, and trend estimates from the database. Refer to section 6.3 and Appendix D in the annual IMBCR report for additional details on the annual analysis and estimates.

DENSITY & OCCUPANCY

Q1) When would I use a density estimate vs an occupancy estimate?

Density is the number of birds/km2 within a stratum, which is abundance per unit area. Occupancy is the probability that a 125-m radius plot within a stratum contains at least one individual of the species. You can also think of occupancy as the proportion of plots within a stratum that contain the species or how widespread the species is within the stratum. For most species, density and occupancy will probably be closely related; a species that is more abundant will also be more widespread.

Density is useful when you need to know how abundant a species is, possibly for project-planning to determine a population impact (e.g., you can multiply the density by the project area to estimate how many individuals could be impacted by the project). Occupancy may be more useful for tracking less common species of concern because it requires fewer detections to estimate than density. Occupancy is also a useful metric to describe the distribution of a species within a state, region, or management unit. Additionally, trends in occupancy can be more interpretable as changes in species range, whereas trends in abundance may primarily reflect changes in population size.

Q2) Can I still use a density or occupancy estimate if the CV is >50%?

On the RMADC, we classify density and occupancy estimates as informative if their CV is <50% (CV stands for coefficient of variation and is a measure of precision). This is a somewhat arbitrary decision because there's no statistical threshold for a low or high CV. Thus, this means that estimates with a CV >50% aren't necessarily uninformative. An estimate might have a larger CV one year because we had fewer detections within the stratum or fewer points on which it was detected, so we're less certain of the true population density or occupancy compared to an estimate with a smaller CV. Remember, we're estimating population size by sampling across a stratum because a complete census to count the actual population size is not practical.

You can also look at the 95% credible interval (CI) to see what the range is for the true population density or occupancy (a 95% CI is the upper and lower bounds containing the true population change with 95% certainty). A wider interval means we're less certain of the true population size. If you have a population target in mind for a management plan or decision, see if the 95% CI includes that target.

If you want to use or report the most recent density or occupancy estimate and the CV is >50%, that's ok. Just be sure to include the CV and 95% CI for context.

Q3) How can you produce a density or occupancy estimate for a stratum when the species was not detected in a particular year?

One of the advantages of IMBCR is that we can leverage detections across the program to inform detection probabilities in smaller strata with fewer detections for a particular species. This is especially beneficial for less-common species. This also means that we can provide a density or occupancy estimate when a species has been detected in a stratum in a previous year, but not in the current year. In these cases, the estimate will likely be zero or very close to zero. We include them on the RMADC because we consider them to be legitimate estimates of density or occupancy because the species occurs in the area of interest, but simply wasn't detected that particular year.

Q4) Is it ok to average density or occupancy estimates over time to inform project planning?

Yes, it is ok to average the density or occupancy estimates across years. This likely will give you a more accurate idea for population size within a stratum, especially for irruptive species, like Evening Grosbeak, and grassland birds that have more variability in population size year to year. However, you can't average the CV's or 95% CI. You can describe the overall precision of estimates by considering the range of CVs. For example, if most of the CV's for density estimates were around 40-60%, and if you consider 50% to be moderately informative, you could consider the density estimates for this species in this stratum to be moderately informative.

TRENDS

Q1) What's the difference between trend based on density & trend based on occupancy?

Both are population change/year. The former describes change/year in density (birds/km2) and the latter is change/year in point occupancy (proportion of plots where the species occurs). Although density & occupancy have different units, trend on either is interpreted the same: the percent change/year within a given stratum. Trend based on occupancy may be better for less common species with fewer detections because occupancy requires fewer detections to estimate. Supported trends are those for which we are at least 90% certain are increasing or decreasing (i.e., f >0.9). Trend based on density will often be similar to trend based on occupancy, especially if both estimates have a higher f-value (f is our confidence that a population is truly increasing or decreasing within a stratum). However, it is entirely possible to have a decreasing trend based on density and an increasing trend based on occupancy, if for example, a species' distribution within a stratum is changing.

Q2) How big is too big of a decline? When should I be worried?

The first step to understanding and interpreting trends is to look at the annual density or occupancy estimates that the trend is summarizing and the number of detections recorded in each year. This will help you evaluate whether the species has declined consistently from year to year or if there was a major drop in just the most recent year and that's what's driving the trend. For example, a stratum may be located on the periphery of a species' range and we detected a few individuals of that species in one year and then haven't detected any since. This will likely result in a supported declining trend, but if we detect more individuals in later years, the trend could reverse. On the other hand, if we consistently have 3-4 detections/year of that species for 5 years and then none the following 5 years, there could be something actually going on with that population.

Secondly, it's important to consider natural variability. A species may have a large decline one year, but then a minimal decline or slight increase the following year due to natural variability. Irruptive species, like crossbills, may be very abundant one year but absent the following year because of poor cone crop production. Grassland birds are also notorious for moving around year-to-year to track precipitation and food resources, so it's important to see if trends are consistent in subsequent years for these types of species.

Lastly, what are your organization's management objectives? Those will dictate your priorities and how much risk you're willing to take. For example, if you work for a state wildlife agency and are responsible for the viability of wildlife populations within your state, then you may be concerned with any population decline even if it's small (e.g., a 1% decline/yr). However, if you work for a land management agency and managing wildlife populations is just one of many objectives, you may not be as concerned until the population decline is larger. It could be helpful for your organization to set a management objective describing when to intervene or change management strategies or policy based on population trend. If you're interested in developing management objectives based on your organizational goals and would like some guidance on the process, contact Jen (jennifer.timmer@birdconservancy.org) to discuss options.

Q3) What if the trend estimate is positive one year, then negative the following year?

Some species have greater interannual variability as they track food or nesting resources, so they may be absent or fewer in number within a particular region. This could cause their trend estimate to change direction over time. Additionally, some strata may occur on the periphery of a species' range, so it may be detected once and then not again for several years, resulting in a declining trend. Then if it's detected in later years, this could result in a positive trend. However, the stronger a trend (i.e., the larger the increase or decrease) that also has a high f-value (i.e., f >0.9), and the more years it's consistently in one direction, the less likely it is to flip direction.

Q4) How useful is a trend estimate that's based on <10yr of monitoring data?

Again, this depends on your objectives. If you want to evaluate the effects of management or conservation activities on birds that occurred within a specific management unit (i.e., stratum) during a shorter time period, then a 5-10yr trend might be relevant. If, however, you wish to update a longer-term management plan, like a Forest Plan or State Wildlife Action Plan, then a longer trend (10+years) is probably more useful. Also, we only produce trends for strata that have been surveyed for at least 3yr.

FOR FURTHER GUIDANCE ON WHEN TO PAY ATTENTION TO A TREND WITHIN A MANAGEMENT UNIT (STRATUM) OR REGION (SUPERSTRATUM), REFER TO THE CHECKLIST BELOW:

• Check to see if the trend is supported and we're confident about the direction of change. If the f value is >0.9, it means we're at least 90% certain of the direction of population change. Trends with f > 0.9 will be highlighted in yellow on the RMADC. Look at both trend based on density and trend based on occupancy -- they're interpreted the same way (% change/year). If both have an f value >0.9, it's ok to focus on one or the other (use trend based on density if you're more interested in how the abundance of a population has changed and use trend based on occupancy if you're interested in how the distribution of a population has changed).

• Look at the 95% credible interval to see how certain we are about the magnitude of change. This interval means we have captured the true population trend with 95% certainty. Wider intervals simply mean we're less certain about how much a population has increased or decreased during the monitoring period. 95% CI that do not include 1 correspond to f values >0.975 and these estimates are highlighted in gold on the RMADC. In these cases, we are certain a population is increasing or decreasing because the interval is entirely above or below 1.

• An estimate may be statistically significant (i.e., has an f value >0.9), but is it biologically significant? This is where the Density and Occupancy tables on the RMADC come in handy. In the Density table, see how many detections we have for the species each year in the stratum. If we didn't detect it at all for several years and then detected a handful of individuals in recent years, this could result in a supported increasing trend. Similarly, we may have a few detections in early years but haven't detected it recently, which could result in a supported declining trend. Either scenario could reflect a real population increase or decrease, but it's less certain when it's based on just a few detections (e.g., fewer than 10 total detections). In the Occupancy table, see on how many points the species was detected. If it was detected on fewer than 5 total points, this could mean the stratum is located on the edge of a species' range or the species does not occur throughout the stratum, so we're less certain about its population trend.

• In addition to looking at the number of detections and number of points on which a species was detected, also look at the density and occupancy estimates themselves. For example, if a species has density estimates that consistently increase over time and then the estimate decreases greatly in a recent year, this could cause a decreasing trend that may not be biologically significant. We may have a reduced sampling effort that year and simply detected fewer individuals. It's also important to think about a species' biology when interpreting estimates. Irruptive species like Red Crossbill or grassland birds like Lark Bunting tend to have more interannual variation in estimates because they move around in search of food and nesting resources. Density or occupancy estimates that go up and down year to year are less likely to result in a supported trend.

• Below are a few examples that might be helpful when interpreting trends:

Biologically & statistically supported trend

In the Trend table, we have a supported, increasing trend based on density for Northern Flicker in San Juan National Forest (Colorado). The f value is 1.0 and the 95% CI is 2.2-9.2, so we're 100% certain this population increased from 2008-2025, and it increased between 2 and 9% each year. In the Density table, in 2008, we estimated 1.7 Flickers/km2 and in 2025, we estimated 6.6 Flickers/km2. We consistently detected >10 Flickers in most years and they were detected on 3-27 points over time (Occupancy table). San Juan NF also occurs within the range of Northern Flickers and contains habitat components that support Flickers.

Statistically supported trend that's not biologically meaningful

We have a supported, increasing trend for Gray Flycatcher in San Juan NF. The f value is 0.99 and the 95% CI is 2.0-37.8, so we're 99% certain this population has increased from 2008-2025, and it increased between 2 and 38% ea/yr. However, we've only detected 8 Flycatchers total from 2008-2025 (0-3 detections/yr) and on 0-3 surveyed points. The San Juan NF is on the eastern edge of the Gray Flycatcher's range and this mostly high-elevation Forest does not contain much of the Flycatcher's preferred breeding habitat components, like sagebrush, desert shrub, and dry open woodlands.

Stable population trend

Some species may have a trend that's very close to 1 indicating a stable population. In San Juan NF, Chipping Sparrow has a slight decreasing trend based on density (-0.57%/yr) and a slight decreasing trend based on occupancy (-1.06%/yr), and the f-values are 0.57 and 0.68, respectively. Note, the f-values will not be high if the trend estimate is close to 1 because the population is not likely increasing or decreasing. Therefore, it's helpful to look at the density and occupancy tables. For Chipping Sparrow, we had numerous detections each year in the San Juan NF and the density estimate in 2008 was 10.9 birds/km2 and 9.2 birds/km2 in 2025. Chipping Sparrow was also detected on numerous surveyed points in the San Juan NF and the occupancy estimate in 2008 was 0.13 and 0.1 in 2025.

NEW RMADC FEATURES IN THE WORKS

Please note, the features below are on our biometricians' agendas, but we need additional funding to complete them and host them on the RMADC. If you have available funding to assist with a feature, please reach out to Jen.

Q1) Can you provide density estimates by specific habitat types?

We don't currently have this feature on the RMADC but we are working on it. This feature would allow you to estimate density for a primary habitat for a given stratum and year. At each survey point, technicians record primary habitat, like mixed conifer, grassland, and sagebrush shrubland. The analysis for this feature would only use survey points recorded as a particular habitat type to estimate density for that habitat type. This feature will be particularly useful for habitat-specialists (I'm looking at you, Sagebrush Sparrow) or for project planning where a project is expected to impact a particular habitat type within a stratum.

Q2) Can I get trend estimates for a specific time period to reflect when we were doing habitat treatments within a Forest?

We don't currently have this flexible trend estimator on the RMADC but are working on it. This feature would allow you to specify the time period for generating a species' trend. For example, if you revised your Forest Plan in 2010, you could adjust the start year from 2008 (when we began monitoring) to 2010 to generate a population trend for 2010-2025 to better match the life cycle of that Plan.

Q3) Do you have a map showing where a species is increasing or decreasing within a region?

We don't currently have this feature on the RMADC but we are working on it. We plan to include maps that will provide a visual for trend estimates across a state or region. Each stratum boundary within the state or region will be color-coded to reflect an increasing trend (blue) or a decreasing trend (red) and the thickness of the outline will signify the magnitude of the change (e.g., a trend >10% will have a thick blue border compared to a 3% increase). Users will be able to export this map.