Ozone Biomonitoring at Purchase Knob

Great Smoky Mountains National Park


About the Ozone Biomonitoring Garden

Ground-level ozone exists in the lowest layer of the atmosphere. High levels of ground-level ozone are produced when nitrogen oxides and sulfur dioxide are emitted into the atmosphere, and ultraviolet radiation from the sun causes them to react and produce ozone. Ozone can damage our lungs and causes ecological damage. It causes physical and physiological injury in plants, and the clearest signs of damage are the purple spots, also called stippling, on their leaves. As sensitive plants are exposed to more ozone, the purple color will spread across their leaves. You can identify ozone damage from other issues with the leaf because ozone injury does not cross over any veins.

An image with of serrated leaves with injuries on them. Some leaves are withered and dead and
    there are three red, labeled arrows. An arrow is pointed at purple areas and is labeled "stippling." The arrow pointed at a yellow area of
    a leaf is labeled "chlorosis," and the third arrow is poined at a dead leaf and is labeled "necrosis."
Photo | National Park Service


To monitor the effects of ozone on plants in Great Smoky Mountains National Park, researchers established an ozone biomonitoring garden at Purchase Knob in 2003. The garden acts as a tool to monitor the ecological effects of ozone, as well as a community science project in which high school students contribute to a long-term ecological dataset.

Purchase Knob is a high-elevation site (~1500m above sea level) that sits at the eastern edge of the National Park, shown below.

Photo of eight people collecting data on crownbeard in a garden. Most people are looking at plants and holding clipboards with
          paper. Photo from the National Park Service.
National Park Service

Yellow crownbeard (Verbesina occidentalis) is an herbaceous perennial plant that is native to the Great Smoky Mountains. While the stems only last one growing season, the same root balls will put out new stems each year. The species is sensitive to ozone, making it a good indicator species for the early signs of ozone-related damage in the ecosystem. Another highly sensitive species in the garden is Sochan (Rudbeckia laciniata) also called Cut-leaf Coneflower. This is a culturally significant plant to the Aniyvwiya (Cherokee) people eaten and used as an early spring tonic.

During the growing season, high school interns, park staff, and school groups collect data on the plants. At a nearby weather station, sensors collect hourly data on ozone levels, temperature, and humidity.

Use the scroll bar below to see how leaf injury changes over the season for this plant stem from 2012. The yellow (first) bar represents chlorosis, the purple (second) is purpling/ozone injury, and black (third) is necrosis. Blank leaves are missing, since they have either fallen or have not grown yet
Photo of yellow crownbeard. Three tall stems full of leaves, all close together, with small yellow flowers.
          Photo by Richard Spellenberg.
Richard Spellenberg


Choose an Observation Date

Figure adapted from Chen et al. 2017. Ozone Concentration and Foliar Injury Analysis at Purchase Knob Garden. National Center for Atmospheric Research.

Ecological Trends and Correlations

Ecological and biological data are often very messy. Individual organisms are unique in their genetic makeup and developmental path, so they will respond differently under the same circumstances. The degree of difference between individual organisms determines how clearly you can see a relationship between cause and effect. The more differences there are between individuals, the harder it is to see the trend with our eyes, though there are statistical methods to account for some of the "noise." Trends that are noisy are said to have low correlation (correlation, or R, is closer to 0). Less noisy trends, where the data cluster tightly around the trend line, are said to have high correlation (R closer to 1). It is easier to make conclusions about data with higher correlations, but data with lower correlation can be harder to interpret.

Use the slider in the following plot to see what trends look like at different levels of correlation. The plot shows simulated data with a predictor variable (x axis) that affects the response variable (y axis). When you select an R value, 100 data points will be generated with that R value.

Choose an Correlation Value



Exploring Data from the Biomonitoring Garden

Instructions for the interactive panels
The panels below show data from the ozone biomonitoring garden from 2004 through 2023. Foliar injury data are from yellow crownbeard, which is the only species that has been consistently monitored every year.

The last tab in each panel is an informational tab. Click the i icon to see more about the variables in the plot.

Use the widgets on the left-hand panels to select the data shown in the panels. In each panel, you can select the data tab to look at the data being shown in the plots and download the data being displayed.

Note on the trend tines
In the first two sections below, the trend lines are the trends predicted by models that use all the data. When zooming in on those plots, the trends will stay the same, since they are based the whole dataset. In the last section, trend lines will refresh as you change the input data points, as they are based only on the data currently shown in the plot.

The plots show R2 instead of R correlation). R2 is used more frequently by statisticians because it is better at describing the amount of "variance" predicted by the predictors (in this case, how well ozone explains the variation in foliar injury).

Annual trends in ozone exposure and foliar injury

Select Data



Ozone Levels

4th-highest 8-hr average
The average ozone level for every 8-hr period is calculated throughout the year. This is the 4th highest value.

W126
A weighted average of hourly ozone exposure that weighs high values more strongly (daylight hours, 8am-8pm).

AOT40
The accumulated hourly ozone exposure (sum hourly values) for concentrations above 40 ppb (daylight hours, 8am-8pm).

SUM06
The accumulated hourly ozone exposure for concentrations above 0.06 ppm (24 hours).

M12
The mean hourly ozone exposure during 12 daylight hours (8am-8pm).

Foliar Injury

Ozone exposure produces purple spots on leaves, and if ozone levels are high, eventually the spots will cover the whole area of the leaf.

Injury Amount
The proportion of leaves on the plant that have any amount of ozone foliar injury (e.g., 0.125 if 2 out of 16 leaves had purpling).

Injury Severity
The average proportion of leaf area covered in ozone-produced purpling on a plant (e.g., 0.4 if there are two leaves, one with 20% purpling and one with 60% purpling).

Does high annual ozone exposure produce more foliar injury?

Select Data


Annual Foliar Injury vs. Ozone Exposure

The data shown here are the same as the data from the previous set of plots. They are summaries of foliar injury and ozone exposure by the end of the growing season.

Ozone Metrics
4th-highest 8-hr average
The average ozone level for every 8-hr period is calculated throughout the year. This is the 4th highest value.

W126
A weighted average of hourly ozone exposure that weighs high values more strongly (daylight hours, 8am-8pm).

AOT40
The accumulated hourly ozone exposure (sum hourly values) for concentrations above 40 ppb (daylight hours, 8am-8pm).

SUM06
The accumulated hourly ozone exposure for concentrations above 0.06 ppm (24 hours).

M12
The mean hourly ozone exposure during 12 daylight hours (8am-8pm).



Foliar Injury Metrics
Injury Amount
The proportion of leaves on the plant that have any amount of ozone foliar injury (e.g., 0.125 if 2 out of 16 leaves had purpling).

Injury Severity
The average proportion of leaf area covered in ozone-produced purpling on a plant (e.g., 0.4 if there are two leaves, one with 20% purpling and one with 60% purpling).

Does dry weather protect plants from ozone-induced foliar injury?

Select Data




Changes between observations
The data in this section represents the changes in foliar injury, ozone levels, and weather between observations. The previous sections used annual data (i.e., the last foliar injury observed on each stem and cumulative ozone from the growing season). This dataset includes multiple data points for the same plant stem, each representing the differences between two observations of the stem.

Variables
The ozone exposure metrics on the x-axis are the accumulated ozone since the last observation, so the values for each day after the previous observation are added together.

The foliar injury data are differences between the current observation and the previous observation of the same stem.

Accumulated growing degree days (AGDD) are one way of summarizing temperature across time. It averages the amount of heat above a temperature threshold, in this case 10°C (50°F). For each observation in the dataset, the AGDD is the sum of the growing degree days for all the days since the last observation. Learn more about AGDD from the National Phenology Network.

Vapor pressure deficit (VPD) is a way of summarizing humidity, relative to temperature. It is the difference between the maximum amount of moisture that the air can hold and the actual amount of moisture in the air. The saturation point increases with temperature, so warm air can hold more moisture than cold air. The dataset includes the average VPD from the previous observation to the current observation.