When a Giant Falls: Tracking New Growth in Amazon Tree Fall Gaps

What happens in the rainforest when a giant tree comes crashing down? Through the Amazon Research Initiative for Educators (ARIE), K-12 teachers stepped into the role of field scientists at the Amazon Conservatory of Tropical Studies (ACTS) in Peru, measuring how the understory responds when a break in the canopy lets sunlight pour in. Now in its second year, their study found that ground cover at the center of each gap jumped by more than 2,300%, with growth fading quickly toward the shaded forest edge. By collecting real data on an under-studied process, these educator-researchers are revealing tree falls as powerful engines of forest renewal and bringing authentic Amazon science back to their classrooms.

2026 Field Report: Tree Fall Plot Study (PDF)

Date:  July 13-17, 026

Location: Explorama’s Napo-Sucusari Biological Reserve, Loreto Peru

Study Site: The Amazon Conservatory for Tropical Studies (ACTS).  Forest floor.

Lead Researcher: Dr. Jeannine Richards

Team Members: Erika Stafne*, Sarah Wages, Leovie Diaz, Katy Kuei, Debby Prestridge

(* = author of this report)

Lay/Marketing Summary

When a tree falls in a rainforest, it creates an open space, allowing light to reach the forest floor. This project measures how quickly plants respond to the newly available light, both as ground cover and canopy cover, by measuring how quickly they grow.

Abstract

When a tree falls, it creates an opportunity for new growth in the understory of the Amazon Rainforest. The forest canopy is very dense, with limited light for understory plants, until a piece of the canopy is gone. These events are under-studied in this ecosystem, and the response and growth have not yet been documented. Our goal is to understand the response rate and abiotic factors that are associated with a tree fall gap and how they change over time. This is a multi-year study that was started one year ago. Using 15 meter transects extending laterally from fallen tree trunks (0 m) into the surrounding forest (15 m) across three tree fall sites, we collected multi-year data (2025–2026) on percent ground cover, canopy cover, and secondary growth (diameter at breast height – DBH). Beginning in 2026, data was also collected on leaf Brix sugar concentration, microclimatic parameters (lux, temperature, humidity), and lepidopteran community composition. 

The results show a strong, consistent inverse relationship between distance from the gap epicenter and plant growth rates. Average percent ground cover change spiked at 0 m (exceeding +2,300%) before declining sharply toward baseline forest levels at 15 m. Canopy cover changes shifted from maximum net positive growth at the gap epicenter to a loss at the shaded edge. Similarly, trunk secondary growth (DBH) exhibited a negative correlation with transect distance, driven by light attenuation from over 210 lux at gap centers to under 75 lux at shaded ends. Leaf Brix readings confirmed higher sugar concentrations in open canopy conditions (4º Bx) compared to shade (1º Bx), corroborating elevated photosynthetic activity. Lepidopteran trapping data was inconclusive due to trap height inconsistencies. Overall, these findings highlight the critical role of canopy gaps as localized engines of secondary succession and biomass production in Amazonian understory ecosystems.

Introduction Light is a limiting factor in a rainforest ecosystem, with the understory typically receiving very little sunlight. Few studies have been conducted on how plant growth changes when light becomes suddenly available in greater quantities. Tree falls provide an opportunity for a significant increase in sunlight reaching the forest floor, as a large opening in the canopy results, and tree falls happen frequently throughout the Amazon Rainforest. This project investigates the changes that occur in plant growth after tree falls at the Amazon Conservatory for Tropical Studies (ACTS) field station. Studying the changes in plant growth after a tree fall can help us understand general rainforest ecology, recovery of areas after a disturbance, and long term competition within the forest understory and canopy. Documenting these changes will help us to comprehend what is actively happening in the forest. The goal of this project is to measure the rate of plant growth in response to a light gap created by a tree fall. Our assumption is that the rate of plant growth will be highest in the center of the light gap, nearest the trunk of the tree fall, and will decrease as we move along the gradient away from the tree. We also expect that there will be a higher temperature and light intensity, and higher rate of photosynthesis, as indicated by a higher sugar content, in leaves collected from the open sunlight as compared to the shade, where there will be higher humidity. Additionally, we believe there will be more butterflies and larger winged organisms caught in the open sunlight area and more moths and smaller winged organisms caught in the shaded understory.

Materials and Methods

In July, 2025, two tree fall sites (Tree Fall 1 and 2) were selected where trees had fallen recently. Data was collected at both sites again in July, 2026, as well as at a third site of a newly fallen tree (Tree Fall 3). At each site, a meter tape was stretched from the broken end of the fallen tree along the length of the trunk up to 50 meters, depending on the length of the trunk. At Tree Fall 1 and 2, a random numbers generator was used to determine the locations along the trunk of three sampling transects. This was repeated at Tree Fall 3. At each randomly generated sampling point along the trunk, a 15 meter transect line was laid out perpendicular to the trunk. At 0, 5, 10, and 15 meters along this tape, data was collected (Figure 1).

Percent ground cover measurements were collected using a square-meter pvc quadrat placed centrally on the transect tape and between 0-1, 5-6, 10-11, 15-16 meters. Held at waist height, two researchers independently assessed coverage and came to consensus using increments of 10% (Figure 2). Additionally, a spherical densiometer was used to take canopy cover measurements. Using the spherical densiometer, four photos of the canopy cover were taken at 90º intervals. The spherical densiometer is a hand-held convex mirror marked with a grid pattern containing 96 squares. To calculate canopy cover, each square in the grid is divided into four sections and the amount of reflected cover in the square is recorded with a range of 0-4. Those numbers are added up and multiplied by 1.04 to calculate overstory density percentage (Figure 3). Trees with a DBH over 1 cm located within 0.5 m of each side of the transects had been marked with orange flagging tape. These trees were measured for secondary growth (DBH) using a DBH tape and any new trees that had reached 1 cm in DBH were measured and added to the data (Figure 4). If a branch was reachable from a tree it was also measured for primary growth, with pink flagging tape marking both ends of the branch. A tape measure was used to measure the length of the marked branches. Trees over 1 cm in DBH and within 0.5 meters of the transect lines of Tree Fall 3 were marked with orange flags; their branches marked with pink flagging tape. Light sensors (HOBOS) were retrieved from sites 1 and 2, but no new sensors were placed. These methods were repeated at three separate tree falls and will be repeated at these sites every year.

At Tree Fall 1 and 3, at each sampling point, a microbit sensor was used to collect temperature, humidity, and light intensity and a PocketLab sensor was used to measure heat index, temperature, humidity, and light intensity (Figure 5). Data was collected for two minutes at each sampling point along each transect and then averaged. One representative transect line was chosen for each Tree Fall site to demonstrate overall data trends in the table below (Table 17).

At Tree Fall 1, Pop-up Butterfly Bait Traps were placed in the shade at ground level around the 15 meter mark on the transect line (Figure 6). A second net was placed in the sun at the 0 meter mark, significantly higher than the other net due to branch limitations. A mixture of ½ banana to 10 ml of locally brewed cane sugar rum (aguardiente de caña) was mashed and placed below the net. Nets were set in the morning and checked every 24 hours for three days, with fresh banana mash placed after 48 hours. Moths and butterflies were counted, collected and measured.

At Tree Fall 1, a Brix refractometer was used to calculate sugar levels in leaves to measure photosynthesis (Figure 7). Leaves were collected from similar plants at the 0 meter sample site (open canopy/sun) and 15 meter sample site (more closed canopy/shade). Cells in the leaves were broken by crushing them in a plastic bag with a rock. The liquid was refracted and placed on the refractometer. Two replications were completed with both samples.

Results

Ground Cover: Across all data sets, the average percent change in ground cover is inversely proportional to the transect distance from the gap center (0 m). The maximum growth spike occurs uniformly at the center point, exceeding +2,300% on average, followed by a steep, exponential decline as distance increases away from the tree trunk (Figure 8). In Tree Fall 1, the percent change drops from+3,600% at 0 m to 0% at 15 m (Figure 9). Diverging from Tree Fall 1, Tree Fall 2 records its lowest growth spike at 10 m (+60%) before experiencing an upward trend at 15 m, rising back up to a +100% (Figure 10).

Canopy Cover: Across all data sets, the data generally show an inverse relationship between the change in canopy cover and the distance from the disturbance center (0 m), transitioning from net positive growth near the center to a net loss at 15 m (Figure 11). CIn Tree Fall 1, average canopy cover growth decreased steadily along the transect, dropping from a peak of +28% at 0 m to a -6% at 15 m (Figure 12). Tree Fall 2 diverges from this linear downward trend, peaking at 10 m with a +13% increase in canopy cover while maintaining lower positive growth at 0 m (~+6%) and 5 m (~+3%), before dropping to -6% at 15 m (Figure 13).

DBH: Across all three datasets, the scatter plots trendlines demonstrate a consistent negative correlation between the transect distance from the gap center (0 m) and annual secondary growth (change in DBH from 2025 to 2026). As the distance along the transect increases from roughly 3.5 m out to 15 m, the overall magnitude of trunk diameter growth decreases toward zero (Figures 14-16).

Microclimate Data: The data collected across all Tree Fall Sites show a consistent inverse relationship between the transect distance and light intensity. A temperature variation of approximately 1 to 2 degrees was observed among the three Tree Fall sites. Temperature consistently peaks at or near the gap epicenter (0 m) and gradually decreases or stays steady moving toward the shaded understory (15 m) across all three sites. Relative humidity generally trends upward moving away from the gap center into the shaded canopy, though values remain fairly narrow across the board (63.0% to 67.5%). Tree Falls 1 and 2 display minor U-shaped fluctuations, starting around 64–65% at 0 m, dropping slightly at mid-transect, and returning close to baseline at 15 m. Tree Fall 3 displays a noticeable positive trend, rising from 63.5% at 0 m to a peak of 67.5% at 10 m before settling at 66.5% at 15 m. Tree Falls 1 and 2 recorded an average temperature of 32ºC with light intensity values generally exceeding 210 lux. Tree Fall 3 recorded a lower average temperature of 30.5ºC with light values between 72-209 lux (Table 1, Figure 17.1- Figure 17.6).

Brix Data: The Brix value of leaves collected from the open canopy indicates higher sugar content compared to the Brix value of leaves collected from the shade where the canopy was more closed (Table 18). 

Lepidoptera Data: Over the three visits to the butterfly trap, 20 moths and 7 butterflies were collected. The total count for the trap in the shade was 17, which was greater than the total number of specimens collected in the sun, which was 10. The data for organism number showed no measurable trend and the shade trap was set up incorrectly for the first collection (human error). The aspect ratio stayed relatively constant, fluctuating tightly between 1.5 and 1.9  (Figure 18).

Discussion

The primary driver of understory dynamics following a gap disturbance is light availability. The exponential drop in ground cover growth and DBH secondary growth along the transect lines directly mirrors the attenuation of solar radiation (Lux) from 0 m to 15 m. The prediction that plant growth rates (ground cover, canopy cover, and DBH) peak at the gap center and decline along the gradient away from the trunk is strongly supported across sites. At 15 m, ground cover growth flattens to baseline levels, demonstrating a return to baseline forestry conditions. The Brix data supports the photosynthetic prediction. Leaves in open canopy conditions exhibited higher sugar levels (4º Bx) than those in shade (1º Bx), confirming elevated photosynthetic activity in response to increased light. Brix measures the approximate concentration of dissolved sugars by determining the refraction of light through a fluid sample. Although both samples produced relatively low Brix values, likely due to sample dilution from adding 5 mL of distilled water added to facilitate the extraction of leaf juice, the relative difference clearly indicates greater sugar concentrations in open-canopy foliage.

Microclimate data are essential for understanding ecological processes and predicting changes in biological responses. Microclimate predictions were partially aligned. Light intensity strongly correlated with proximity to the gap, but temperature (30º C – 32º C) and relative humidity (63%-68%) showed relatively narrow, subtle variations across transects rather than dramatic microclimate shifts. Across sites, higher temperatures at the gap epicenter (0 m) correspond with lower relative humidity readings, whereas cooler temperatures farther along the transect (10 m–15 m) align with higher relative humidity values, showing an inverse relationship between temperature and humidity. In contrast, light intensity showed a direct correlation with temperature. Across all transects, environmental shifts are remarkably subtle, with temperature varying by only 0.1°C to 1.0°C and humidity fluctuating by less than 4% along the entire 15-meter gradient. The data show that light availability is the primary microclimate variable changing along the transect, whereas temperature and relative humidity remain remarkably stable across the entire gradient.

However, methodological limitations impacted data consistency. Due to variances in equipment orientation at Tree Falls 1 and 2, data anomalies existed, skewing average values away from the median (Figure 17 and Figures 17.2, 17.4, and 17.6). Additionally, a ‘null’ value was recorded for transect point 2 at Tree Fall 3 because the PocketLab failed to obtain a reading after it was accidentally covered during data collection due to human error (Figure 17.6).

The prediction regarding insect distribution (more butterflies in sun versus more moths in shade), could not be validated due to low sample counts (20 moths, 7 butterflies total) and inconsistent trap setup. The shade trap yielded more total specimens (17) than the sun trap (10), but lack of standard placement heights complicates this result. Morphometric data showed that wing aspect ratios remained remarkably tight across individuals (1.5-1.9) despite variations in overall body size. This indicates that structural aerodynamic proportions in local lepidopterans may be constrained by functional flight constraints regardless of microhabitat or sun exposure.

These findings highlight the fundamental role of canopy gaps in driving secondary succession in the Amazonian understory. The massive explosion of ground cover (> 2,300%) at gap epicenters demonstrates how localized disturbances which increase light penetration can activate dormant seed banks and/or spur sapling growth. This rapid growth at gap epicenters serves as an engine for forest canopy turnover and long-term floral diversity in tropical ecosystems. However, our study was limited by brief and inconsistent spot-reading for microclimate data (PocketLab/micro:bit) rather than continuous diurnal logging, as well as minor human error in trap setup and sensor orientation and obstruction.

Next Steps

• Adjust Sensor Protocol: Ensure PocketLab and micro:bit sensors are mounted on unobstructed tripods to avoid “covered sensor” null data entries or ensure the instrument’s orientation is similar for all transect points. Calibrate the instruments and allow at least 2 minutes to acclimate to the new transect point. Note additional location observations like distance from nearby river.
• Standardize Trap Heights: Hang both the sun and shade butterfly traps at identical, standardized heights to eliminate vertical bias and ensure that the food plate is hanging the same distance beneath both traps.
• Re-deploy Continuous Loggers: Reestablish 24-hour HOBO light and temperature loggers at 0 m and 15 m marks across Tree Falls 1, 2, and 3 prior to leaving the field site.
• Consider Rapid Understory Growth Rate: Due to the rapid growth rate, access to sampling sites can be intimidating to research assistants new to the rainforest. Begin sampling at the site with the least groundcover; a new tree fall site if possible, to practice research protocols and build confidence before returning to overgrown sites.

Suggestions for further research:

• Soil Nutrient & Moisture Dynamics: Expand future ARIE cohorts to test soil chemistry (nitrogen, phosphorus, organic matter decomposition rate) along the transects to see if root death from the fallen tree creates a nutrient hotspot alongside the light gap.
• Decomposition Rates: Monitor the decomposition rate of the trunk itself.
• Track Canopy Closure Dynamics: Measure how many years it takes for neighboring lateral branches to re-close the canopy gap or a new tree fills the space, and identify the tipping point where ground cover growth halts due to returning shade.

References and Citations

Princeton University. (2016, January 7). In rainforests, battle for sunlight shapes forest structure: New finding sheds light on tropical forests’ influence on global climate. ScienceDaily. https://www.sciencedaily.com/releases/2016/01/160107185003.htm

2023: Field Report: Tree Fall Plot Biodiversity Study

2024: Field Report: Tree Fall Plot Biodiversity Study

2025: Field Report: Tree Fall/Epiphytes Survey

Classroom Connections

6–12 Interdisciplinary Science

• Curriculum Integration: Students adapt the ARIE field protocol using transects, sampling points (0 m, 5 m, 10 m, 15 m), quadrats, and sensor technology to design and conduct an independent field study measuring environmental gradients in their local urban or suburban community. Applying this ecological framework beyond the Amazon, students collect real-world data across local gradients such as:
  • Litter Density and Urban Waste: Measuring trash accumulation along a gradient extending from a high-traffic subway entrance or school cafeteria out toward a green park space.
  • Urban Heat Islands and Tree Prevalence: Tracking surface/air temperature and ambient light changes moving from an unshaded asphalt parking lot into a neighborhood with trees or a community garden.
  • Biodiversity and Edge Effects: Assessing soil moisture, ground cover, or insect diversity along a transect moving from a heavily trampled athletic field into an undisturbed woodland border.
  • Noise Pollution: Quantifying sound levels (decibels) along a gradient extending away from a busy roadway into a quiet residential block.
• NGSS Alignment:
  • MS/HS-ETS1-2: Evaluate competing design solutions for maintaining biodiversity and ecosystem services.
  • HS-ESS3-4: Evaluate or refine a technological solution that reduces impacts of human activities on natural systems.
  • SEP 3 (Planning and Carrying Out Investigations): Plan and conduct an investigation individually and collaboratively to produce data to serve as the basis for evidence.

High School Earth & Environmental Science (Grades 9–12)

• Curriculum Integration: Students analyze real field data from the ARIE Amazon Tree Plot Study to model secondary succession, light availability gradients, and microclimate dynamics. Using raw percent ground cover and DBH metrics, students calculate rates of change in plant biomass following gap disturbance events.
• NGSS Alignment:
  • HS-LS2-6: Evaluate claims, evidence, and reasoning that the complex interactions in ecosystems maintain relatively consistent numbers and types of organisms in stable conditions, but changing conditions may result in a new ecosystem.
  • HS-LS2-7: Design, evaluate, and refine a solution for reducing the impacts of human activities and natural hazards on the environment and biodiversity.

High School Biology & STEAM (Grades 9–12)

• Curriculum Integration: Connect the field study’s Brix leaf sugar data directly to unit lessons on cellular energetics and photosynthetic rates under varying light spectra. Integrating the ‘A’ in STEAM, students develop an Interactive Infographic Series mapping light attenuation (Lux) against physiological plant responses, or build a Digital Storytelling Archive featuring field audio, photo-layers of canopy gaps, and student-produced mini-podcasts on Amazonian ecosystem resilience.
• NGSS Alignment:
  • HS-LS1-5: Use a model to illustrate how photosynthesis transforms light energy into stored chemical energy.

Middle School Life Science & Ecology (Grades 6–8)

• Curriculum Integration: Adapt the ARIE quadrat and transect methodology into a localized schoolyard micro-disturbance lab. Students lay transects out from shaded building walls or tree canopy edges into sunny lawns, collecting real-time Lux, temperature, and relative humidity data via micro:bit or PocketLab sensors to observe physical microclimatic gradients.
• NGSS Alignment:
  • MS-LS2-1: Analyze and interpret data to provide evidence for the effects of resource availability on organisms and populations of organisms in an ecosystem.
  • MS-LS2-4: Construct an argument supported by empirical evidence that changes to physical or biological components of an ecosystem affect populations.

High School Mathematics & Data Science (Grades 9–12)

• Curriculum Integration: Utilize the scatter plots and bar graphs from Figures 8–16 (ground cover, canopy cover, and DBH scatter plots) for lessons in regression analysis, trendlines, negative correlations, and handling field data anomalies (such as sensor null values or edge-effect outliers).
• NGSS / CCSS Integration:
  • CCSS.MATH.CONTENT.HSS.ID.B.6: Represent data on two quantitative variables on a scatter plot, and describe how the variables are related.